OrcaFlex Manual

* The preview only display some random pages of manuals. You can download full content via the form below.

The preview is being generated... Please wait a moment!
  • Submitted by:
  • File size: 3.9 MB
  • File type: application/pdf
  • Words: 144,273
  • Pages: 429
Report / DMCA this file Add to bookmark

Description

w

 

 

 

                     

OrcaFlex  Manual     Version  9.4a               Orcina  Ltd.   Daltongate   Ulverston   Cumbria   LA12  7AJ   UK   Telephone:   Fax:   E-­‐mail:   Web  Site:  

+44  (0)  1229  584742   +44  (0)  1229  587191   [email protected]   www.orcina.com

1  

w

 

Contents  

 

CONTENTS   1   INTRODUCTION   1.1   1.2   1.3   1.4   1.5   1.6   1.7   1.8   1.9   1.10  

11  

Installing  OrcaFlex   Running  OrcaFlex   Parallel  Processing   Distributed  OrcaFlex   Orcina  Licence  Monitor   Demonstration  Version   OrcaFlex  Examples   Validation  and  QA   Orcina   References  and  Links  

11   13   14   15   15   15   15   16   16   16  

2   TUTORIAL   2.1   2.2   2.3   2.4   2.5   2.6   2.7   2.8   2.9   2.10  

21  

Getting  Started   Building  a  Simple  System   Adding  a  Line   Adjusting  the  View   Static  Analysis   Dynamic  Analysis   Multiple  Views   Looking  at  Results   Getting  Output   Input  Data  

21   21   21   22   22   23   23   24   24   24  

3   USER  INTERFACE   3.1  

3.2  

3.3  

25  

Introduction  

25  

3.1.1   3.1.2   3.1.3   3.1.4   3.1.5   3.1.6  

25   25   26   27   28   28  

Program  Windows   The  Model   Model  States   Toolbar   Status  Bar   Mouse  and  Keyboard  Actions  

OrcaFlex  Model  Files  

31  

3.2.1   3.2.2   3.2.3  

31   32   36  

Data  Files   Text  Data  Files   Simulation  Files  

Model  Browser  

37  

3.3.1   3.3.2  

39   39  

Model  Browser  Views   Move  Selected  Objects  Wizard  

3  

w

 

Contents    

3.4   3.5  

Libraries  

40  

3.4.1   3.4.2  

40   43  

Menus   3.5.1   3.5.2   3.5.3   3.5.4   3.5.5   3.5.6   3.5.7   3.5.8   3.5.9   3.5.10   3.5.11   3.5.12  

3.6  

3.7  

3.8   3.9  

Using  Libraries   Building  a  Library  

44   File  Menu   Edit  Menu   Model  Menu   Calculation  Menu   View  Menu   Replay  Menu   Graph  Menu   Results  Menu   Tools  Menu   Workspace  Menu   Window  Menu   Help  Menu  

44   45   46   47   48   49   49   50   50   50   51   51  

3D  Views  

52  

3.6.1   3.6.2   3.6.3   3.6.4   3.6.5   3.6.6   3.6.7   3.6.8   3.6.9   3.6.10  

53   53   54   55   56   58   58   58   58   59  

View  Parameters   View  Control   Navigating  in  3D  Views   Shaded  Graphics   How  Objects  are  Drawn   Selecting  Objects   Creating  and  Destroying  Objects   Dragging  Objects   Connecting  Objects   Printing,  Copying  and  Exporting  Views  

Replays  

59  

3.7.1   3.7.2   3.7.3   3.7.4   3.7.5  

60   60   61   61   63  

Replay  Parameters   Replay  Control   Custom  Replays   Custom  Replay  Wizard   Superimpose  Times  

Data  Forms  

63  

3.8.1   3.8.2  

64   64  

Data  Fields   Data  Form  Editing  

Results  

65  

3.9.1   3.9.2   3.9.3   3.9.4   3.9.5   3.9.6   3.9.7   3.9.8   3.9.9   3.9.10   3.9.11  

65   67   67   68   68   69   69   70   71   72   72  

Producing  Results   Selecting  Variables   Summary  and  Full  Results   Statistics   Linked  Statistics   Offset  Tables   Line  Clashing  Report   Time  History  and  XY  Graphs   Range  Graphs   Offset  Graphs   Spectral  Response  Graphs  

4  

w

 

Contents  

 

3.9.12   3.9.13  

Extreme  Statistics  Results   Presenting  OrcaFlex  Results  

72   75  

3.10   Graphs   3.10.1  

3.11   3.12   3.13   3.14   3.15   3.16  

76   Modifying  Graphs  

77  

Spreadsheets   Text  Windows   Workspaces   Comparing  Data   Preferences   Printing  and  Exporting  

78   78   78   79   80   82  

4   AUTOMATION   4.1   4.2  

4.3  

83  

Introduction   Batch  Processing  

83   83  

4.2.1   4.2.2   4.2.3   4.2.4   4.2.5   4.2.6   4.2.7   4.2.8   4.2.9  

83   85   85   85   88   93   93   93   96  

Introduction   Script  Files   Script  Syntax   Script  Commands   Examples  of  setting  data   Handling  Script  Errors   Obtaining  Variable  Names   Automating  Script  Generation   Automating  Text  Data  File  Generation  

Post-­‐processing   4.3.1   4.3.2   4.3.3   4.3.4   4.3.5   4.3.6   4.3.7   4.3.8   4.3.9   4.3.10  

97  

Introduction   OrcaFlex  Spreadsheet   Instruction  Format   Pre-­‐defined  commands   Basic  commands   Time  History  and  related  commands   Range  Graph  commands   Data  commands   Instructions  Wizard   Duplicate  Instructions  

5   THEORY   5.1   5.2   5.3   5.4   5.5  

5.6  

111  

Coordinate  Systems   Direction  Conventions   Object  Connections   Interpolation  Methods   Static  Analysis   5.5.1   5.5.2   5.5.3  

97   98   100   101   102   103   103   104   105   107  

111   112   113   113   115  

Line  Statics   Buoy  and  Vessel  Statics   Vessel  Multiple  Statics  

115   119   119  

Dynamic  Analysis  

120  

5  

w

 

Contents    

5.6.1   5.6.2  

5.7   5.8   5.9   5.10  

Calculation  Method   Ramping  

121   123  

Friction  Theory   Spectral  Response  Analysis   Extreme  Statistics  Theory   Environment  Theory   5.10.1   5.10.2   5.10.3   5.10.4   5.10.5   5.10.6  

123   126   127   129  

Buoyancy  Variation  with  Depth   Current  Theory   Seabed  Theory   Seabed  Non-­‐Linear  Soil  Model  Theory   Morison's  Equation   Waves  

5.11   Vessel  Theory   5.11.1   5.11.2   5.11.3   5.11.4   5.11.5   5.11.6   5.11.7  

129   129   130   131   137   138  

145  

Vessel  Rotations   RAOs  and  Phases   RAO  Quality  Checks   Hydrodynamic  and  Wind  Damping   Stiffness,  Added  Mass  and  Damping   Impulse  Response  and  Convolution   Wave  Drift  Loads  

145   146   147   149   151   152   153  

5.12   Line  Theory  

155  

5.12.1   5.12.2   5.12.3   5.12.4   5.12.5   5.12.6   5.12.7   5.12.8   5.12.9   5.12.10   5.12.11   5.12.12   5.12.13   5.12.14   5.12.15   5.12.16   5.12.17   5.12.18   5.12.19   5.12.20   5.12.21  

155   156   157   158   159   161   161   162   162   163   164   164   166   167   168   169   172   173   173   174   175  

Overview   Structural  Model  Details   Calculation  Stages   Calculation  Stage  1  Tension  Forces   Calculation  Stage  2  Bend  Moments   Calculation  Stage  3  Shear  Forces   Calculation  Stage  4  Torsion  Moments   Calculation  Stage  5  Total  Load   Line  End  Orientation   Line  Local  Orientation   Treatment  of  Compression   Contents  Flow  Effects   Line  Pressure  Effects   Pipe  Stress  Calculation   Pipe  Stress  Matrix   Hydrodynamic  and  Aerodynamic  Loads   Drag  Chains   Line  End  Conditions   Interaction  with  the  Sea  Surface   Interaction  with  Seabed  and  Shapes   Clashing  

5.13   6D  Buoy  Theory   5.13.1   5.13.2   5.13.3   5.13.4   5.13.5  

177  

Overview   Lumped  Buoy  Added  Mass,  Damping  and  Drag   Spar  Buoy  and  Towed  Fish  Added  Mass  and  Damping   Spar  Buoy  and  Towed  Fish  Drag   Contact  Forces  

6  

177   179   180   182   184  

w

 

Contents  

 

5.14   3D  Buoy  Theory   5.15   Winch  Theory   5.16   Shape  Theory  

185   186   187  

6   SYSTEM  MODELLING:  DATA  AND  RESULTS   6.1   6.2   6.3  

Modelling  Introduction   Data  in  Time  History  Files   Variable  Data   6.3.1  

6.4  

6.5  

External  Functions  

193  

194  

Statics   Dynamics   Integration  &  Time  Steps   Explicit  Integration   Implicit  Integration   Numerical  Damping   Response  Calculation   Properties  Report   Drawing   Results  

195   197   197   198   199   200   201   201   201   202  

Environment   6.5.1   6.5.2   6.5.3   6.5.4   6.5.5   6.5.6   6.5.7   6.5.8   6.5.9   6.5.10   6.5.11   6.5.12   6.5.13   6.5.14   6.5.15   6.5.16   6.5.17   6.5.18   6.5.19   6.5.20   6.5.21   6.5.22   6.5.23  

6.6   6.7  

189   190   192  

General  Data   6.4.1   6.4.2   6.4.3   6.4.4   6.4.5   6.4.6   6.4.7   6.4.8   6.4.9   6.4.10  

202  

Sea  Data   Sea  Density  Data   Seabed  Data   Wave  Data   Data  for  Regular  Waves   Data  for  Random  Waves   Data  for  JONSWAP  and  ISSC  Spectra   Data  for  Ochi-­‐Hubble  Spectrum   Data  for  Torsethaugen  Spectrum   Data  for  Gaussian  Swell  Spectrum   Data  for  User  Defined  Spectrum   Data  for  Time  History  Waves   Data  for  User  Specified  Components   Data  for  Response  Calculation   Waves  Preview   Modelling  Design  Waves   Setting  up  a  Random  Sea   Current  Data   Wind  Data   Drawing  Data   External  Functions   Results   Wave  Scatter  Conversion  

Solid  Friction  Coefficients  Data   Vessels   6.7.1  

189  

202   203   204   207   209   209   210   211   212   212   212   213   214   214   214   215   217   219   221   222   223   223   224  

228   229  

Vessel  Data  

230  

7  

w

 

Contents    

6.7.2   6.7.3   6.7.4   6.7.5  

6.8  

239   263   265   267  

Lines   6.8.1   6.8.2   6.8.3   6.8.4   6.8.5   6.8.6   6.8.7   6.8.8   6.8.9   6.8.10   6.8.11   6.8.12   6.8.13   6.8.14   6.8.15   6.8.16   6.8.17   6.8.18   6.8.19  

6.9  

Vessel  Types   Modelling  Vessel  Slow  Drift   Vessel  Response  Reports   Vessel  Results  

269   Line  Data   Line  Types   Attachments   Rayleigh  Damping   Line  Results   Drag  Chain  Results   Flex  Joint  Results   Line  Setup  Wizard   Line  Type  Wizard   Chain   Rope/Wire   Line  with  Floats   Homogeneous  Pipe   Hoses  and  Umbilicals   Modelling  Stress  Joints   Modelling  Bend  Restrictors   Modelling  non-­‐linear  homogeneous  pipes   Line  Ends   Modelling  Compression  in  Flexibles  

6D  Buoys   6.9.1   6.9.2   6.9.3   6.9.4   6.9.5   6.9.6   6.9.7   6.9.8   6.9.9   6.9.10   6.9.11   6.9.12   6.9.13   6.9.14   6.9.15   6.9.16   6.9.17   6.9.18   6.9.19  

343  

Wings   Common  Data   Applied  Loads   Wing  Data   Wing  Type  Data   Lumped  Buoy  Properties   Lumped  Buoy  Drawing  Data   Spar  Buoy  and  Towed  Fish  Properties   Spar  Buoy  and  Towed  Fish  Added  Mass  and  Damping   Spar  Buoy  and  Towed  Fish  Drag   Spar  Buoy  and  Towed  Fish  Drawing   Shaded  Drawing   Other  uses   External  Functions   Properties  Report   Results   Buoy  Hydrodynamics   Hydrodynamic  Properties  of  a  Rectangular  Box   Modelling  a  Surface-­‐Piercing  Buoy  

6.10   3D  Buoys   6.10.1   6.10.2   6.10.3  

271   286   296   300   303   314   315   315   316   317   322   325   329   331   333   335   337   339   342   344   345   347   347   348   350   351   352   354   355   356   356   358   358   358   359   361   362   364  

367  

Data   Properties  Report   Results  

368   369   369  

6.11   Winches  

370   8  

w

 

Contents  

 

6.11.1   6.11.2   6.11.3   6.11.4   6.11.5   6.11.6   6.11.7   6.11.8  

Data   Wire  Properties   Control   Control  by  Stage   Control  by  Whole  Simulation   Drive  Unit   External  Functions   Results  

371   371   372   372   373   373   374   374  

6.12   Links   6.12.1   6.12.2  

375   Data   Results  

375   377  

6.13   Shapes   6.13.1   6.13.2   6.13.3   6.13.4   6.13.5   6.13.6   6.13.7  

377   Data   Blocks   Cylinders   Curved  Plates   Planes   Drawing   Results  

378   379   380   381   382   382   383  

6.14   All  Objects  Data  Form  

383  

7   MODAL  ANALYSIS   7.1  

387  

Modal  Analysis  Theory  

388  

8   FATIGUE  ANALYSIS  

391  

8.1   8.2   8.3   8.4   8.5   8.6   8.7   8.8   8.9   8.10   8.11   8.12   8.13  

Commands   Data   Load  Cases  Data  for  Regular  Analysis   Load  Cases  Data  for  Rainflow  Analysis   Load  Cases  Data  for  Spectral  Analysis   Load  Cases  Data  for  SHEAR7   Components  Data   Analysis  Data   S-­‐N  and  T-­‐N  Curves   Integration  Parameters   Results   Fatigue  Points   How  Damage  is  Calculated  

9   VIV  TOOLBOX   9.1   9.2  

405  

Frequency  Domain  Models   9.1.1   9.1.2  

392   393   394   394   395   397   397   398   399   400   400   401   401  

405  

VIVA   SHEAR7  

405   410  

Time  Domain  Models  

417  

9  

w

 

Contents    

9.2.1   9.2.2   9.2.3  

Wake  Oscillator  Models   Vortex  Tracking  Models   VIV  Drawing  

420   423   429  

 

10  

w

 

Introduction,  Installing  OrcaFlex  

 

1

INTRODUCTION  

Welcome   to   OrcaFlex   (version   9.4a),   a   marine   dynamics   program   developed   by   Orcina   for   static   and   dynamic   analysis   of  a   wide   range   of  offshore   systems,   including   all   types   of   marine   risers   (rigid   and   flexible),   global   analysis,   moorings,  installation  and  towed  systems.   OrcaFlex  provides  fast  and  accurate  analysis  of  catenary  systems  such  as  flexible  risers  and  umbilical  cables  under   wave   and   current   loads   and   externally   imposed   motions.   OrcaFlex   makes   extensive   use   of   graphics   to   assist   understanding.   The   program   can   be   operated   in   batch   mode   for   routine   analysis   work   and   there   are   also   special   facilities  for  post-­‐processing  your  results  including  fully  integrated  fatigue  analysis  capabilities.   OrcaFlex   is   a   fully   3D   non-­‐linear   time   domain   finite   element   program   capable   of   dealing   with   arbitrarily   large   deflections   of  the   flexible   from  the   initial   configuration.   A   lumped   mass   element   is   used   which   greatly   simplifies   the   mathematical   formulation   and   allows   quick   and   efficient   development   of   the   program   to   include   additional   force   terms  and  constraints  on  the  system  in  response  to  new  engineering  requirements.   In   addition   to   the   time   domain   features,   modal   analysis   can   be   performed   for   individual   lines   and   RAOs   can   be   calculated  for  any  results  variable  using  the  Spectral  Response  Analysis  feature.   OrcaFlex   is   also   used   for   applications   in   the   Defence,   Oceanography   and   Renewable   energy   sectors.   OrcaFlex   is   fully   3D   and   can   handle   multi-­‐line   systems,   floating   lines,   line   dynamics   after   release,   etc.   Inputs   include   ship   motions,   regular  and  random  waves.  Results  output  includes  animated  replay  plus  full  graphical  and  numerical  presentation.   If  you  are  new  to  OrcaFlex  then  please  see  the  tutorial  and  examples.   For  further  details  of  OrcaFlex  and  our  other  software,  please  contact   Orcina  or  your  Orcina  agent.   Copyright  notice  

Copyright  Orcina  Ltd.  1987-­‐2010.  All  rights  reserved.  

1.1

INSTALLING  ORCAFLEX  

Hardware  Requirements   OrcaFlex  can  be  installed  and  run  on  any  computer  that  has:   x

Windows  XP,  Windows  Vista  or  Windows  7.  Both  32  bit  and  64  bit  versions  of  Windows  are  supported.  

x

If  you  are  using  small  fonts  (96dpi)  the  screen  resolution  must  be  at  least  1024×768.  If  you  are  using  large  fonts   (120dpi)  the  screen  resolution  must  be  at  least  1280×1024.  

However,  OrcaFlex  is  a  powerful  package  and  to  get  the  best  results  we  would  recommend:   x

A   powerful   processor   with  fast   floating   point   and   memory   performance.   This   is   the   most   important   factor   since   OrcaFlex  is  a  computation-­‐intensive  program  and  simulation  run  times  can  be  long  for  complex  models.  

x

At   least   2GB   of   memory.   This   is   less   important   than   processor   performance   but   some   aspects   of   OrcaFlex   do   perform   better   when   more   memory   is   available,   especially   on   multi-­‐core   systems.   If   you   have   a   multi-­‐core   system  with  a  64  bit  version  of  Windows  then  you  may  benefit  from  fitting  even  more  memory.  

x

A  multi-­‐core  system  to  take  advantage  of  OrcaFlex's  multi-­‐threading  capabilities.  

x

As   much   disk   space   as   you   require   to   store   simulation   files.   Simulation   files   vary   in   size,   but   can   be   100's   of   megabytes  each  for  complex  models.  

x

A  screen  resolution  of  1280×1024  or  greater  with  32  bit  colour.  

x

A   DirectX   9   compatible   graphics   card   with   at   least   256MB   memory   for   the   most   effective   use   of   the   shaded   graphics  facility.  

x

Microsoft   Excel   (Excel   2000,   or   later)   in   order   to   use   the  OrcaFlex   automation   facilities.   This   requires   the   32   bit   version  of  Excel.   Note:  

Although  OrcaFlex  is  a  32  bit  program,  the  64  bit  versions  of  Windows  run  32  bit  programs  very   efficiently  and   have  certain   advantages   over  32   bit   versions  of  Windows.  Most  notably  the  64  bit   versions   of   Windows   are   able   to   make   use   of   larger   amounts   of   memory.   This   can   benefit   OrcaFlex,   and  indeed  other  programs.  In  addition  we  have  found  the  64  bit  versions  of  Windows  to  be  more   effective   at   multi-­‐threaded   calculations.   For  these   reasons   we   currently   recommend   64  bit   Vista/7   as  the  best  platforms  for  running  OrcaFlex.  

11  

w

 

Introduction,  Installing  OrcaFlex    

Installation   To  install  OrcaFlex:   x

You  will  need  to  install  from  an  account  with  administrator  privileges.  

x

If   installing   from   CD,   insert   the   OrcaFlex   CD   and   run   the   Autorun.exe   program   on   the   CD   (on   many   machines   this  program  will  run  automatically  when  you  insert  the  CD).  Then  click  on  'Install  OrcaFlex'.  

x

If   you   have   received   OrcaFlex   by   e-­‐mail   or   from   the   web   you   will   have   a   zip   file,   and   possibly   a   number   of   licence  files   (.lic).   Extract   the   files   from   the   zip   file   to   some  temporary  location,   and  save   the   licence  files   to   the   same  folder.  Then  run  the  extracted  file  Setup.exe.  

x

You  will  also  need  to  install  the  OrcaFlex  dongle  supplied  by  Orcina.  See  below  for  details.  

For   further   details,   including   information   on   network   and   silent   installation,   click   on   Read   Me   on   the   Autorun   menu   or  open  the  file  Installation  Guide.pdf  on  the  CD.  If  you  have  any  difficulty  installing  OrcaFlex  please  contact  Orcina   or  your  Orcina  agent.   Orcina  Shell  Extension   When  you  install  OrcaFlex  the  Orcina  Shell  Extension  is  also  installed.  This  integrates  with  Windows  Explorer,  and   associates   the   data   and   simulation   file   types   (.dat   and   .sim)   with   OrcaFlex.   You   can   then   open   an   OrcaFlex   file   by   simply  double-­‐clicking  the  filename  in  Explorer.  The  shell  extension  also  provides  file  properties  information,  such   as  which  version  of  OrcaFlex  wrote  the  file  and  the  Comments  text  for  the  model  in  the  file.  For  details  see  the  file   CD:\OrcShlEx\ReadMe.htm  on  the  OrcaFlex  CD.   Installing  the  Dongle   OrcaFlex  is  supplied   with  a   dongle,  a  small  hardware  device  that  must   be  attached  to  the  machine  or  to  the  network   to  which  the  machine  is  attached.   Note:  

The  dongle   is  effectively  your  licence  to  run  one  copy  (or  more,   if  the  dongle  is  enabled  for   more   copies)  of  OrcaFlex.  It  is,  in  essence,  what  you  have  purchased  or  leased,  and  it  should  be  treated   with  appropriate  care  and  security.  If  you  lose  your  dongle  you  cannot  run  OrcaFlex.  

Warning:  

Orcina  can  normally  resupply  disks  or  manuals  (a  charge  being  made  to  cover  costs)  if  they  are  lost   or  damaged.  But  we  can  only  supply  a  new  dongle  in  the  case  where  the  old  dongle  is  returned  to   us.  

Dongles  labelled  'Hxxx'  (where  xxx  is  the  dongle  number)  must  be  plugged  into  the  machine  on  which  OrcaFlex  is   run.   Dongles   labelled   'Nxxx'   can   be   used   in   the   same   way   as   'Hxxx'   dongles,   but   they   can   also   be   used   over   a   network,  allowing  the  program  to  be  shared  by  multiple  users.  In  the  latter  case  the  dongle  should  be  installed  by   your  network  administrator;  instructions  can  be  found  in  the  Dongle  directory  on  the  OrcaFlex  CD.   Types  of  Dongle  

Dongles   are   available   for   either   parallel   or   USB   ports,   and   these   are   functionally   equivalent   so   far   as   OrcaFlex   is   concerned.  In  general,  USB  dongles  are  preferred,  since  they  seem  to   be   more  reliable.  In  any  case,  parallel  ports   are   becoming   less   common  on  new  machines.   By  default,   'N'   dongles   can   hold   up   to   10  OrcaFlex   licences   for   use   over   a   network.  We  can  supply  dongles  with  larger  capacities  on  request.   Dongle  Troubleshooting   We  supply,  with  OrcaFlex,  a  dongle  utility  program  called  OrcaDongle.  If  OrcaFlex  cannot  find  the  dongle  then  this   program  may   be  used  to  check  that  the  dongle  is   working  correctly  and  has  the  expected   number  of  licences.   For   details  see  the  OrcaDongle  help  file.   The  OrcaDongle  program  is   included  on  the  OrcaFlex  CD,  and  you  may  choose  to   install  it  from  the  Autorun  menu  in   the  same  way  as  OrcaFlex.  It  is  also  available  for  download  from   www.orcina.com/Support/Dongle.   Also   on   our   website,   users   of   network   dongles   may   find   the   Orcina   Licence   Monitor   to   be   useful.   This   application   keeps  track  of  the  number  of  OrcaFlex  licences  claimed  on  a  network  at  any  time.   Diagnostics  

If  OrcaFlex  fails  to  start,  with  the  error  that  it  can't  obtain  a  licence,  then  please  check  the  following.     x

If   you   are   using   a   network   dongle,   are   all   the   licences   in   use?   The   Orcina   Licence   Monitor   may   be   of   use   in   determining  this.  If  they  are,  you  will  need  to  wait  until  a  licence  becomes  free  before  you  can  run  OrcaFlex.  

12  

w

 

Introduction,  Running  OrcaFlex  

 

x

If  you  are  using  a  local  dongle,  is  it  plugged  into  your  machine?  If  so,  is   the  dongle  device  driver  installed?  You   can   check   this   by   running   OrcaDongle.   If   the   driver   is   not   present,   it   may   have   been   uninstalled   by   another   program:  if  so,  you  can  fix  this  by  Repairing  the  OrcaFlex  installation  (from  the  Windows  Control  Panel,  select   'Add  or  Remove  Programs'  (XP)  or  Programs  /  Programs  and  Features  (Vista),  select  the  OrcaFlex  entry,  select   Change   then   Repair).   If   this  still   fails,   you   can   install   the   driver  by  downloading  from   our   website,   and  running,   the  file  Hasp-­‐Setup.msi.  

x

Does  the  dongle  you  are  using  have  an  OrcaFlex  licence  on  it?  Again,  you  can  check  this  with  OrcaDongle.  

x

Do   you   have   a   licence   file   for   the   dongle   you  wish  to   access?   This  file  will   be  named  Nxxx.lic   or  Hxxx.lic   (where   xxx   is   the   dongle   number)   and   will   be   in   the   OrcaFlex   installation   folder.   If   not,   then   you   should   be   able   to  copy   the  required  file(s)  from  the  root  level  of  the  OrcaFlex  CD  into  the  installation  folder.  

If  none  of  these  help,  then   please  contact  us  at  Orcina  with  a  description  of  the  problem.  Ideally,  please  also  email   to   us  the  diagnostics  file   named  OrcLog.txt  which  OrcaFlex  will  have  written  on  failing   to  find  a  licence.   This  file  can  be   found   in   the   folder   "%appdata%/Orcina/OrcaFlex":   to   open   this  folder,   select   Start  menu  |   Run...   and  enter   the  text   between  the  quotes  (including  the  '%'  characters).    

1.2

RUNNING  ORCAFLEX  

 A   shortcut   to   run   OrcaFlex   is   set   up   on   the   Start   menu   when   you   install   OrcaFlex   (see   Start\Programs\Orcina   Software\).   This  shortcut  passes  no   parameters  to  OrcaFlex  so  it  gives  the  default  start-­‐up  behaviour;  see  below.  If  this  is   not   suitable  you  can  configure  the  start-­‐up  behaviour  using  command-­‐line  parameters,  for  example  by  setting  up  your   own  shortcuts  with  particular  parameter  settings.   Default  Start-­‐up   OrcaFlex   has   two   basic   modules:   full   OrcaFlex   and   statics-­‐only   OrcaFlex.   A   full   OrcaFlex   licence   is   needed   for   dynamic  analysis.   When   you   run   OrcaFlex   it   looks   for   an   Orcina   dongle   from   which   it   can   claim   an   OrcaFlex   licence   (either   a   full   licence  or  a  statics-­‐only  licence).  By  default,  it  first  looks  for  a  licence  on  a  local  dongle  (i.e.  one  in  local  mode  and   connected   to   the   local   machine)   and   if   none   is   found   then   it   looks   for   a   licence   on   a   network   dongle   (i.e.   one   in   network   mode   and   accessed   via   a   licence   manager   over   the   network).   This   default   behaviour   can   be   changed   by   command-­‐line  parameters.   If   OrcaFlex   finds  a  network  dongle   and   there   is   a   choice   of  which  licences   to   claim  from   it,  then  OrcaFlex   displays   a   Choose  Modules  dialog  to  ask  you  which  modules  you  want  to  claim.  This  helps  you  share  the  licences  with  other   users   of   that   network   dongle.   For   example   if   the   network   dongle   contains   both   a   full   licence   and   a   statics-­‐only   licence  then  you  can  choose  to  use  the  statics-­‐only  licence,  if  that  is  all  you  need,  so  that  the  full  licence  is  left  free  for   others   to   use  when   you  do  not   need   it   yourself.   The   Choose   Modules   dialog   can   be  suppressed   using  command-­‐line   parameters.   Command  Line  Parameters   OrcaFlex  can  accept  various  parameters  on  the  command  line  to  modify  the  way  it  starts  up.  The  syntax  is:   OrcaFlex.exe  Filename  Option1  Option2  ...  etc.   Filename  is  optional.  If  present  it  should  be  the  name  of  an  OrcaFlex  data  file  (.dat  or  .yml)  or  simulation  file  (.sim)   and  after  starting  up  OrcaFlex  will  automatically  open  that  file.   Option1,  Option2  etc.  are  optional  parameters  that  allow  you  configure  the  start-­‐up  behaviour.  They  can  be  any  of   the   following   switches.   For   the   first   character   of   an   option   switch,   the   hyphen   character   '-­‐'   can   be   used   as   an   alternative  to  the  '/'  character.   Dongle  Search  switches  

By   default   the   program   searches   first   for   a   licence   on   a   local   dongle   and   then   for   a   licence   on   a   network   dongle.   The   following  switches  allow  you  to  modify  this  default  behaviour.   x

/LocalDongle  Only  search  for  licences  on  a  local  dongle.  No  search  will  be  made  for  network  dongles.  

x

/NetworkDongle   Only   search   for   licences   on   a   network   dongle.   Any   local   dongle   will   be   ignored.   This   can   be   useful  if  you  have  a  local  dongle  but  want  to  use  a  network  dongle  that  has  licences  for  more  modules.  

13  

w

 

Introduction,  Parallel  Processing    

Module  Choice  switch  

This  switch  is  only  relevant  if  the  dongle  found  is  a  network  dongle   and  there  is  a  choice  of  licences  to  claim  from   that  dongle.  You  can  specify  your  choice  using  the  following  command  line  switch:   x

/DisableDynamics   Choose   the   statics-­‐only  basic   licence.   This  is  sometimes   useful   when   using   a   network   dongle   since  it  allows  you  to  leave  full  licences  free  for  other  users  when  you  only  need  a  statics-­‐only  licence.  

If   you   do   not   specify   all   the   choices   then   the   program   displays   the  Choose   Modules  dialog   to   ask   for   your   remaining   choices.  You  can  suppress  this  dialog  using  the  following  switch.   x

/DisableInteractiveStartup  Do  not  display  the  Choose  Modules  dialog.  The  program  behaves  the  same  as  if  the   user  clicks  OK  on  that  dialog  without  changing  any  module  choices.  

Batch  Calculation  switches  

These  switches  allow  you  to  instruct  OrcaFlex  to  start   a   batch  calculation  as  soon  as  the  program  has  loaded.  The   following  switches  are  available:   x

/Batch   Start   a   batch   calculation   as   soon   as   the   program   has   loaded.   The   batch   calculation   will   contain   all   the   files   specified   on  the   command   line   (you   can   have   more  than   one)   in   the   order   in   which  they  are   specified.   You   can  use  relative  paths  which  will  be  relative  to  the  working  directory.  

x

/CloseAfterBatch  Instructs  the  program  to  close  once  the  batch  is  complete.  

x

/BatchAnalysisStatics,  /BatchAnalysisDynamics   specify   what   type   of  analysis   to   perform   to   the   specified   files.   If   these  parameters  are  missing  then  the  program  defaults  to  dynamic  analysis.  

Process  Priority  switches  

These   switches   determine   the   processing   priority   of   OrcaFlex.   The   available   switches   are   /RealtimePriority,   /HighPriority,  /AboveNormalPriority,  /NormalPriority,  /BelowNormalPriority,  /LowPriority.   ThickLines  switch  

The  /ThickLines  switch  allows  you  to  specify  a  minimum  thickness  for  lines  drawn  on  OrcaFlex  3D  View  windows.   For   example   using   the   switch   /ThickLines=5   forces   OrcaFlex   to   draw   all   lines   at   a   thickness   of   at   least   5.   If   no   value   is  specified  (i.e.  the  switch  is  /ThickLines)  then  the  minimum  thickness  i s  taken  to  be  2.   This  switch  has  been  added  to  make  OrcaFlex  3D  Views  clearer  when  projected  onto  a  large  screen.   ThreadCount  switch  

The   /ThreadCount   switch   allows   you   to   set   the   number   of   execution   threads   used   by   OrcaFlex   for   parallel   processing.   For   example   /ThreadCount=1   forces   OrcaFlex   to   use   a   single   execution   thread   which   has   the   effect   of   disabling  parallel  processing.  

1.3

PARALLEL  PROCESSING  

Machines   with  multiple   processors   or   processors  with  multiple   cores   are  becoming   increasingly  common.   OrcaFlex   can  make  good  use  of  the  additional  processing  capacity  afforded  by  such  machines.  For  up  to  date  information  on   hardware  choice  for  OrcaFlex  please  refer  to  www.orcina.com/Support/Benchmark.   OrcaFlex   performs   the   calculations   of   the   model's   Line   objects   in   parallel.   This   means   that,   interactively   at   least,   performance   is   only   improved   for   models   with   more   than   one   Line   object.   However,   for   models   with   more   than   one   Line  performance  is  significantly  improved.   Both   batch   processing   and   fatigue   calculations   process   their   jobs   and   load   cases   concurrently,   using   all   available   processor  cores.   Note,   however,   that   the   OrcaFlex   spreadsheet   is   currently   only   able   to   make   use   of  a   single   processor   core.   We   plan   to  address  this  limitation  in  a  future  release.   Thread  count  

OrcaFlex  manages  a  number  of  execution  threads  to  perform  the  parallel  calculations.  The  number  of  these  threads   (the   thread   count)   defaults   to   the   number   of  physical   processor   cores   available   on   your   machine   as   reported   by  the   operating  system.  This  default  will  work  well  for  most  cases.  Should  you  wish  to  change  it  you  can  use  the  Tools  |  Set   Thread  Count  menu  item.  The  thread  count  can  also  be  controlled  by  a  command  line  switch.  

14  

w

 

Introduction,  Distributed  OrcaFlex  

 

Hyperthreading  

Some   Intel   processors   offer   a   technology   called   hyperthreading.   Such   processors   can   process   multiple   execution   threads  in  parallel  by  making  use  of  under-­‐used  resources  on   the  processor.  Hyperthreaded  processors  appear  to   the  operating  system  as  2  distinct,  logical  processors.   Sadly,  the  real   world   performance   of   such  chips   does  not   live   up   t o  the  marketing  hype.   At   best   this  technology  can   give   improvements   of   around   10-­‐20%.   However,   the   performance   of   hyperthreading   under   OrcaFlex   varies   considerably   with   the   OrcaFlex   model   being   analysed.   In   the   worst   cases   using   hyperthreading   results   in   performance  twice  as  slow  as  without!   For   this   reason   we   recommend   that   you   don't   attempt   to   use   hyperthreading   when   running   OrcaFlex.   By   default   OrcaFlex  will  use  as  many  threads  as  there  are  true  physical  cores  available  to  your  system.   To  help  understand  this  consider  a  dual  processor,  dual  core  machine  with  hyperthreading  support.  The  operating   system   will   recognise   8   processors.   Of   these   processors,   4   are   true   physical   processor   cores   and   the   other   4   are   virtual  hyperthreaded  processors.  Accordingly  OrcaFlex  will  default  to  using  4  calculation  threads.  

1.4

DISTRIBUTED  ORCAFLEX  

Distributed  OrcaFlex  is  a  suite  of  programs  that  enables  a  collection  of  networked,  OrcaFlex  licensed  computers  to   run   OrcaFlex   jobs,   transparently,   using   spare   processor   time.   For   more   information   about   Distributed   OrcaFlex   please   refer   to   www.orcina.com/Support/DistributedOrcaFlex.   Distributed   OrcaFlex   can   be   downloaded   from   this   address.   OrcaFlex  can  also  make  use  of  machines  with  multiple  processors  using  parallel  processing  technology.  

1.5

ORCINA  LICENCE  MONITOR  

The  Orcina  Licence  Monitor  (OLM)  is  a   service  that  monitors  the  current  number  of  OrcaFlex  licences  claimed  on  a   network  in  real  time.  Other  programs  that  use  the  OrcaFlex  programming  interface  (OrcFxAPI)  such  as  Distributed   OrcaFlex  and  the   OrcaFlex  spreadsheet  are  also  monitored.   You  can  obtain  information  on   each  licence  claimed  that   includes:   x

Network  information:  the  computer  name,  network  address  and  the  user  name.  

x

Licence  information:  the  dongle  name,  the  dongle  type  (network  or  local)  and  the  time  the  licence  was  claimed.  

x

Program   information:   which   modules   are   being   used,   the   version,   and   the   location   of   the   program   which   has   claimed   the   licence   (usually   this   is   OrcaFlex.exe   but   it   can   be   Excel.exe   for   the   OrcaFlex   spreadsheet   for   example).  

OLM  can  be  downloaded  from  www.orcina.com/Support/OrcinaLicenceMonitor.  

1.6

DEMONSTRATION  VERSION  

For  an  overview  of  OrcaFlex,  see  the  Introduction  topic  and  the  tutorial.   The   demonstration   version   of  OrcaFlex   has   some   facilities   disabled  Ȃ   you   cannot   calculate   statics   or   run   simulation,   and  you  cannot  save  files,  print,  export  or  copy  to  the  clipboard.  Otherwise  the  demonstration  version  is  just  like  the   full  version,  so  it  allows  you  to  see  exactly  how  the  program  works.   In   particular   the   demonstration   version   allows   you   to   open   any   prepared   OrcaFlex   data   or   simulation   file.   If   you   open   a   simulation   file   then   you   can   then   examine   the   results,   see   replays   of   the   motion   etc.   There   are   numerous   example   files   provided   on   the   demonstration   DVD.   These   example   files   are   also   available   from   www.orcina.com/SoftwareProducts/OrcaFlex/Examples.   If  you  have  the  full  version  of  OrcaFlex  then  you  can  use  the  demonstration  version  to  show  your  customers  your   OrcaFlex  models  and  results  for  their  system.  To  do  this,  give  them  the  demonstration  version  and  copies  of  your   OrcaFlex   simulation   files.   The   demonstration   version   can   be   downloaded   from   www.orcina.com/SoftwareProducts/OrcaFlex/Demo.  

1.7

ORCAFLEX  EXAMPLES  

OrcaFlex  is  supplied   with  a   DVD  containing  a  comprehensive  collection   of  example  files.  These   examples  can  also  be   found  at  www.orcina.com/SoftwareProducts/OrcaFlex/Examples.  

15  

w

 

Introduction,  Validation  and  QA    

1.8

VALIDATION  AND  QA  

The  OrcaFlex  validation  documents  are  available  from  www.orcina.com/SoftwareProducts/OrcaFlex/Validation.  

1.9

ORCINA  

Orcina   is   a   creative   engineering   software   and   consultancy   company   staffed   by   mechanical   engineers,   naval   architects,   mathematicians   and   software   engineers   with   long   experience   in   such   demanding   environments   as   the   offshore,   marine   and   nuclear   industries.   As   well   as   developing   engineering   software,   we   offer   a   wide   range   of   analysis   and   design   services   with   particular   strength   in   dynamics,   hydrodynamics,   fluid   mechanics   and   mathematical  modelling.   Contact  Details   Orcina  Ltd.   Daltongate   Ulverston   Cumbria   LA12  7AJ   UK   Telephone:  +44  (0)  1229  584742   Fax:  +44  (0)  1229  587191   E-­‐mail:  [email protected]   Web  Site:  www.orcina.com   Orcina  Agents   We  have  agents  in  many  parts  of  the  world.  For  details  please  refer  to   www.orcina.com/ContactOrcina.  

1.10

REFERENCES  AND  LINKS  

References   API,   1993.   API   RP   2A-­‐WSD,   Recommended   Practice   for   Planning,   Designing   and   Constructing   Fixed   Offshore   Platforms  Ȅ  Working  Stress  Design.  American  Petroleum  Institute.   API,   1998.   API   RP   2RD,   Design   of   Risers   for   Floating   Production   Systems   and   Tension-­‐Leg   Platforms.   American   Petroleum  Institute.   API,  2005.  API  RP  2SK,  Design  and  Analysis  of  Stationkeeping  Systems  for  Floating  Structures.   American  Petroleum   Institute.   API.  Comparison  of  Analyses  of  Marine  Drilling  Risers.  API  Bulletin.  2J.   Aubeny   C,   Biscontin   G   and   Zhang   J,   2006.   Seafloor   interaction   with   steel   catenary   risers.   Offshore   Technology   Research  Center  (Texas  A&M  University)  Final  Project  Report  (http://www.mms.gov/tarprojects/510.htm).   Aubeny  C,  Gaudin  C  and  Randolph  M,  2008.  Cyclic  Tests  of  Model  Pipe  in  Kaolin.  OTC  19494,  2008.   Barltrop   N   D   P   and   Adams   A   J,   1991.   Dynamics   of   fixed   marine   structures.   Butterworth   Heinemann   for   MTD.   3rd   Edition.   Batchelor  G  K,  1967.  An  introduction  to  fluid  dynamics.   Cambridge  University  Press.   Blevins  R  D,  2005.  Forces  on  and  Stability  of  a  Cylinder  in  a  Wake.  J.  OMAE,  127,    39-­‐45.   Bridge  C,  Laver  K,  Clukey  E,  Evans  T,  2004.  Steel  Catenary  Riser  Touchdown  Point  Vertical  Interaction  Models.   OTC   16628,  2004.   Carter  D  J  T,  1982.  Prediction  of  Wave  height  and  Period  for  a  Constant  Wind  Velocity  Using  the  JONSWAP  Results,   Ocean  Engineering,  9,    no.  1,  17-­‐33.   Casarella   M   J   and   Parsons   M,   1970.   Cable   Systems   Under   Hydrodynamic   Loading.  Marine   Technology   Society  Journal   4,  No.  4,  27-­‐44.   Chapman  D  A,  1984.  Towed  Cable  Behaviour  During  Ship  Turning  Manoeuvres.  Ocean  Engineering.  11,  No.  4.   Chung   J   and   Hulbert   G   M,   1993.   A   time   integration   algorithm   for   structural   dynamics   with   improved   numerical   dissipation:  The  generalized-­‐Ƚ‡–Š‘†ǤASME  Journal  of  Applied  Mechanics.  60,  371-­‐375.  

16  

w

 

Introduction,  References  and  Links  

 

CMPT,  1998.  Floating  structures:  A  guide  for  design  and  analysis.  Edited  by  Barltrop  N  D  P.   Centre  for  Marine  and   Petroleum  Technology  publication  101/98,  Oilfield  Publications  Limited.   Coles  S,  2001.  An  Introduction  to  Statistical  Modelling  of  Extreme  Values.  Springer.   Cummins  W  E,  1962.  The  impulse  response  function  and  ship  motions.  Schiffstechnik,  9,  101-­‐109.   Dean  R  G,  1965.  Stream  function  representation  of  non-­‐linear  ocean  waves.  J.  Geophys.  Res.,  70,  4561-­‐4572.   Dirlik  T,  1985.  Application  of  computers  in  Fatigue  Analysis.  PhD  Thesis  University  of  Warwick.   DNV-­‐OS-­‐F201,  Dynamic  Risers.   DNV-­‐RP-­‐C205,  Environmental  Conditions  and  Environmental  Loads.   ESDU  71016.  Fluid  forces,  pressures  and  moments  on  rectangular  blocks.  ESDU  71016   ESDU  International,  London.   ESDU   80025.   Mean   forces,   pressures   and   flow   field   velocities   for   circular   cylindrical   structures:   Single   cylinder   with   two-­‐dimensional  flow.  ESDU  80025  ESDU  International,  London.   Falco  M,  Fossati  F  and  Resta  F,  1999.  On  the  vortex  induced  vibration  of  submarine  cables:  Design  optimization  of   wrapped  cables  for  controlling  vibrations.  3rd  International  Symposium  on  Cable  Dynamics,  Trondheim,  Norway.   Faltinsen  O  M,  1990.  Sea  loads  on  ships  and  offshore  structures.  Cambridge  University  Press.   Fenton  J  D,  1979.  A  high-­‐order  cnoidal  wave  theory.  J.  Fluid  Mech.  94,  129-­‐161.   Fenton  J  D,  1985.  A  fifth-­‐order  Stokes  theory  for  steady  waves.   J.  Waterway,  Port,  Coastal  &  Ocean  Eng.  ASCE.  111,   216-­‐234.   Fenton   J  D,   1990.  Non-­‐linear   wave   theories.   Chapter   in   "The   Sea  Ȃ  Volume   9:   Ocean  Engineering   Science",  edited   by   B.  Le  MeHaute  and  D.  M.  Hanes.  Wiley:  New  York.    3-­‐25.   Fenton  J  D,  1995.  Personal  communication  Ȃ  pre-­‐print  of  chapter  in  forthcoming  book  on  cnoidal  wave  theory.   Gregory   R   W   and   Paidoussis   M   P,   1996.   Unstable   oscillation   of   tubular   cantilevers   conveying   fluid:   Part   1:Theory.   Proc.  R.  Soc.293  Series  A,  512-­‐527.   Hartnup  G  C,  Airey  R  G  and  Fraser  J  M,  1987.  Model  Basin  Testing  of  Flexible  Marine  Risers.  OMAE  Houston.   Hoerner  S  F  1965.  Fluid  Dynamic  Drag,  Published  by  the  author  at  Hoerner  Fluid  Dynamics,  NJ  08723,  USA.   Huse  E,  1993.  Interaction  in  Deep-­‐Sea  Riser  Arrays.  OTC  7237,  1993.   Isherwood   R   M,   1987.   A   Revised   Parameterisation   of   the   JONSWAP   Spectrum.   Applied   Ocean   Research,   9,   No.   1   (January),  47-­‐50.   Iwan  W  D,  1981.  The  vortex-­‐induced  oscillation  of  non-­‐uniform  structural  systems.  Journal  of  Sound  and  Vibration,   79,  291-­‐301.   Iwan   W   D   and   Blevins   R   D,   1974.   A   Model   for   Vortex   Induced   Oscillation   of  Structures.   Journal   of   Applied   Mechanics,   September  1974,  581-­‐586.   Kotik  J  and  Mangulis  V,  1962.  On  the  Kramers-­‐Kronig  relations  for  ship  motions.  Int.  Shipbuilding  Progress,  9,  No.  97,   361-­‐368.   Larsen   C   M,   1991.   Flexible   Riser   Analysis   Ȃ   Comparison   of   Results   from   Computer   Programs.   Marine   Structures,   Elsevier  Applied  Science.   Longuet-­‐Higgins   M   S,   1983.   On   the   joint   distribution   of   wave   periods   and   amplitudes   in   a   random   wave   field.   Proceedings  Royal  Society  London,  Series  A,  Mathematical  and  Physical  Sciences.389,  241-­‐258.   Maddox  S  J,  1998.  Fatigue  strength  of  welded  structures.  Woodhead  Publishing  Ltd,  ISBN  1  85573  013  8.   Morison   J   R,   O'Brien   M   D,   Johnson   J   W,   and   Schaaf   S   A,   1950.   The   force   exerted   by   surface   waves   on   piles.   Petrol   Trans  AIME.  189.   Mueller   H   F,   1968.   Hydrodynamic   forces   and   moments   of   streamlined   bodies   of   revolution   at   large   incidence.   Schiffstechnik.  15,  99-­‐104.   Newman   J   N.   1974.   Second-­‐order,   slowly-­‐varying   forces   on   vessels   in   irregular   waves.   Proc   Int   Symp   Dynamics   of   Marine  Vehicles  and  Structures  in  Waves,  Ed.  Bishop  RED  and  Price  WG,  Mech  Eng  Publications  Ltd,  London.   Newman  J  N,  1977.  Marine  Hydrodynamics,  MIT  Press.  

17  

w

 

Introduction,  References  and  Links    

NDP,   1995.   Regulations   relating   to   loadbearing   structures   in   the   petroleum   activities.   Norwegian   Petroleum   Directorate.   Ochi  M  K  and  Hubble  E  N,  1976.  Six-­‐parameter  wave  spectra,  Proc  15th  Coastal  Engineering  Conference,  301-­‐328.   Ochi  M  K,  1973.  On  Prediction  of  Extreme  Values,  J.  Ship  Research,  17,  No.  1,  29-­‐37.   Ochi  M  K,  1998.  Ocean  Waves:  The  Stochastic  Approach,  Cambridge  University  Press.   Oil   Companies   International   Marine   Forum,   1994.   Prediction   of   Wind   and   Current   Loads   on   VLCCs,   2nd   edition,   Witherby  &  Co.,  London.   Paidoussis  M  P,  1970.  Dynamics  of  tubular  cantilevers  conveying  fluid.   J.  Mechanical  Engineering  Science,  12,  No  2,   85-­‐103.   Paidoussis  M  P  and  Deksnis  E  B,  1970.  Articulated  models  of  cantilevers  conveying  fluid:  The  study  of  a  paradox.   J.   Mechanical  Engineering  Science,  12,  No  4,  288-­‐300.   Paidoussis   M   P   and   Lathier   B   E,   1976.   Dynamics   of   Timoshenko   beams   conveying   fluid.   J.   Mechanical   Engineering   Science,  18,  No  4,  210-­‐220.   Palmer   A   C   and   Baldry  J   A   S,   1974.  Lateral   buckling   of   axially  constrained  pipes.  J.   Petroleum   Technology,   Nov   1974,   1283-­‐1284.   Pode  L,  1951.  Tables  for  Computing  the  Equilibrium  Configuration  of  a  Flexible  Cable  in  a  Uniform   Stream.  DTMB   Report.    687.   Principles   of   Naval   Architecture.   Revised   edition,   edited   by   J   P   Comstock,   1967.   Society   of   Naval   Architects   and   Marine  Engineers,  New  York.   Puech  A,  1984.  The  Use  of  Anchors  in  Offshore  Petroleum  Operations.  Editions  Technique.   Randolph   M   and   Quiggin   P,   2009.   Non-­‐linear   hysteretic   seabed   model   for   catenary   pipeline   contact.   OMAE   paper   79259,  2009  (www.orcina.com/Resources/Papers/OMAE2009-­‐79259.pdf).   Rawson   and   Tupper,   1984.   Basic   Ship   Theory   3rd   ed,   2:   Ship   Dynamics   and   Design,   482.   Longman   Scientific   &   Technical  (Harlow).   Rienecker  M  M  and  Fenton   J  D,  1981.  A  Fourier  approximation  method  for  steady  water  waves.   J.  Fluid  Mech.104,   119-­‐137.   Roark  R  J,  1965.  Formulas  for  Stress  and  Strain.  4th  edition  McGraw-­‐Hill.   Sarpkaya  T,  Shoaff  R  L,  1979.  Inviscid  Model  of  Two-­‐Dimensional  Vortex  Shedding  by  a  Circular  Cylinder.   Article   No.   79-­‐0281R,  AIAA  Journal,17,  no.  11,  1193-­‐1200.   Sarpkaya  T,  Shoaff  R  L,  1979.  A  discrete-­‐vortex  analysis  of  flow  about  stationary  and  transversely  oscillating  circular   cylinders.  Report  no.  NPS-­‐69SL79011,  Naval  Postgraduate  School,  Monterey,  California.   Rychlik  I,  1987.  A  new  definition  of  the  rainflow  cycle  counting  method.  Int.  J.  Fatigue  9,  No  2,  119-­‐121.   Skjelbreia  L,  Hendrickson  J,  1961.  Fifth  order  gravity  wave  theory.  Proc.  7th  Conf.  Coastal  Eng.  184-­‐196.   Sobey   R   J,   Goodwin   P,   Thieke   R   J   and   Westberg   R   J,   1987.   Wave   theories.   J.   Waterway,   Port,   Coastal   &   Ocean   Eng.   ASCE  113,  565-­‐587.   Sparks   C,   1980.   Le   comportement   mecanique   des   risers   influence   des   principaux   parametres.   Revue   de   l'Institut   Francais  du  Petrol,  35,  no.  5,  811.   Sparks  C,  1983.  Comportement  mecanique  des  tuyaux  influence  de  la  traction,  de  la  pression  et  du  poids  lineique  :   Application  aux  risers.  Revue  de  l'Institut  Francais  du  Petrol  38,    no.  4,  481.   Standing   RG,   Brendling   WJ,   Wilson   D,   1987.   Recent   Developments   in   the   Analysis   of   Wave   Drift   Forces,   Low-­‐ Frequency  Damping  and  Response.  OTC  paper  5456,  1987.   Tan   Z,   Quiggin   P,   Sheldrake   T,   2007.   Time   domain   simulation   of   the   3D   bending   hysteresis   behaviour   of   an   unbonded  flexible  riser.  OMAE  paper  29315,  2007  (www.orcina.com/Resources/Papers/OMAE2007-­‐29315.pdf).   Taylor   R   and  Valent  P,   1984.   Design   Guide   for  Drag   Embedment   Anchors,  Naval   Civil   Engineering   Laboratory   (USA),   TN  No  N-­‐1688.   Torsethaugen  K  and  Haver  S,  2004.  Simplified  double  peak  spectral  model  for  ocean  waves,  Paper  No.  2004-­‐JSC-­‐193,   ISOPE  2004  Touson,  France.  

18  

w

 

Introduction,  References  and  Links  

 

Thwaites,  1960.  Incompressible  Aerodynamics,  Oxford,  399-­‐401.   Timoshenko  S,1955.  Vibration  Problems  in  Engineering,  van  Nostrand.   Triantafyllou   M   S,   Yue   D  K  P   and   Tein   D  Y   S,   1994.   Damping   of  moored  floating   structures.  OTC  7489,   Houston,   215-­‐ 224.   Tucker  et  al,  1984.  Applied  Ocean  Research,  6,  No  2.   Tucker  M  J,  1991.  Waves  in  Ocean  Engineering.  Ellis  Horwood  Ltd.  (Chichester).   Wichers   J   E   W,   1979.   Slowly   oscillating   mooring   forces   in   single   point   mooring   systems.   BOSS79   (Second   International  Conference  on  Behaviour  of  Offshore  Structures).   Wichers  J  E  W,  1988.  A  Simulation  Model  for  a  Single  Point  Moored  Tanker.   Delft  University  Thesis.   Wu  M,  Saint-­‐Marcoux  J-­‐F,  Blevins  R  D,  Quiggin  P  P,  2008.  Paper  No.  ISOPE-­‐2008-­‐MWU10.  ISOPE  Conference  2008,   Vancouver,  Canada.  (www.orcina.com/Resources/Papers/ISOPE2008-­‐MWU-­‐10.pdf)   Young  A  D,  1989.  Boundary  Layers.  BSP  Professional  Books,  87-­‐91.   Suppliers  of  frequency  domain  VIV  software   SHEAR7  

SBM  Atlantia   1255  Enclave  Parkway,  Suite  1200   Houston,  TX  77077,  USA   Attention:  Dr.  S.  Leverette   Email:  [email protected]   Tel:  +1  281  899  4300   Fax:  +1  281  899  4307   VIVA  

JD  Marine   11777  Katy  Freeway,  Suite  434  South   Houston,  TX  77079,  USA   Phone:  +1  281  531  0888   Fax:  +1  281  531  5888   Email:  [email protected]  

19  

w

 

Tutorial,  Getting  Started  

 

2

TUTORIAL  

2.1

GETTING  STARTED  

This   short   tutorial   gives   you   a   very   quick   run   through   the   model   building   and   results   presentation   features   of   OrcaFlex.   On  completion  of  the  tutorial  we  suggest  that  you  also  look  through  the  pre-­‐run  examples  Ȃ  see  Example  Files.   On  starting  up  OrcaFlex,  you  are   presented  with  a  3D  view  showing  just  a  blue  line  representing  the  sea  surface  and   a  brown  line   representing  the  seabed.  At  the  top   of  the  screen  are  menus,  a  tool  bar  and  a  status  bar  arranged  in  the   manner   common   to   most   Windows   software.   As   usual   in   Windows   software,   nearly   all   actions   can   be   done   in   several   ways:   here,   to   avoid  confusion,   we  will   usually  only  refer   to   one   way  of   doing   the   action  we   want,   generally   using  the  mouse.  

  Figure:  

2.2

The  OrcaFlex  main  window  

BUILDING  A  SIMPLE  SYSTEM  

To  start  with,  we  will  build  a  simple  system  consisting  of  one  line  and  one  vessel  only.   Using   the   mouse,   click   on   the   new   vessel   button    on   the  toolbar.   The   cursor   changes  from   the   usual   pointer   to   a   crosshair  cursor  to  show  that  you  have  now  selected  a  new  object  and  OrcaFlex  is  waiting  for  you  to  decide  where  to   place   it.   Place   the   cursor   anywhere   on   the   screen   and   click   the   mouse   button.   A   "ship"   shape   appears   on   screen,   positioned  at  the  sea  surface,  and  the  cursor  reverts  to  the  pointer  shape.  To  select  the  vessel,  move  the  cursor  close   to   the   vessel   and   click   the   mouse   button   Ȃ   the   message   box   (near   the   top   of   the   3D   view)   will   confirm   when   the   vessel   has   been   selected.   Now   press   and   hold   down   the   mouse   button   and   move   the   mouse   around.   The   vessel   follows   the   mouse   horizontally,   but   remains   at   the   sea   surface.   (To   alter   vessel   vertical   position,   or   other   details,   select  the  vessel  with  the  mouse,  then  double  click  to  open  the  Vessel  data  window.)  

2.3

ADDING  A  LINE  

Now   add   a   line.   Using   the   mouse,   click   on   the   new   line   button   .   The   crosshair   cursor   reappears   Ȃ   move   the   mouse  to  a  point  just  to  the  right  of  the  vessel  and  click.  The  line  appears  as  a  catenary  loop  at  the  mouse  position.   Move  the  mouse  to  a  point  close  to  the  left  hand  end  of  the  line,  press  and  hold  down  the  mouse  button  and  move   the   mouse   around.  The  end  of   the   line  moves   around  following   t he   mouse,   and   the   line   is  redrawn   at   each  position.   Release   the   mouse   button,   move   to   the   right   hand   end,   click   and   drag.   This   time   the   right   hand   end   of   the   line   is   dragged  around.  In  this  way,  you  can  put  the  ends  of  the  lines  roughly  where  you  want  them.  (Final  positioning  to   exact   locations   has   to   be   done   by   typing   in   the   appropriate   numbers   Ȃ   select   the   line   with   the   mouse   and   double   click  to  bring  up  the  line  data  form.)   Move  the  line  ends  until  the   left  hand  end  of  the  line  is  close  to  the  bow  of  the  ship,  the  right  hand   end  lies  above  the   water  and  the  line  hangs  down  into  the  water.  

21  

w

 

Tutorial,  Adjusting  the  View    

At  this  point,  the  line  has  a  default  set  of  properties  and  both  ends  are  at  fixed  positions  relative  to  the  Global  origin.   For   the   moment   we   will   leave   the   line   properties   (length,  mass,   etc.)   at   their   default   values,   but  we   will   connect  the   left  hand  end  to  the  ship.  Do  this  as  follows:   1.

Click  on  the  line  near  the  left  hand  end,  to  select  that  end  of  the  line;  make  sure  you  have  selected  the  line,  not   the   vessel   or   the   sea.   The   message   box  at   the   left   hand   end   of  the   status   bar   tells   you   what   is   currently   selected.   If   you   have   selected   the  wrong   thing,   try  again.   (Note  that   you   don't   have  to   click   at   the  end  of   the   line   in  order   to  select  it  Ȃ  anywhere  in  the  left  hand  half  of  the  line  will  select  the  left  hand  end.  As  a  rule,  it  is  better  to  choose   a  point  well  away  from  any  other  object  when  selecting  something  with  the  mouse.)  

2.

Release  the  mouse  and  move  it  to  the  vessel,  hold  down  the   CTRL  key  and  click.  The  message  box  will  confirm   the   connection   and,   to  indicate   the   connection,   the   triangle   at   the   end   of  the   line   will   now   be   the   same   colour   as   the  vessel.  

Now  select  the  vessel  again  and  drag  it  around  with  the  mouse.  The  left  hand  end  of  the  line  now  moves  with  the   vessel.  Leave  the  vessel  positioned  roughly  as  before  with  the  line  in  a  slack  catenary.  

2.4

ADJUSTING  THE  VIEW  

The   default   view   of   the   system   is   an   elevation   of   the   global   X-­‐Z   plane   Ȃ   you   are   looking   horizontally   along   the   positive   Y   axis.   The   view   direction   (the   direction   you   are   looking)   is   shown   in   the   Window   Title   bar   in   azimuth/elevation   form   (azimuth=270;  elevation=0).   You   can   move   your   view  point   up,   down,   right   or   left,  and   you   can   zoom   in   or   out,   using   the   view  control   buttons     near   the   top   left   corner   of  the   window.   Click   on   each   of  the   top  3   buttons   in   turn:  then   click   again   with   the   SHIFT   key   held   down.   The   SHIFT   key  reverses   the   action   of   the   button.  If  you  want  to  move  the  view  centre  without  rotating,  use  the  scroll  bars  at  the  bottom  and  right  edges  of  the   window.  By  judicious  use  of  the  buttons  and  scroll  bars  you  should  be  able  to  find  any  view  you  like.   Alternatively,   you   can   alter   the   view   with   the   mouse.   Hold   down   the   ALT   key   and   left   mouse   button   and   drag.   A   rectangle   on   screen   shows   the   area   which   will   be   zoomed   to   fill   the   window   when   the   mouse   button   is   released.   SHIFT+ALT+left  mouse  button  zooms  out  Ȃ  the  existing  view  shrinks  to  fit  in  the  rectangle.   Warning:  

OrcaFlex   will   allow   you   to   look   up   at   the   model   from   underneath,   effectively   from   under   the   seabed!  Because  the   view   is  isometric  and  all   lines  are  visible,   it  is   not  always  apparent  that  this   has  occurred.  When  this  has  happened,  the  elevation  angle  is  shown  as  negative  in  the  title  bar.  

There  are  three  shortcut  keys  which  are  particularly  useful  for  controlling  the  view.  For  example  CTRL+P  gives  a  plan   view  from  above;   CTRL+E  gives  an  elevation;   CTRL+Q  rotates  the  view  through  90°  about  the  vertical  axis.  ( CTRL+P   and  CTRL+E  leave  the  view  azimuth  unchanged.)   Now   click   the     button   on   the   3D   View  to   bring   up   the  Edit   View  Parameters   form.   This   gives   a   more   precise   way   of  controlling  the  view  and  is  particularly  useful  if  you  want  to  arrange  exactly  the  same  view  of  2  different  models  Ȃ   say  2  alternative  configurations  for  a  particular  riser  system.  Edit  the  view  parameters  if  you  wish  by  positioning  the   cursor  in  the  appropriate  box  and  editing  as  required.   If   you   should   accidentally   lose   the   model   completely   from   view   (perhaps   by   zooming   in   too   close,   or   moving   the   view  centre  too  far)  there  are  a  number  of  ways  of  retrieving  it:   x

Press  CTRL+T  or  right  click  in  the  view  window  and  select  Reset  to  Default  View.  

x

Press  the  Reset  button  on  the  Edit  View  Parameters  form.  This  also  resets  back  to  the  default  view.  

x

Zoom  out  repeatedly  until  the  model  reappears.  

x

Close   the   3D   View   and   add  a   new  one   (use   the  Window|Add   3D   View  menu  item).   The   new  window   will   have   the  default  view  centre  and  view  size.  

2.5

STATIC  ANALYSIS   Note:  

If  you  are  running  the  demonstration  version  of  OrcaFlex  then  this  facility  is  not  available.  

To  run  a  static  analysis  of  the  system,  click  on  the  calculate  statics  button   .  The  message  box  reports  which  line  is   being   analysed   and   how   many   iterations   have   occurred.   When   the   analysis   is   finished   (almost   instantly   for   this   simple  system)  the  Program  State  message  in  the  centre  of  the  Status  Bar  changes  to  read  "Statics  Complete",  and   the  Static  Analysis  button  changes  to  light  grey  to  indicate  that  this  command  is  no  longer  available.  The  appearance   of   the   line   will   have   changed   a   little.   When   editing   the   model,   OrcaFlex   uses   a   quick   approximation   to   a   catenary  

22  

w

 

Tutorial,  Dynamic  Analysis  

 

shape  for  general  guidance   only,  and  this  shape  is   replaced  with  the  true  catenary   shape  when  static  analysis  has   been  carried  out.  (See  Static  Analysis  for  more  details).   We  can  now   examine  the  results  of  the  static  analysis  by  clicking  on  the  Results   button   Selection  window.  

.  This  opens  a  Results  

You  are  offered  the  following  choices:   x

Results   in   numerical   and   graphical   form,   with  various   further   choices   which  determine   what   the   table   or   graph   will  contain.  

x

Results  for  all  objects  or  one  selected  object.  

Ignore  the  graph  options  for  the  moment,  select  Summary  Results  and  All  Objects,  then  click  Table.  A  summary  of   the   static   analysis   results   is   then   displayed   in   spreadsheet   form.   Results   for   different   objects   are   presented   in   different  sheets.  To  view  more  static  analysis  results  repeat  this  process:  click  on  the  Results  button  and  select  as   before.  

2.6

DYNAMIC  ANALYSIS  

We   are  now  ready  to   run   the  simulation.   If   you   are  running   the   demonstration   version   of   OrcaFlex   then   you   cannot   do  this,  but  instead  you  can  load  up  the  results  of  a  pre-­‐run  simulation  Ȃ  see  Examples.   Click   the   Run   Dynamic   Simulation   button   .   As   the   simulation   progresses,   the   status   bar   reports   current   simulation   time   and   expected   (real)   time   to  finish   the   analysis,   and   the   3D   view  shows   the  motions   of  the   system   as   the  wave  passes  through.   Click  the  Start  Replay  button   .  An  animated  replay  of  the  simulation  is  shown  in  the  3D  view  window.  Use  the   view   control   keys   and   mouse   as   before   to   change   the   view.   The   default   Replay   Period   is   Whole   Simulation.   This   means  that  you  see  the  simulation  start  from  still  water,  the  wave  building  and  with  it  the  motions  of  the  system.   Simulation  time  is  shown  in  the  Status  bar,  top  left.  Negative  time  means  the  wave  is  still  building  up  from  still  water   to  full  amplitude.  At  the  end  of  the  simulation  the  replay  begins  again.   The  replay  consists  of  a  series  of  "frames"  at   equal  intervals  of  time.   Just  as  you  can   "zoom"  in  and  out  in  space   for  a   closer  view,  so  OrcaFlex  lets  you  "zoom"  in  and  out  in  time.  Click  on  the  Replay  Parameters  button   ,  edit  Interval   to  0.5s  and  click  OK.  The  animated  replay  is  now  much  jerkier  than  before  because  fewer  frames  are  being  shown.   Now  click  again  on  Replay  Parameters,  set  Replay  Period  to  Latest  Wave  and  click  on  the  Continuous  box  to  deselect.   The  replay  period  shown  is  at  the  end  of  the  simulation  and  has  duration  of  a  single  wave  period.  At  the  end  of  the   wave  period  the  replay  pauses,  then  begins  again.   Now  click  on  the  Replay  Step  button    to  pause  the  replay.  Clicking  repeatedly  on  this  button  steps  through  the   replay  one  frame  at  a  time  Ȃ  a  very  useful  facility  for  examining  a  particular  part  of  the  motion  in  detail.  Click  with   the  SHIFT  key  held  down  to  step  backwards.   You  can  then  restart  the  animation  by  clicking  on  'Start  Replay'  as  before.  To  slow  down  or  speed  up  the  replay,  click   on  Replay  Parameters  and  adjust  the  speed.  Alternatively  use  the  shortcuts   CTRL+F  and   SHIFT+CTRL+F  to  make  the   replay  faster  or  slower  respectively.   To  exit  from  replay  mode  click  on  the  Stop  Replay  button  

2.7

.  

MULTIPLE  VIEWS  

You  can  add  another  view  of  the  system  if  you  wish  by  clicking  on  the   View  button   .  Click  again  to  add  a  third   view,  etc.  Each  view  can  be  manipulated  independently  to  give,  say,  simultaneous  plan  and  elevation  views.  To  make   all   views   replay   together,   click   on  Replay   Control   and   check   the   All   Views   box.   To   remove   an   unwanted   view  simply   close   its   view   window.   To   rearrange   the   screen   and   make   best   use   of   the   space,   click   Window   and   choose   Tile   Vertical  (F4)  or  Tile  Horizontal  (SHIFT+F4).   Alternatively,  you   can   minimise  windows   so  that   they  appear   as   small   icons   on   the   background,   or   you   can   re-­‐size   them   or   move   them   around   manually   with   the   mouse.   These   are   standard   Windows   operations   which   may   be   useful   if   you   want   to   tidy   up   the   screen   without   having   to   close   a   window  down  completely.  

23  

w

 

Tutorial,  Looking  at  Results    

2.8

LOOKING  AT  RESULTS  

Now  click  on  the  Results  button  

.  This  opens  a  Results  Selection  window.  

You  are  offered  the  following  choices:   x

Results  as  Tables  or  Graphs,  with  various  further  choices  which  determine  what  the  table  or  graph  will  contain.  

x

Results  for  all  objects  or  one  selected  object.  

Select   Time   History   for   any   line,   then   select   Effective   Tension   at   End   A   and   click   the   Graph   button.   The   graph   appears  in  a  new  window.  You  can  call  up  time  histories  of  a  wide  range  of  parameters  for  most  objects.  For  lines,   you   can   also   call   up   Range   Graphs   of   effective   tension,   curvature,   bend   moment   and   many   other   variables.   These   show   maximum,   mean   and   minimum   values   of   the   variable   plotted   against   position   along   the   line.   Detailed   numerical  results  are  available  by  selecting  Summary  Results,  Full  Results,  Statistics  and  Linked  Statistics.   Time  history  and  range  graph  results  are  also  available  in  numerical  form  Ȃ  select  the  variable  you  want  and  press   the   Values   button.   The   results   can   be   exported   as   Excel   compatible  spreadsheets   for   further   processing   as   required.   Further   numerical   results   are   available   in   tabular   form   by   selecting   Summary   Results,   Full   Results,   Statistics   and   Linked  Statistics.   Results  Post-­‐Processing  

Extra  post-­‐processing  facilities  are  available  through  Excel  spreadsheets.  

2.9

GETTING  OUTPUT  

You   can   get   printed   copies   of   data,   results   tables,   system   views   and   results   graphs   by   means   of   the   File  |  Print   menu,   or   by   clicking   Print   on   the   pop-­‐up   menu.   Output   can   also   be   transferred   into   a   word   processor   or   other   application,  either  using  copy+paste  via  the  clipboard  or  else  export/import  via  a  file.   Note:  

2.10

Printing  and  export  facilities  are  not  available  in  the  demonstration  version  of  OrcaFlex.  

INPUT  DATA  

Take  a  look  through  the  input  data  forms.  Start  by  resetting  the  program:  click  on  the  Reset  button   .  This  returns   OrcaFlex   to   the   reset   state,   in   which   you   can   edit   the   data   freely.   (While   a   simulation   is   active   you   can   only   edit   certain  non-­‐critical  items,  such  as  the  colours  used  for  drawing.)   Now  click  on  the  Model  Browser  button  

.  This  displays  the  data  structure  in  tree  form  in  the  Model  Browser.  

Select   an   item   and   double   click   with   the   mouse   to   bring   up   the   data   form.   Many   of   the   data   items   are   self   explanatory.  For  details  of  a  data  item,  select  the  item   with  the  mouse  and  press  the   F1  key.  Alternatively  use  the   question  mark  Help  icon  in  the  top  right  corner  of  the  form.  Have  a  look  around  all  the  object  data  forms  available  to   get  an  idea  of  the  capabilities  of  OrcaFlex.   End  of  Tutorial   We   hope   you   have   found   this   tutorial   useful.   To   familiarise   yourself   with  OrcaFlex,   try   building   and   running   models   of  a  number  of  different  systems.  The  manual  also  includes  a  range  of   examples  which  expand  on  particular  points  of   interest  or  difficulty.   Finally,  please  remember  that  we  at  Orcina  are  on  call  to  handle  your  questions  if  you  are  stuck.  

24  

w

 

User  Interface,  Introduction  

 

3

USER  INTERFACE  

3.1

INTRODUCTION  

3.1.1

Program  Windows  

OrcaFlex   is   based   upon   a   main   window  that   contains   the  Menus,   a   Tool   Bar,   a   Status  Bar   and   usually   at  least   one  3D   view.  The  window  caption  shows  the  program  version  and  the  file  name  for  the  current  model.  

  Figure:  

The  OrcaFlex  main  window  

Within  this  main  window,  any  number  of  child  windows  can  be  placed  which  may  be:   3D  View  Windows  

showing  3D  pictorial  views  of  the  model  

Graph  Windows  

showing  results  in  graphical  form  

Spreadsheet  Windows  

showing  results  in  numerical  form  

Text  Windows  

reporting  status  

Additional  temporary  windows  are   popped  up,  such  as  Data  Forms  for   each  object  in   the  model  (allowing   data  to  be   viewed  and  modified)  and  dialogue  windows  (used  to  specify  details  for  program  actions  such  as  loading  and  saving   files).  While  one  of  these  temporary  windows  is  present  you  can  only  work  inside  that  window  Ȃ  you  must  dismiss   the  temporary  window  before  you  can  use  other  windows,  the  menus  or  toolbar.   The  actions  that  you  can  perform  at  any  time  depend  on  the  current  Model  State.   Arranging  Windows   3D   View,   Graph,   Spreadsheet   and   Text   Windows   may   be   tiled   so   that   they   sit   side-­‐by-­‐side,   but   they   must   remain   within  the  bounds  of  the  main  window.  The  program  rearranges  the  windows  every  time  a  new  window  is  created.  

3.1.2

The  Model  

OrcaFlex   works   by   building   a   mathematical   computer   model   of   your   system.   This   model   consists   of   a   number   of   objects  that  represent  the  parts  of  the  system  Ȃ  e.g.  vessels,  buoys,  lines  etc.   Each  object  has  a  name,   which  can  be  any  length.  Object  names  are   not  case-­‐sensitive,  so   Riser,  riser  and   RISER   would  all  refer  to  the  same  object.  This  behaviour  is  the  same  as  for  Windows  file  names.   The  model  always  has  two  standard  objects:   x

General  contains  general  data,  such  as  title,  units  etc.  

x

Environment  represents  the  sea,  seabed,  waves,  current  etc.  

You  can  then  use  the  Model  Browser  or  the  toolbar  to  add  other  objects  to  represent  the  parts  of  your  system.  There   is  no  limit,  other  than  the  capacity  of  your  computer,  to  the  number  of  objects  you  can  add  to  the  model.  At  any  time,   you  can  save  your  model  to  a  data  file.  

25  

w

 

User  Interface,  Introduction    

3.1.3

Model  States  

OrcaFlex  builds  and  analyses  a  mathematical  model  of  the  system  being  analysed,  the  model  being  built  up  from  a   series  of  interconnected  objects,  such  as  Lines,  Vessels  and  Buoys.  For  more  details  see  Modelling  and  Analysis.   OrcaFlex  works  on  the  model  by  moving  through  a  sequence  of  states,  the  current  state  being  shown  on  the  status   bar.  The  following  diagram  shows  the  sequence  of  states  used  and  the  actions,  results  etc.  available  in  each  state.  

RESET Calculate Static Position Reset

Calculating Statics

Edit or Reset

STATICS COMPLETE Reset Run Pause SIMULATION UNSTABLE

Simulating

SIMULATION Reset PAUSED

Run Extend Simulation SIMULATION COMPLETE

Reset  

Figure:  

Model  States  

The  states  used  are  as  follows:   Reset  

The  state  in  which  OrcaFlex  starts.  In  Reset  state  you  can  freely  change  the  model  and  edit  the  data.  No  results  are   available.   Calculating  Statics  

OrcaFlex  is  calculating  the  statics  position  of  the  model.  You  can  abort  the  calculation  by   CLICKING  the  Reset  button.   Statics  Complete  

The   statics   calculation   is  complete   and   the   static   position  results   are   available.   You   are   allowed   to   make   changes   to   the  model  when  in  this  state  but  if  you  make  any  changes  (except  for  very  minor  changes  like  colours  used)  then  the   model  will  be  automatically  reset  and  the  statics  results  will  be  lost.   Simulating  

The  dynamic  simulation  is  running.  The  results  of  the  simulation  so  far  are  available  and  you  can  examine  the  model   data,  but  only  make  minor  changes  (e.g.  colours  used).  You  cannot  store  the  simulation  to  a  file  while  simulating  Ȃ   you  must  pause  the  simulation  first.  

26  

w

 

User  Interface,  Introduction  

 

Simulation  Paused  

There  is  a  simulation  active,  but  it  is  paused.  The  results  so  far  are  available  and  you  can  examine  the  model  data.   You  can  also  store  the  part-­‐run  simulation  to  a  file.   Simulation  Complete  

The  simulation  is  complete.   The  simulation  results  are  available  and  you  can  store  the  results  to  a  simulation  f ile  for   later   examination.   You   must   reset   the   model,   by   CLICKING   on   the   Reset   button,   before   significant   changes   to   the   model  can  be  made.   You  can  use  the  Extend  Dynamic  Simulation  facility  if  you  wish  to  simulate  for  a  further  period  of  time.   Simulation  Unstable  

The   simulation   has   become   unstable.   The   simulation   results   are   available   and   you   can   store   the   results   to   a   simulation   file   for   later   examination.   This   allows   you   to   try   and   understand   why   the   simulation   has   become   unstable.   You   may   also   want   to   examine   the   results   up   until   the   point   at   which   the   simulation   became   unstable.   However,   please   treat   these   results   with   caution   Ȃ   because   the   simulation   eventually   went   unstable   this   indicates   that  the  dynamic  simulation  may  not  have  converged  at  earlier  simulation  times.   You  must  reset  the  model,  by  CLICKING  on  the  Reset  button,  before  significant  changes  to  the  model  can  be  made.   Typical  model  state  flow  

To  illustrate  how  model  states  work,  here  is  an  example  of  a  typical  working  pattern:   1.

In  Reset  state,  open  a  new  model  from  a  data  file  or  use  the  current  model  as  the  starting  point  for  a  new  model.  

2.

In  Reset  state,  add  or  remove  objects  and  edit  the  model  data  as  required  for  the  new  model.  It  is  generally  best   to  use  a  very  simple  model  in  the  early  stages  of  design  and  only  add  more  features  when  the  simple  model  is   satisfactory.  

3.

Run   a   static   analysis   (to   get   to   Statics   Complete   state)   and   examine   the   static   position   results.   Make   any   corrections   to   the   model   that   are   needed   Ȃ   this   will   automatically   reset   the   model.   Steps   (2)   and   (3)   are   repeated  as  required.  

4.

Run  a  simulation  and  monitor  the  results  during  the  simulation  (in  Simulating  state).  

5.

If  further  changes  to  the  model  are  needed  then   Reset  the  model  and  edit  the  model  accordingly.  Steps  (2)  to   (5)  are  repeated  as  required.  

6.

Finalise   the   model,   perhaps   improving   the   discretisation   (for   example   by   reducing   the   time   step   sizes   or   increasing   the   number   of   segments   used   for   Lines).   Run   a   final   complete   simulation   (to   reach   Simulation  Complete  state)  and  generate  reports  using  the  results.  

3.1.4

Toolbar  

The   toolbar   holds   a   variety   of   buttons   that   provide   quick   access   to   the   most   frequently   used   menu   items.   The   selection  of  buttons  available  varies  with  the  current  Program  State.   Button   Action                    

Equivalent  Menu  Item  

Open  

File  |  Open  

Save  

File  |  Save  

Model  Browser  

Model  |  Model  Browser  

New  Vessel  

Model  |  New  Vessel  

New  Line  

Model  |  New  Line  

New  6D  Buoy  

Model  |  New  6D  Buoy  

New  3D  Buoy  

Model  |  New  3D  Buoy  

New  Winch  

Model  |  New  Winch  

New  Link  

Model  |  New  Link  

27  

w

 

User  Interface,  Introduction    

Button   Action                          

3.1.5

Equivalent  Menu  Item  

New  Shape  

Model  |  New  Shape  

Calculate  Statics  

Calculation  |  Single  Statics  

Run  Simulation  

Calculation  |  Run  Dynamic  Simulation  

Pause  Simulation  

Calculation  |  Pause  Dynamic  Simulation  

Reset  

Calculation  |  Reset  

Start  Replay  

Replay  |  Start  Replay  

Stop  Replay  

Replay  |  Stop  Replay  

Step  Replay  Forwards  

Replay  |  Step  Replay  Forwards  

Edit  Replay  Parameters  

Replay  |  Edit  Replay  Parameters  

Add  New  3D  View  

Window  |  Add  3D  View  

Examine  Results  

Results  |  Select  Results  

Help  Contents  and  Index  

Help  |  OrcaFlex  Help  

Status  Bar  

The  Status  Bar  is  divided  into  three  fields:   The  Message  Box  

This  is  at  the  left  hand  end.  It  shows  information  about  the  progress  of  the  current  action,  such  as  the  name  of  the   currently  selected  object,  or  the  current  iteration  number  or  simulation  time.  Error  messages  are  also  shown  here.   When   a   statics   calculation   is  done   messages   showing   the   progress   of   the   calculation  are   shown  in   the  message   box.   To   see   all   the   messages   from   the   statics   calculation   CLICK   on   the   message   box   Ȃ   the   Statics  Progress  Window   will   then  be  opened.   The  Program  State  Indicator  

In  the  centre  and  shows  which  state  the  program  is  in  (see  Model  States).   The  Information  Box  

This  is  on  the  right.  It  shows  additional  information,  including:   x

The  global  coordinates  of  the  position  of  the  cursor,  in  the  current  view  plane.  

x

Distances  when  using  the  measuring  tape  tool.  

3.1.6

Mouse  and  Keyboard  Actions  

As  well  as  the  standard  Windows  mouse  operations  such  as  selection  and  dragging  OrcaFlex  uses  some  specialised   actions.   Clicking   the   right   mouse   button   over   a   3D   View,   Graph   or   Text   Window   displays   a   pop-­‐up   menu   of   frequently  used  actions,  such  as  Copy,  Paste,  Export  etc.  For  wire  frame  3D  Views  and  Graph  Windows  the  mouse   can   be   used   for   zooming.   Simply   hold   the   ALT   key   down   and   using   the   left   mouse   button,   drag   a   box   over   the   region   you  want  to  view.   All  of  the  menu  items  can  be  selected  from  the  keyboard  by  pressing  ALT  followed  by  the  underlined  letters.   Example:  

To  exit  from  the  program  (menu:  File  |  Exit)  press  ALT+F  then  X,  or  ALT  then  F  then  X  

A  number  of  frequently  used  menu  items  may  also  be  accessed  by  shortcut  keys,  such  as   CTRL+R  to  start  a  replay.   See  the  tables  below.  The  shortcut  keys  are  also  displayed  on  the  OrcaFlex   menus.  We  suggest  that  as  you  become   more  familiar  with  the  operation  of  OrcaFlex  that  you   memorise  some  of  the  shortcut  keys  for  actions  that  you  use   frequently.  

28  

w

 

User  Interface,  Introduction  

 

Keys  on  Main  Window  

New  model  

CTRL+N  

Open  file  

CTRL+O  

Save  file  

CTRL+S  

Open  data  

SHIFT+CTRL+O  

Save  data  

SHIFT+CTRL+S  

Help  

F1  

Print  

F7  

Show  /  hide  Model  Browser  

F6  

Switch  between  Model  Browser  and  Main  Window  

SHIFT+F6  

Calculate  static  position  

F9  

Run  dynamic  simulation  

F10  

Pause  dynamic  simulation  

F11  

Reset  

F12  

Open  results  selection  form  

F5  

Go  to  next  window  

CTRL+F6  

Go  to  previous  window  

SHIFT+CTRL+F6  

Tile  windows  vertically  

F4  

Tile  windows  horizontally  

SHIFT+F4  

Close  selected  window  

CTRL+F4  

Close  program  

ALT+F4  

Keys  on  Model  Browser  

Edit  data  

Enter  

Rename  object  

F2  

Switch  to  Main  Window  

SHIFT+F6  

Locate  

F3  

Move  selected  objects  

CTRL+M  

Hide  

CTRL+H  

Show  

CTRL+S  

Hide  all  objects  

SHIFT+CTRL+H  

Show  all  objects  

SHIFT+CTRL+S  

View  by  Groups  

SHIFT+CTRL+G  

View  by  Types  

SHIFT+CTRL+T  

Lock  /  Unlock  objects  

CTRL+L  

Cut  

CTRL+X  

Copy  

CTRL+C  

Paste  

CTRL+V  

Delete  

DELETE  

Close  browser  

ESC  

Keys  on  Data  Forms  

Help  

F1  

Go  to  next  data  form  

F6  

Go  to  previous  data  form  

SHIFT+F6  

29  

w

 

User  Interface,  Introduction    

Display  batch  script  names  for  currently  selected   data  item  or  table.  

F7  

Display  Properties  Report  

ALT+ENTER  

Show  connections  report  

F8  

Copy  form  

F9  

Export  form  

F10  

Print  form  

CTRL+P  

Open  calculator  

F12  

Data  Selection  Keys  

Go  to  next  data  item  or  table  

TAB  

Go  to  previous  data  item  or  table  

SHIFT+TAB  

Go  to  data  item  or  table  labelled  with  underlined  letter  

ALT+LETTER  

Move  around  within  a  table  

ĸĺĹĻ  

Select  multiple  cells  in  table  

SHIFT  +  ĸĺĹĻ   SHIFT+HOME   SHIFT+END  

Go  to  first  or  last  column  in  table  

HOME  END  

Go  up  or  down  table  several  rows  at  a  time  

PGUP  PGDN  

Data  Editing  Keys  

Enter  new  value  for  selected  cell   Edit  current  value  of  selected  cell  

Type  new  value   F2  

Open  drop-­‐down  list  

ALT  +  ĹĻ  

Move  around  within  new  data  value  being  entered  

ĸĺHOME  END  

Accept  edit  

RETURN  

Accept  edit  and  go  to  adjacent  cell  in  table  

ĹĻ  

Cancel  edit  

ESC  

Cut  selected  cell(s)  to  clipboard  

CTRL+X  

Copy  selected  cell(s)  to  clipboard  

CTRL+C  

Paste  from  clipboard  

CTRL+V  

Fill  selection  from  top  (copy  top  cell  down)  

CTRL+D  

Fill  selection  from  left  (copy  leftmost  cell  to  right)  

CTRL+R  

Fill  selection  from  bottom  (copy  bottom  cell  up)  

CTRL+U   SHIFT+CTRL+D  

Fill  selection  from  right  (copy  rightmost  cell  to  left)  

CTRL+L   SHIFT+CTRL+R  

Insert  new  rows  in  table  

INSERT  

Delete  selected  rows  from  table  

DELETE  

3D  View  Control  Keys  

Elevation  view  

CTRL+E  

Plan  view  

CTRL+P  

Rotate  viewpoint  up  (increment  view  elevation  angle)  

CTRL+ALT+Ĺ  

Rotate  viewpoint  down  (decrement  view  elevation  angle)  

CTRL+ALT+Ļ  

Rotate  viewpoint  right  (increment  view  azimuth  angle)  

CTRL+ALT+ĺ  

Rotate  viewpoint  left  (decrement  view  azimuth  angle)  

CTRL+ALT+ĸ  

30  

w

 

User  Interface,  OrcaFlex  Model  Files  

 

Rotate  viewpoint  +90°  

CTRL+Q  

Rotate  viewpoint  -­‐90°  

SHIFT+CTRL+Q  

Zoom  In  

CTRL+I  

Zoom  Out  

SHIFT+CTRL+I  

Move  view  centre  Ȃ  fine  adjustment  

ĸĺĹĻ  

Move  view  centre  Ȃ  coarse  adjustment  

CTRL  +  ĸĺĹĻ  

Edit  view  parameters  for  current  3D  view  

CTRL+W  

Reset  to  default  view  

CTRL+T  

Set  as  default  view  

SHIFT+CTRL+T  

Show  entire  model  

CTRL+ALT+T  

3D  View  Control  Keys  (for  wire  frame  graphics  only)  

Show  /  Hide  local  axes  

CTRL+Y  

Show  /  Hide  node  axes  

CTRL+ALT+Y  

Undo  most  recent  drag  

CTRL+Z  

Lock/Unlock  selected  object  

CTRL+L  

Place  new  object  

SPACE  or  RETURN  

Edit  selected  object  

CTRL+F2  

Cut  selected  object  to  clipboard  

CTRL+X  

Copy  selected  object,  or  view  if  none  selected,   to  clipboard  

CTRL+C  

Paste  object  from  clipboard  (followed  by  mouse  click   or  RETURN  to  position  the  new  object)  

CTRL+V  

Delete  selected  object  

DELETE  

Measuring  tape  tool  

SHIFT+CTRL+drag  

Replay  Control  Keys  

Start  /  Stop  replay  

CTRL+R  

Replay  faster  

CTRL+F  

Replay  slower  

SHIFT+CTRL+F  

Step  forwards  one  frame  in  the  replay  and  pause  

CTRL+A  

Step  backwards  one  frame  in  the  replay  and  pause  

CTRL+B  

Edit  replay  parameters  

CTRL+D  

3.2

ORCAFLEX  MODEL  FILES  

3.2.1

Data  Files  

OrcaFlex  models  are  saved  to  either  binary  data  files  (.dat)  or  text  data  files  (.yml).   All  versions  of  OrcaFlex  can  read  binary  data  files.  Text  data  files  were  only  introduced  in  version  9.3a  and  so  cannot   be  read  by  older  versions  of  the  program.   Binary  data   files   have   strong   version   compatibility   features.   For   example,   when   OrcaFlex   attempts   to   open   a   binary   data   file   written   by   a   later   version   of   the   program   it   is   able   to   report   informative   compatibility   warnings.   The   program   is   not   able   to   be   as   helpful   and   informative   when   working   with   text   data   files   across   program   versions.   Whilst   we   strive   to  achieve   as   much   compatibility   as   possible   for   text   data   files   across   program   versions,   we   cannot   achieve  the  same  level  of  compatibility  as  that  for  binary  data  files.   Text   data   files,   as   written   by   OrcaFlex,   contain   only   data   that   is   active   in   the   model.   For   example,   if   implicit   time   integration  is  selected  in  the  model  then  all  data  relating  to  explicit  time  integration  is  excluded  from  the  text  data   file.   On   the   other   hand,   binary  data   files   contain   all   data   whether   or   not   it   is  active.   The   fact   that   the   binary  data   file  

31  

w

 

User  Interface,  OrcaFlex  Model  Files    

contains   inactive   data   can   be   very   useful   and   so,   in   general,   we   would   recommend   that   model   building   and   development  is  performed  using  the  binary  data  file.   Text  data  files  can  be  created  without  the  use  of  OrcaFlex  simply  by  entering  text  into  a  text  editor.  In  general  we   would   not   advocate   this   approach   to   model   building.   For   very   simple   systems   it   may   be   a   practical   approach   but   more   complex   models   are   usually   much   easier   to   build   and   inspect   using   the   full   capabilities   and   visualisation   strengths  of  OrcaFlex.  On  the  other  hand,  text  data  files  can  be  very  effective  when  making  minor  changes  to  existing   models.   Using   text   data   files   for   such   minor   variations   of   existing   models   makes   it   much   easier   to   monitor   just   what   has  been  changed,  for  example  by  using  standard  text  differencing  programs.   Text   data   files   are   highly   readable   and   self-­‐documenting   which   makes   them   ideal   for   QA   and   archival   purposes.   Another  application  well  suited  to  the  use  of  text  data  files  is  automation.  

3.2.2

Text  Data  Files  

Text   data   files   are   used   to   define   and   represent   OrcaFlex   models   in   a   human   readable   and   easily   editable   format.   Text  data  files  can  be  opened  and  saved  by  OrcaFlex.  A  very  simple  example  is  shown  below:   General:      StageDuration:          -­  10.0          -­  50.0   Lines:      -­  Name:  Line1          Length,  TargetSegmentLength:              -­  [60.0,  5.0]              -­  [40.0,  2.0]              -­  [120.0,  10.0]  

This   example  first   defines  a  10s   build-­‐up  stage  followed  by  stage  1   with  50s   duration.   Then  a  Line  is  created   and   named   "Line1".   Finally   the   section   data   is   specified:   three   sections   are   created   with   varying   section   lengths   and   segment  lengths.  Default  values  are  used  for  all  data  which  are  not  specified.   Note:  

The   formatting   (colour,   bold,   italic   etc.)   in   the   examples   here   has   been   added   to   aid   readability,   and  is  not  a  feature  or  requirement  of  text  data  files  themselves.  

YAML  file  format   Text  data  files  use  a  standard  file  format  called  YAML  and  should  be  saved  with  the  .yml  file  extension.  The  YAML  file   format  was  chosen  because  it  is  extremely  easy  to  read  and  write.   YAML  files  are  plain  text  files  and  so  can  be  edited  in  any  text  editor.  We  have  found  Notepad++  to  be  a  very  effective   editor  for  YAML  files.  Notepad++  has  a  tabbed  interface  for  easy  editing  of  multiple  files  and  has  code  folding  and   syntax  highlighting  facilities  that  work  well  with  YAML  files.   Note:  

YAML  files  must  be  saved  with  the  UTF-­‐8  character  encoding.  

More  details  on  the  YAML  format  and  Notepad++  can  be  obtained  from  the  following  web  sites:   x

http://en.wikipedia.org/wiki/YAML  Ȃ  YAML  page  on  Wikipedia.  

x

http://www.yaml.org/  Ȃ  Official  YAML  homepage.  

x

http://www.yaml.org/spec/  Ȃ  Complete  technical  specification  of  YAML.  

x

http://notepad-­‐plus.sourceforge.net/  Ȃ  Notepad++.  

Elements  of  a  text  data  file   The  most  basic  element  of  a  text  data  file  is  the  name/value  pair:   UnitsSystem:  SI   The  name  (UnitsSystem)  is  written  first,  followed  by  a   colon  (:),  then  a   SPACE,  and   then  the  value  (SI).  The  names   used  in  text  data  files  are  the  same  as  used  to  identify  data  items  in   batch  script  files.   Names  and  values  in  YAML  files  can  contain  spaces  and  other  punctuation:   General:      StaticsMethod:  Whole  System  statics   Lines:  

32  

w

 

User  Interface,  OrcaFlex  Model  Files  

 

   -­  Name:  12"  Riser      -­  Name:  Umbilical,  upper      -­  Name:  £"!$%^&*(){}[]=+-­_#~'@:;;/?.>,<\|  

This   example  also  contains  a   list.  New  items  in  a  list  are   introduced  by  a  dash  ( -­‐)  followed  by  a   SPACE.  Items  in  a  list   can  span  more  than  a  single  line:   Lines:      -­  Name:  Riser          TopEnd:  End  B          ContentsDensity:  0.8      -­  Name:  Umbilical          TopEnd:  End  A          ContentsDensity:  0.0   Outline   indentation   is   used   to   delimit   blocks   in   a   YAML   file.   This   concept,   known   as   significant   indentation,   is   perhaps  a  little  unusual  as  most  data  formats  and  programming  languages  use  symbols  to  indicate  the   beginnings   and  ends  of  blocks.  To  understand  this  better  consider  the  following  example:   General:      UnitsSystem:  SI      StaticsMethod:  Whole  System  statics   Environment:      WaterDepth:  80   The   two   lines   immediately   following  General:   which   are   indented   by   two  spaces,   form   a   single   block.   This   block   is   ended   by   Environment:   because   it   is   not   indented.   A   second   block   follows   Environment:   containing   a   single   name/value  pair  which  defines  the  water  depth.   Indentation  must  be  made  with   SPACE  characters  rather  than   TAB  characters.  It  does  not  matter  how  many  spaces   are   used   so   long   as   the   indentation   is   consistent   within   each   block.   However,   it   is   good   practice   to   use   the   same   indentation  throughout  a  file.  OrcaFlex  itself  uses  indentation  of  two  spaces  when  it  writes  YAML  files.   Lists  are  commonly  used  to  represent  tables  of  data:   Lines:      -­  Name:  Line1          LineType,  Length,  TargetSegmentLength:              -­  [Line  Type1,  60,  5]              -­  [Line  Type1,  40,  2]              -­  [Line  Type2,  120,  10]  

The  name  LineType,  Length,  TargetSegmentLength  indicates  three  columns  of  data,   LineType,  Length   and  TargetSegmentLength  which  are   interpreted   in  that   order.   The   comma   (,)   character   is  used   as   a   separator.   Note   that   you   do   not   have   to   present   the   data   in   the   same   o rder   as   it   appears   in   OrcaFlex.   The   following   example   is   equivalent  to  the  previous  example:   Lines:      -­  Name:  Line1          Length,  TargetSegmentLength,  LineType:              -­  [60,  5,  Line  Type1]              -­  [40,  2,  Line  Type1]              -­  [120,  10,  Line  Type2]  

You  can,  if  you  wish,  omit  columns,  in  which  case  default  values  will  be  used:   Lines:      -­  Name:  Line1          LineType,  Length:              -­  [Line  Type1,  60]              -­  [Line  Type1,  40]              -­  [Line  Type2,  120]  

Some  data  are  closely  related  to  each  other  and  can  naturally  be   grouped  in  a  text  data  file:   3DBuoys:      -­  Name:  3D  Buoy1          InitialPosition:  [0,  0,  10]  

33  

w

 

User  Interface,  OrcaFlex  Model  Files    

       DragArea:  [100,  100,  30]          Pen:  [4,  Solid,  Yellow]  

Without  grouping  the  file  would  be  significantly  longer:   3DBuoys:      -­  Name:  3D  Buoy1          InitialX:  0          InitialY:  0          InitialZ:  10          DragAreaX:  100          DragAreaY:  100          DragAreaZ:  30          PenWidth:  4          PenStyle:  Solid          PenColour:  Yellow   The  majority  of  grouped  data  are  X,Y,Z  components  and  we  adopt  the  convention  that  these  components  appear  in   that  order  when  grouped.   YAML   files   may   contain   comments   which   are   introduced   by   a   hash   (#)   character   followed   by   a   SPACE.   All   subsequent   text  on   the  same  line  is  comment  and  is   ignored  when   OrcaFlex   reads  a  text  data  file.  Comments  are   not   preserved  by  OrcaFlex  and  any  user  comments  in  a  manually  edited  YAML  file  opened  with  OrcaFlex  will  be  lost  if   the  file  is  saved.  Comments  are  formatted  in  green  in  the  following  example:   General:      #  Statics      StaticsMethod:  Whole  System  statics      BuoysIncludedInStatics:  Individually  Specified      #  Dynamics      StageDuration:          -­  8          -­  16      TargetLogSampleInterval:  0.1      #  Integration      SimulationIntegrationMethod:  Implicit      ImplicitConstantTimeStep:  0.1  

A  text  data  file  can  be  rather  large,  particularly  if  it  contains  vessel  hydrodynamic  data.  Code  folding  editors  can  help   somewhat,  but  even  so  such  files  can  be  awkward  to  work  with.   The  IncludeFile  identifier  allows  you  to  move   data  into  a  separate  file  which  is  then  included  in  the  main  file:   #  File:  C:\Desktop\main.yml   VesselTypes:      -­  Name:  FPSO          IncludeFile:  FPSO.yml   Vessels:      -­  Name:  Vessel1          VesselType:  FPSO   The  included  file  contains  just  the  data  for  the  vessel  type:   #  File:  C:\Desktop\FPSO.yml   Length:  240   RAOResponseUnits:  degrees   RAOWaveUnit:  amplitude   WavesReferredToBy:  period  (s)   #  ...  remainder  of  large  file  omitted  ...   As  well  as  making  the  main  file  shorter  and  more  readable,  using  this  approach  can  offer  significant  QA  benefits.   In  this   example   we  have  used  a   relative   path  and  so  the   program   will  look  for  FPSO.yml  in  the  same   directory  as  the   main  text  data  file.   A  text  data  file  saved  by  OrcaFlex  contains  some  extra  information:   %YAML  1.1  

#  Program:  OrcaFlex  9.3a  

34  

w

 

User  Interface,  OrcaFlex  Model  Files  

 

#  File:  C:\Desktop\untitled.yml   #  Created:  12:35  on  21/07/2009   #  User:  jamie   #  Machine:  holly   -­-­-­  

General:      #  Statics      StaticsMethod:  Whole  System  statics      BuoysIncludedInStatics:  Individually  Specified      #  Dynamics      StageDuration:          -­  8          -­  16      TargetLogSampleInterval:  0.1      #  Integration      SimulationIntegrationMethod:  Implicit      ImplicitConstantTimeStep:  0.1   Environment:      #  Seabed      SeabedType:  Flat      WaterDepth:  100      SeabedModel:  Linear      SeabedNormalStiffness:  100      #  Current      RefCurrentSpeed:  0.4      RefCurrentDirection:  180   ...  

The   section   between   the   -­-­-­   and   ...   lines   is   the   main   body   of   the   file   and   is   known   in   YAML   terminology   as   a   document.   Everything   else   is   in   fact   optional   and   can   be   omitted.   A   YAML   file   can   contain   multiple   documents,   separated   by   -­-­-­   lines   but  OrcaFlex   has   no   special   treatment   for   such   multi-­‐document   files   and   all   data   is   read  into   a  single  OrcaFlex  model.   The   first   line   (%YAML   1.1)   is   known   as   the   YAML   directive   and   specifies   which   version   of   YAML   the   file   adheres   to.   The   YAML   directive   can  be  omitted.   The  rest   of   the   header   contains   a  number  of   comments  detailing   the   version   of   OrcaFlex  which  created  the  file,  the  file  name  etc.  Again,  these  comments  can  be  omitted.   Ordering  issues   The  order  in  which  the  data  appear  in  a  text  data  file  is  very  important.  OrcaFlex  processes  the  file  line  by  line  in   the   order  in  which  it  appears  in  the  file.   Any  references  (e.g.  Lines   referring  to  Line   Types)  must  be  ordered  so  that  the   referenced  object  appears  before   any   references   to  it.  So  Line   Types   appear   before   Lines   in   the   file.   Similarly   Vessels   and   3D/6D   Buoys   appear   before   Lines,   Links,   Winches   and   Shapes   so   that   any   connection   references   (e.g.   a   Line   connected   to   a   Vessel)   can   be   ordered  correctly.   The  other  ordering  issue  relates  to   inactive  data.  Data  which  are  not  currently  available  are  known  as  inactive  data.   For  example,  data  relating  to  the  explicit  solver  are  inactive  when  the  implicit  solver  is  selected.  Inactive  data  cannot   be  specified  in  a  text  data  file.   This   rule   has   implications   for   the   order   in   which   data   are   presented   in   the   text   data   file.   Consider   the   following   example:   General:      InnerTimeStep:  0.01      SimulationIntegrationMethod:  Explicit   Since   the   default   integration   method   is   the   implicit   solver   the   attempt   to   set   the   explicit   time   step   (InnerTimeStep)  will  fail  because  it  is  inactive  data.  The  solution  is  to  set  the  integration  method  before  setting   the  time  step:   General:      SimulationIntegrationMethod:  Explicit      InnerTimeStep:  0.01  

35  

w

 

User  Interface,  OrcaFlex  Model  Files    

This  principle  applies  in  general  Ȃ  you  should  set  as  soon  as  possible  all  data  which  influences  whether  other  data   are  active.   Automation   Text   data   files   can   easily   be   modified   and/or   generated   by   computer   programs/scripts.   This   means   that   the   text   data   file   format,   combined   with   a   text   processing   script   language   (e.g.   Python,   Perl,   Ruby   etc.),   can   form   a   very   effective   automation   tool.   The   OrcaFlex   Spreadsheet   provides   a   simple,   yet   effective,   facility   for   automating   the   production  of  text  data  files.   Some  specialist  features  have  been  included  in  the  text  data  file  to  aid  with  automation  tasks,  as  illustrated  in  the   following  example:   BaseFile:  base.dat   Riser:      ContentsDensity:  0.8      Length[1]:  180   When  this  text  data  file  is  loaded  in  OrcaFlex  the  program  does  the  following:   1.

Opens  the  OrcaFlex  binary  data  file  named  base.dat,  located  in  the  same  directory  as  the  text  data  file.  

2.

Sets  the  contents  density  for  the  OrcaFlex  Line  called  "Riser"  to  0.8.  

3.

Sets  the  length  of  the  first  section  of  "Riser"  to  180.  

The   BaseFile   identifier   differs   from   IncludeFile   in   that   it   is   able   to   load   either   binary   or   text   data   files   (IncludeFile  only  works   with   text  data   files).   In   addition  BaseFile  clears   all   existing  data   in   the   model  before   loading   the   contents   of   the   specified   file.   On   the   other   hand,   IncludeFile   acts   incrementally,   starting   from   whatever  state  the  model  is  in  when  the  IncludeFile  identifier  is  encountered.   Standard   text   data   files   typically   specify   the   entire   model.   The   common   automation   task   of   making   systematic   variations   to   a   base   case   requires   the   ability   to   specify   an   existing   object   for   which   data   modifications   are   to   be   made.  This  is  done  using  the  object's  name  Ȃ  in  the  example  above  the  Riser:  line  performs  this  step.   In  a  similar  vein  it  is  a  common   requirement  to  modify   data  for  certain  items  in  a  list  or   table   without  specifying  the   entire  table.  The  indexing  syntax  (Length[1]  in  the  example)  performs  this  task.  Note  that  as  for  batch  script  files  the   indices  are  always  1-­‐based.   Manually  edited  text  data  files   Saving  a  text  data  file,  then  editing  it  is  a  good  way  to  create  a  base  file  for  automation,  or  to  discover  data  names   and   data   structure   for   an   object.   However,   please   be   aware   that   this   is   a   one   way   process.   OrcaFlex   reads   and   interprets  a  text  data  file  line  by  line  to  build  the  model  incrementally,  discarding  the  lines  once  processed.  When   saving   a   file   OrcaFlex   exports   each   object,   including   any   default   values.   Consequently   the   save   process   is   not   the   inverse   of  the   load   process   and   any   manual   modifications   to   the   input   file   will   be   overwritten   when   the   file   is   saved   by  OrcaFlex.   In  the  short  automation  example  above,  if  the  model  created  when  this  file  is  loaded  is  saved,  the  text  data  file  would   contain   data   for   all   the   objects   imported   by   the   BaseFile   command,   the   full   data   for   the   line   Riser   and   other   default  data  not  specified  in  the  input  file.  

3.2.3

Simulation  Files  

Results  from  OrcaFlex  calculations  (statics  or  dynamics)  are   saved  to  simulation  files  (.sim).  These  are  binary  files   containing  the  following  sections:   x

The  model  data.  This  section  is  essentially  a  binary  data  file.  

x

The   latest   calculated   state   (positions,   loads   etc.)   of   the   model.   This   section   allows   static   state   results   to   be   retrieved  and  also  enables  partially-­‐run  dynamic  simulations  to  be  continued.  

x

The   log   file   which   contains   results   for   a   dynamic   simulation.   This   section   is   not   present   for   static   state   simulation  files.  

Simulation  files  can  be  generated  in  a  number  of  different  ways:   x

Interactively   from   the   main   OrcaFlex   window.   After   a   calculation   (statics   or   dynamics)   has   performed   then   a   simulation  can  be  saved  using  the  File  |  Save  or  File  |  Save  As  menu  items.  

36  

w

 

User  Interface,  Model  Browser  

 

x

From  the  batch  processing  form.  

x

From  Distributed  OrcaFlex.  

x

From  the  OrcaFlex  programming  interface  (OrcFxAPI).  

Similarly,  results  can  be  post-­‐processed  from  simulation  files  in  a  number  of  different  ways:   x

Interactively  from  the  results  form.  

x

From  the  OrcaFlex  spreadsheet.  

x

From  the  OrcaFlex  programming  interface.  

3.3

MODEL  BROWSER  

At   any   time   you   can   use   the   Model   Browser   to   see   what   objects   you   have   in   your   model.   To   display   the   model   browser,  use  the  model  browser  button   (F6  to  open  the  model  browser).  

 or  the  Model  |  Model  Browser  menu  item  or  use  the  keyboard  shortcuts  

  Figure:  

Model  Browser  

The  Model  Browser  consists  of  a  list  of  all  the  objects  in  the  model,  arranged  into  categories  according  to  object  type.   Several  symbols  are  used  in  the  list  of  objects:    

Categories  can  be  opened,  to  show  their  contents,  or  closed,  to  simplify  viewing  a  complex  model.    

 

Objects.  Use  double  click  to  view  or  edit  the  object's  data.    

 

Locked.  These  objects  cannot  be  dragged  by  the  mouse  in  the  3D  View.    

37  

w

 

User  Interface,  Model  Browser    

You   can   navigate   the   list   and   select   the   object   required   by   clicking   with   the   mouse,   or   using   the   arrow   keys   and   return.  If  the  list  is  longer  than  the  window  then  you  can  either  enlarge  the  window  or  use  the  scroll  bar.   Note:  

More  than  one  object  can  be  selected  in  the  model  browser.  This  allows  you  to  perform  the  same   action  (e.g.  delete,  copy,  hide,  show,  locate)  on  many  objects  at  once.  To  select  more  than  one  object   you   use   the   standard   Windows   key   presses   CTRL+CLICK   to   add   to   a   selection   and   SHIFT+CLICK   to   extend  a  selection.  

Hint:  

If  you  have  all  objects  in  the  model  browser  selected  then  it  can  be  difficult  to  de-­‐select  them.  The   simplest   way  is   to   use   CTRL+CLICK  to  de-­‐select   one   item   and   then   to   CLICK  that   item   again   to   select   it  alone.  

Model  Browser  Facilities   The  model  browser  menus,  and  its  pop-­‐up  menu,  provide  the  following  model  management  facilities.  For  details  of   keyboard  shortcuts  see  Keys  on  Model  Browser.   Add  

Add  a  new  object  to  the  model.   Delete  

Delete  the  selected  object  from  the  model.   Cut/Copy  

Cut  or  Copy  the  selected  object  to  the  clipboard.   Paste  

Paste   an   object   from   the   clipboard   into   the   model.   If   the   object   is  the  Variable   Data  then  all  the   variable   data   tables   are  pasted  in,  with  tables  being  renamed  if  necessary  to  avoid  clashing  with  existing  variable  data  n ames.   Note:  

You  can  use  Cut/Copy  and  Paste  to  transfer  objects  between  two  copies  of  OrcaFlex  running  on  the   same   machine.   You   can   also   use   it   to   transfer   objects   between   two   OrcaFlex   data   files   (open   the   source  file   and   copy  the  object   to   the   clipboard,   t hen   open   the   destination   file   and   paste   the   object   back  from  the  clipboard),  but  the  Library  facility  (see  below)  provides  an  easier  way  of  achieving   the  same  thing.  

Move  Selected  Objects  

Opens  the  Move  Selected  Objects  Wizard.   Locate  

Finds   and   highlights   the   object   in   any   open  3D   view   windows.   This   is   useful   in   complex   models   where   many   objects   are  on  the  3D  view.  The  highlighting  method  is  determined  by  the  Locate  Object  Method  preference.   Edit  

Open   the   object's   data   form.   This   action   can   also   be   invoked   by   double-­‐clicking   an   item,   or   by   selecting   it   and   pressing  RETURN.   Rename  

Rename  the  selected  object.  You  can  also  rename  by  single-­‐clicking  the  selected  object.   Lock/Unlock  

Lock  or  unlock  the  selected  object.   Hide/Hide  All/Show/Show  All  

Control  whether  the  objects  are  drawn  on  3D  views.   Reorder  

You  can  use  drag+drop  with  the  mouse  to  reorder  objects  in  the  model.  This  is  useful  if  you  are  working  on  the  static   position  of  one  particular  line  Ȃ  you  can  drag  it  up  to  the  top  of  the  list  of  lines,  so  that  it  will  be  tackled  first  when   OrcaFlex  does  the  static  analysis.  

38  

w

 

User  Interface,  Model  Browser  

 

Library  

The  Library  menu  facilities  allow  you  to  open  a  second  data  file.  You  can  then   Import  objects  from  that  second  file   into  the  current  model.  You  can  also  import  using  drag+drop  with  the  mouse.  For  details  see  Libraries.   Notes:  

The  second  data   file  is   referred   to   as   the   library   model,   but   in   fact   it   can  be   any   OrcaFlex   data   file.   The   library   facilities   therefore   provide   an   easy   way   to   move   objects   between   different   OrcaFlex   data  files.  

 

If  the  object  being  imported  is  the  variable  data  then  all  the  variable  data  tables  are  transferred,   with  tables  being  renamed  if  necessary  to  avoid  clashing  with  existing  variable  data  names.  

Switch  to  Main  Window  

The  browser's   Window   menu  enables   you   to   switch  focus   to   the   main   form  without  closing   the  browser   window.   A   corresponding  command  on  the  main  form's  Window  menu  switches  focus  back.  

3.3.1

Model  Browser  Views  

There  are  2  ways  of  viewing  objects  in  the  model  browser:  by  Types  or  by  Groups.  You  can  switch  between  views   by  clicking  on  the  model  browser  View  |  View  by  Groups/Types  menu  items,  or  though  the  popup  menu.   Types  View  

This   is   the   traditional   model   browser   view.   The   browser   has   a   number   of   folders   containing   objects   of   the   same   type.   For   example   all   the   lines   are   contained   in   a   folder  called   "Lines".   Objects   can  be  reordered   within   a   folder  but   they  cannot  be  moved  to  a  different  folder.   To  select  this  view  you  should  click  the  View  |  View  by  Types  menu  item.   Groups  View  

This  view  allows  you  to  customise  how  the  objects  are  arranged  in  the  model  browser.  You  can  add  any  number  of   browser   groups   to   the   browser.   These   groups   are   simply   folders   in   the   browser   tree.   Groups   can   contain   any   number  of  objects  or  other  groups.  In  this  way  a  hierarchical  structure  for  the  model  can  be  created.   To  select  this  view  you  should  click  the  View  |  View  by  Groups  menu  item.   To  add  groups  you  select  the  Edit  |  Add  Group  menu  item  or  use  the  popup  menu.  Groups  can  be  renamed  in  the   same   way   as   other  objects.  Objects   can  be   added   to   a  group  by  dragging   the  objects  onto   the   group.   Any  number   of   objects  can  be  added  to  a  group  in  one  operation   by  first  selecting  the  objects  and  then  d ragging  them.  This  multiple   selection  is  performed  using  the  standard  Windows  key  presses   CTRL+CLICK  to  add  to  a  selection  and  SHIFT+CLICK  to   extend   a   selection.   Groups   can   be   dragged   into   other   groups   and   so   a   hierarchical   structure   for   the   model   can   be   created.   As   well   as   allowing   you   the   freedom   to   structure   your   model   however   you   like,   the   Groups   View   allows   you   to   perform   the   same   action   (e.g.   delete,   copy,   hide,   show,   locate)   on   all   objects   in   a   group.   The   grouping   structure   is   also  used  when  cycling  through  data  forms-­‐  clicking  the  Next  button  takes  you  to  the  next  object  in  the  groups  view.  

3.3.2

Move  Selected  Objects  Wizard  

This  wizard  allows  you  to  move  and  rotate  a  number  of  objects  en  masse.  The  wizard  is  most  useful  w hen  you  select   multiple  objects,  a  group  or  a  number  of  groups  or  even  the  entire  model.   To  use   the   wizard  you  must   first  open  the   Model  Browser  and  select  the  objects  which  you  wish  to  move.   Then  click   Move  Selected  Objects  on  the  browser's  edit  menu  (also  available  from  the  popup  menu).   Selecting  objects   Before  using  the   wizard  you  must  select  (in  the  model  browser)  the  objects  which  you  wish  to  move.  There  are  a   variety  of  ways  in  which  you  can  do  this.  We  list  a  few  of  the  more  useful  methods  below:   x

Select  a  single  object.  

x

Select   multiple   objects.   You   can   do   this   in   the   model   browser   using   CTRL+CLICK   to   add   to   a   selection   and   SHIFT+CLICK  to  extend  a  selection.  

x

Select  an  object  type   folder.   This  works  when  the  model  browser  is  in  Types  View  mode.  For   example  select  the   Lines  folder  if  you  wish  to  move  all  the  lines  in  a  model.  

39  

w

 

User  Interface,  Libraries    

x

Select  a  group.  This  works  when  the  model  browser  is  in  Groups  View  mode.  This  allows  you  to  move  all  objects   in  that  group.  

x

Select   the   entire   model.   This  is  easiest   to   do   when   the   model   browser   is  in  Groups   View  mode.   The   first   item   in   the  model  browser  is  titled  "Model".  Select  this  item  if  you  wish  to  move  all  objects  in  the  model.  

There   is   no   limitation   to   the   type   of  selections   you   can   make.   If   y ou   wish   to   move   2   groups   then   select   both   of  them   (using  CTRL+CLICK)  and  open  the  wizard.   Note:  

If   your   selection   includes   an   item   which   contains   other   objects   (e.g.   a   group   or   an   object   type   folder)  then  all  objects  contained  by  that  item  will  be  moved  by  the  wizard.  

Points   The  wizard  shows  a  list  of  the  points  associated  with  each  selected  object.  For  objects  like  buoys,  vessels  and  shapes   a   single   point   is   shown.   For   objects   like   lines,   links   and   winches   with   multiple   connection   points   the   list   shows   each   connection  point  for  that  object.  The  list  also  shows  the  global  coordinates  of  each  point.   For  each  point  you  have  the  option  of  including  or   excluding  it  in  the   move   operation.  This  might  be  useful  if  you   wanted   to   move   only   the   End   A   line   connection   points   and   leave   the   End   B   connection   points   unchanged,   for   example.   Move  specified  by   There  are  4  methods  of  specifying  how  the  objects  are  moved.   Displacement  

For  this  method  you  specify  a  position  change  (i.e.  a  displacement)  which  will  be  applied  to  all  the  points  included  in   the  move  operation.   Polar  Displacement  

This   method   is   similar   to   the   Displacement   method.   Here   you   specify   a   direction   and   distance   which   determine   a   position  change.  This  is  applied  to  all  the  points  included  in  the  move  operation.   New  Position  

Here   you   give   a  reference   point   and   its  new  position.  The   same   displacement   is  applied   to   all   other   points  included   in  the  move.   Rotation  

This   method   rotates   the   included   points   in   the   horizontal   plane.   You   specify   an   angle   of  rotation   and   a   central   point   about   which   the   rotation   is   performed.   Note   that   the   environment   data   (e.g.   wave   and   current   directions,   seabed   direction  etc.)  is  not  included  in  the  rotation.   Moving  the  objects   Once  you  have  decided  which  objects  to  include  in  the  move  and  how  the  move  is  specified  you  are  ready  to  actually   move   the   objects.   This   is   done   by   clicking   the   Move   button.   If   you   change   your   mind   and   decide   not   to   move   the   objects  then  simply  click  the  Close  button.  

3.4

LIBRARIES  

An  OrcaFlex  Library  is  a  collection  of  OrcaFlex  objects  (line  types,  lines,  buoys  etc.)  stored  in  an  ordinary  OrcaFlex   data  file.  For  example,  a  library  may  contain  all  the  standard  Line  Types  that  you  use  regularly.  Once  such  a  library   file   has   been   built   you   can   quickly   build   new   models   using   the   library   Ȃ   this   gives   faster   model   building   and   can   make  QA  procedures  safer.   To   open   a   library   file,   use   the   File   |   Libraries   menu   or   the   Library   menu   on   the   Model   Browser.   Note   that   any   OrcaFlex  data  file  can  be  opened  as  a  library  file,  and  this  makes  it  easy  to  use  the  model  browser  to  copy  objects   from  one  model  to  another.  

3.4.1

Using  Libraries  

Libraries   allow   you   to   easily   import   objects  from   one   OrcaFlex   model   to   another.   To   do   this  run  OrcaFlex   and   open   the   model  browser   by  clicking  the   model   browser   button   F2.  The  model  browser  should  look  like:  

40  

 or  the  Model   |  Model  Browser  menu  item,  or  pressing  

w

 

User  Interface,  Libraries  

 

  Now  you  open  your  file  as  a   library.   To  do  this  click  the  open  button   file.  Now  the  model  browser  will  look  like:  

 on  the  model   browser  and  select  your   data  

  We   are   now   going   to   copy   some   objects   from   the   right   hand   pane   to   the   left   hand   pane.   To   do   so   select   the   required   line  types  and  click  the  import  button   .  As  an  alternative  to  the  import  button  the  objects  can  be   dragged  from   the  right  hand  pane  to  the  left  hand  pane  or  the  Library  |  Import  menu  item  can  be  used.  

41  

w

 

User  Interface,  Libraries    

Note   that   you   can   select   a   number   of   objects   and   import   them   all   in   one   go.   You   do   this   by   using   the   standard   Windows   key   presses   CTRL+CLICK   to   add   to   a   selection   and   SHIFT+CLICK   to   extend   a   selection.   If   you   do   this   the   library  will  look  like:  

  Once  you  have  imported  the  required  objects  you  can  close  the  library  by  selecting  the  Library  |  Close  menu  item   on  the  model  browser.  Now  the  model  browser  looks  like:  

42  

w

 

User  Interface,  Menus  

 

  Here  are  some  other  points  about  using  library  files:   x

Because   library   files   are   simply   ordinary   OrcaFlex   data   files,   you   can   temporarily   treat   any   OrcaFlex   data   file   as   a  library.  This  allows  you  to  import  objects  from  one  OrcaFlex  data  file  to  another.  

x

You   can   re-­‐size   the   model   browser   by   dragging   its   border.   You   can   also   control   the   relative   sizes   of   its   two   panes,  by  dragging  the  right  border  of  the  left  pane.  

x

You   can   view,   but   not  edit,   the   data   for   a   library  model  object,  by  double   clicking  it  in   the   Model   Browser  or  by   selecting  it  and  using  the  pop-­‐up  menu.  

x

When  an  object  is  imported  from  a  library,  the  current  model  may  already  have  an  object  of  that  name.   In  this   case  OrcaFlex  automatically  gives  the  object  a  new  name  based  on  the  old  name.  

3.4.2

Building  a  Library  

A  library  file  is  simply  an  OrcaFlex  data  file   Ȃ  you  can  use  any  OrcaFlex  data  file  as  a  library.  In  practice  it  is  most   convenient  to  put  your  commonly  used  OrcaFlex  objects  into  files  designated  as  OrcaFlex  library  files.   You  build  a  library  file  in  the  same  way  as  you  build  a  standard  OrcaFlex  data  file.  Starting  with  a  blank  model  you   can  add  objects  in  the  usual  way  and  set  their  data.  Typically,  however,  you  would  want  to  reuse  objects  that  had   previously  been  created  and  used  for  a  project.   To  do  this  you   would   open  the   model  browser  and  load   your   project  data  file  as  a  library  using  the  open  button     on   the   model   browser.   Then   you   import   the   required   objects   as   described   in  Using   Libraries.   This   procedure   can   be   repeated  with  a  number  of  different  data  files  until  you  have  all  the  objects  you  wish  to  keep  in  the  library.  Then  you   should  close  the  model  browser  and  save  the  data  file  by  clicking  the   data  file  can  now  be  used  as  a  library.  

 button  on  the  main  OrcaFlex  form.  This  

Notes:  

Because   they   are   OrcaFlex   models,   libraries   contain   General   and   Environment   data,   but   these   would  not  usually  be  used,  except  perhaps  for  the  General  data  Comment  field,  which  can  act  as  a   title  for  the  library.  

 

Because   the   library   file   is   just   an   ordinary   OrcaFlex   data   file,   it   can   also   be   opened   using   File   |   Open.  This  allows  you  to  edit  the  data  of  the  objects  in  the  library.  

You   can   set   up   as   many   library   files   as   you   wish.   For   example   you   might   have   separate   libraries   for   Line   Types,   Attachment   Types,   Vessel   Types,   Variable   Data   Sources   etc.,   or   you   may   choose   to   use   just   one   library   for   everything.  The  model  browser's  Library  menu  contains  a  list  of  the  most  recently  used  libraries.  

43  

w

 

User  Interface,  Menus    

3.5

MENUS  

OrcaFlex  has  the  following  menus:   x

The   File   menu   has   the   file   opening   and   saving   commands,   plus   commands   for   printing   or   exporting   data   or   results  and  managing  libraries.  

x

The  Edit  menu  has  data  and  object  editing  facilities.  

x

The  Model  menu  gives  access  to  the  model  building  facilities.  

x

The  Calculation  menu  provides  commands  for  starting  and  stopping  analyses,  including  batch  processing.  

x

The  View  menu  provides  view  control.  

x

The  Replay  menu  provides  replay  control.  

x

The  Graph  menu  gives  you  access  to  facilities  related  to  the  currently  active  graph  window.  

x

The  Results  menu  leads  to  the  results  facilities.  

x

The  Tools  menu  allows  you  adjust  preferences  and  to  lock  or  unlock  objects.  

x

The  Workspace  menu  allows  you  to  save  and  restore  collections  of  view,  graph  and  spreadsheet  windows.  

x

The  Window  menu  gives  access  to  the  various  windows  that  are  available,  and  allows  you  to  adjust  the  layout  of   your  windows.  

x

The  Help  menu  leads  to  the  various  help  documentation  that  is  available.  

3.5.1

File  Menu  

 New  

Deletes  all  objects  from  the  model  and  resets  data  to  default  values.    Open  

Open  an  OrcaFlex  file  Ȃ  either  a  data  file  (.dat  or  .yml)  or  a  simulation  file  (.sim).   You   can   also   open   an  OrcaFlex  file   by   dragging   and   dropping   it  onto   the   OrcaFlex   window.   For  example   if   you   have   Windows  Explorer  running  in  one  window  and  OrcaFlex   running  in  another  then  you  can  ask  OrcaFlex  to  open  a  file   by  simply  dragging  it  from  Explorer  and  dropping  it  over  the  OrcaFlex  window.   If   you   open   a   data   file   then   OrcaFlex   reads   in   the   data,   whereas   if   you   select   a   simulation   file   then   OrcaFlex   reads   in   both   the   data   and   the   simulation   results.   To   read   just   the   data   from   a   simulation   file,   you   can   use   the   Open   Data   menu  item.    Save  

Save  an  OrcaFlex  file   Ȃ   either  a  data  file  (.dat  or  .yml)  or   a   simulation  file  (.sim)  Ȃ  to  the  currently  selected  file   name.   If  a  file  of  that  name  already  exists  then  it  is  overwritten.   If   calculation   results   (either  statics   or   dynamics)   are   available   then   a   simulation   file  will   be   saved.   Otherwise   a   data   file  will  be  saved.   Note:  

You  cannot  save  a  dynamic  simulation  while  it  is  running  Ȃ  you  must  pause  the  simulation  first.  

 Save  As  

This  is  the  same  as  Save  but  allows  you  to  specify  the  file  name  to  save  to.  If  a  file  of  that  name  already  exists  then   you  are  asked  whether  to  overwrite  the  file.   When  saving  data  you  can  choose  either  the  binary  file  format  (.dat)  or  the  text  file  format  (.yml)  from  the  Save  as   type  drop  down  list.   Open  Data  

Read   the   data   from   an   existing   data   file   or   simulation   file,   replacing   the   existing   model.   If   a   simulation   file   is   specified  then  OrcaFlex  reads  just  the  data  from  it,  ignoring  the  simulation  results  in  the  file.  

44  

w

 

User  Interface,  Menus  

 

Save  Data  

Save   the   data   to   the   currently   selected   file   name,   using   extension   .dat   or   .yml.   If   a   file   of   that   name   already   exists   then  it  is  overwritten.   Save  Data  As  

This  is  the  same  as  Save  Data  but  allows  you  to  specify  the  file  name  to  save  to.  If  a  file  of  that  name  already  exists   then  you  are  asked  whether  to  overwrite  the  file.   You  can  choose  either  the  binary  file  format  (.dat)  or  the  text  file  format  (.yml)  from  the  Save  as  type  drop  down  list.   Compare  Data  

Compares  the  data  of  two  OrcaFlex  models.  See  Comparing  Data  for  details.   Properties  

Displays  the  system  file  properties  dialog  for  the  current  file.  This  is  mainly  intended  t o  make  it  easier  to  find  the  full   path  for  files  with  long  names.   Submit  to  Distributed  OrcaFlex  

Submit   the   current   file   for   processing   by  Distributed   OrcaFlex.   For   this   option   to   be   available,   either   the   Distributed   OrcaFlex  Viewer  or  Client  must  also  be  installed  on  the  machine.   Libraries  

You   can   create   new   libraries   of   OrcaFlex   objects,   or   open   existing   libraries.   You   can   then   import   objects   from   the   library  into  your  existing  model,  or  export  objects  from  your  existing  model  to  the  library.   Export  

Display   the   Export   form,   allowing   you   to   export   Data,   3D   Views,   Graphs,   Spreadsheets   or   Text   Windows.   See   also   Copy.   Selected  Printer  

Allows  you  to  change  the  selected  printer.   Printer  Setup  

Calls  up  the  Printer  Setup  dialogue  window.  This  standard  Windows  dialogue  is  used  to  select  which  printer  to  use,   and  allows  you  to  control  the  way   that  it  is  used  Ȃ  the  details  vary  from  printer  to  printer,  and  depend  on  the  printer   manufacturer's   device   driver   currently   installed.   Please   refer   to   the   manuals   for   your   printer   as   well   as   the   Microsoft  documentation.    Print  

Display  the  Print  form,  allowing  you  to  print  Data,  3D  Views,  Graphs,  Spreadsheets  or  Text  Windows.  See  Printing.   Most  Recent  Files  

List  of  the  most  recently  used  files.  Selecting  an  item  on  the  list  causes  the  file  to  be  opened.   Exit  

Close  OrcaFlex.  

3.5.2

Edit  Menu  

 Undo  Drag  

Undo  the  most  recent  drag.  This  is  useful  if  you  accidentally  drag  an  object.    Cut  

Copies  the  current  selection  to  the  clipboard  and  then  deletes  it.  

45  

w

 

User  Interface,  Menus    

 Copy  

If  there  is  a  currently  selected  object  (see  Selecting  Objects),  then  that  object  is  copied  to  the  clipboard.  You  can  then   use  Edit  |  Paste  to  create  duplicate  copies  of  the  object.  The  data  for  the  object  is  copied  to  the  clipboard  in  text  form,   from  where  it  can  be  pasted  into  a  word  processor  document.   Note:  

After  pasting  into  a  word  processor,  you  will  probably  need  to  put  the  text  into  a  fixed  space  font   since  much  of  the  data  is  in  tables.  

If   there   is   no   currently   selected   object   then   the   currently   selected   3D  view,   text   window,   graph   or   spreadsheet   is   copied  to  the  clipboard.    Paste  

Insert   object  from   clipboard.   This  can   be   used  to   duplicate   an   object   several   times   within   the   model.   After   selecting   Paste,  the  object  is  inserted  at  the  next  mouse  CLICK  position  in  a  3D  view.   If  the  current  window  is  a  Spreadsheet  then  the  contents  of  the  clipboard  are  pasted  into  the  spreadsheet.    Delete  

If   the   active   window   is   a   3D   View   then   the   currently   selected   object   is   deleted.   Before   the   object   is   deleted,   any   connected  objects  are  disconnected,  and  any  graphs  associated  with  the  object  are  closed.   If  the  active  window  is  a  Spreadsheet  then  the  selected  cells  are  cleared.   Select  All  

Selects  all  the  cells  in  a  Spreadsheet.   Copy  All  Data  

Copy   the   whole   model   to  the   clipboard.   The   model   data   is   copied   to   the   clipboard   in   text   form,   from   where   it  can   be   pasted  into  a  word  processor  document.  

3.5.3

Model  Menu  

 Model  Browser  

Toggles  the  visibility  of  the  Model  Browser.    New  Vessel    New  Line    New  6D  Buoy    New  3D  Buoy    New  Winch    New  Link    New  Shape  

Create  new  objects.  The  mouse  cursor  changes  to  the  New  Object  symbol   .  The  object  is  placed  at  the  position  of   the  next  mouse   CLICK  within  a  3D  View.  A   three  dimensional  position  is  generated  by  finding  the  point  where  the   mouse  CLICK   position   falls  on  a  plane   normal  to  the  view  direction  and  passing  through  the  View  Centre.  Vessels  are   always  placed  initially  at  the  sea  surface.   Show  Connections  Report  

Displays  a  spreadsheet  containing  information  about  all  object  connections  in  the  model.  

46  

w

 

User  Interface,  Menus  

 

Truncate  Object  Names  

Old   versions   of   OrcaFlex   (before   7.4b)   cannot   read   files   that   contain   long   object   names,   i.e.   longer   than   10   characters.  This  menu  item  truncates  any  long  object  names  in  the  model.  You  should  do  this  if  you  wish  to  send  a   file  to  another  user  whose  version  of  OrcaFlex  is  older  than  7.4b.   Delete  Unused  Types  

Deletes   any  types   (e.g.  Line  Types,  Clump   Types  etc.)   that   are   not  in   use.   This  is  sometimes   useful   to   simplify   a   data   file,  or  to  find  out  which  types  are  in  use.   Delete  Unused  Variable  Data  Sources  

Deletes  any  variable  data  sources  that  are  not  in  use.  This  is  sometimes  useful  to  simplify  a  data  file,  or  to  find  out   which  variable  data  sources  are  in  use.   Use  Calculated  Positions  

This  menu  item  is  available  after  a  successful  static  iteration  or  when  t he  simulation  is  finished  or  paused.   If  the  model  is  in  the  statics  complete  state  then  clicking  the  menu  item  sets  the  initial  positions  of  buoys,  vessels   and   free   line   ends   to   be   the   calculated   static   positions.   This   can   be   desirable   when   setting   up   a   model,   since   the   positions  found  are  likely  to  be  good  estimates  for  the  next  statics  calculation.   If   the   model   is   in   the   simulation   paused   or   stopped   state,   then   clicking   the   menu   item   sets   the   initial   positions   of   buoys   and   free   line   ends   to   be   the   latest  positions   in   the   simulation.   This   is   useful   when   OrcaFlex   statics   fails   to  find   an  equilibrium  configuration.  In  such  cases  you  can  use  dynamics  with  no  wave  motion  to  find  the  static  equilibrium   position  and  then  click  Use  Calculated  Positions.   If   a   replay   is   active   then   clicking   the   menu   item   sets   the   initial   positions   of   buoys   and   free   line   ends   to   be   the   positions  at  the  latest  replay  time.   Use  Specified  Starting  Shape  for  Lines  

This  menu  item  is  an   extension  of   Use  Calculated  Positions.  As  well  as  setting  the  initial  positions  of  buoys,  vessels   and  free  line  ends  it  modifies  data  for  all  Lines  in  the  following  way:   1.

The  Step  1  Statics  Method  is  set  to  User  Specified.  

2.

The   User   Specified   Starting   Shape   data   are   set   to   the   calculated   node   positions.   As   described   above   these   positions  are  either  the  results  of  a  static  calculation  or  the  results  of  a  dynamic  simulation.  

Use  Static  Line  End  Orientations  

This   menu   item   is   only   available   after   a   successful   static   analysis.   Clicking   the   menu   item   sets   the   line   end   orientation  data,  for  all  line  ends  in  the  model  that  have  zero  connection  stiffness,  to  the  orientations  found  in  the   static  analysis.  This  is  done  as  follows.   x

For   any   line   end   with   zero   bend   connection   stiffness,   the   end   azimuth   and   end   declination   will   be   set   to   the   azimuth  and  declination  of  the  end  node,  as  found  by  the  static  analysis.  

x

If   the   line   includes   torsion   and   the   line   end   connection   twist   stiffness   is   zero,   then   the   end   gamma   will   be   set   to   the  gamma  of  the  end  node,  as  found  by  the  static  analysis.  

This   action   can   be   useful   if  you   want   to   set   the   line   end   orientation   to   that   which   gives   zero   end   moments   when   the   line  is  in  its  static  position.  To  do  this  first  set  the   end  connection  stiffness  values  to  zero,  then  run  the  static  analysis   and  then  click  the   Use  Static  Line  End  Orientations  menu  item.   You  can  then  set  the  end  connection  stiffness  to   their  actual  values.  

3.5.4

Calculation  Menu  

 Single  Statics  

Start   the   single   statics   calculation   (see  Static   Analysis).   Progress   and   any  error   messages   that   occur   are  reported   in   the   Statics   Progress   Window,   which   is   shown   as   a   minimised   window   icon.   The   statics   calculation   can   be   interrupted  by  CLICKING  the  Reset  button.  

47  

w

 

User  Interface,  Menus    

Multiple  Statics  

Starts  the  multiple  offset  statics  calculation  (see   Multiple  Statics).  Progress  and  any  error  messages  that  occur  are   reported  in  the  Statics  Progress  Window,  which  is  shown  as  a  minimised  window  icon.  The  statics  calculation  can  be   interrupted  by  CLICKING  the  Reset  button.    Run  Dynamic  Simulation  

Start   a   dynamic   simulation   (see  Dynamic   Analysis).   If   necessary,  OrcaFlex   will   automatically   do   a   statics   calculation   first.  During  the  simulation,  the  Status  Bar  shows  the  current  simulation  time  and  an  estimate  of  the  time  that  the   simulation  will  take,  and  all  3D  View  windows  and  Graphs  are  updated  at  regular  intervals.    Pause  Dynamic  Simulation  

Pause  the  simulation.  To  save  the  results  of  a  part-­‐run  simulation  you  need  to  pause  it  first.  The  simulation  can  be   restarted  by  CLICKING  the  Run  button.   Extend  Dynamic  Simulation  

This  facility  is  only  available  when  the  current  simulation  is  either  paused  or  completed.  It  adds  another  stage  to  the   current   simulation,   without   having   to   reset.   You   are   asked   to   specify   the   length   of   the   new   stage.   You   can   then   continue  the  simulation,   without  having  to  restart   it  from  scratch.  This  is  particularly  useful  if  you  have  a  simulation   that  has  not  been  run  for  long  enough.   Note   that   data   for   the   new  stage,   e.g.   for  winch   control   and   vessel   prescribed   motion,   are   initially   set   to   be   the   same   as   for   the   previous   stage.   However,   the   data   for   the   new   stage   can   be   edited   because   the   new   stage   has   not   yet   started.    Reset  

Reset  the  model,  discarding  any  existing  results.  The  model  can  then  be  edited  or  a  new  model  loaded.    View  Warnings  

Displays   a   window   allowing   you   to   review   all   warnings   displayed   by   OrcaFlex   during   a   calculation   (statics   or   dynamics).   This   feature   is   particularly   useful   for   simulations   run   in   batch   mode   or   by   Distributed   OrcaFlex.   In   these   circumstances  warnings  are  not  displayed  since  to  do  so  would  require  user  intervention.   Line  Setup  Wizard  

Opens  the  Line  Setup  Wizard.  The  wizard  is  only  available  when  the  current  simulation  is  in  Reset  state.   Wave  Scatter  Conversion  

Opens   the   Wave   Scatter   Conversion   form.   This   facility   converts   a   scatter   table   of   sea   states   to   a   scatter   table   of   regular  (i.e.  individual)  waves.   Batch  Processing  

Run  a  batch  of  analyses  automatically  while  the  program  is  unattended.  See   Batch  Processing  for  details.  

3.5.5

View  Menu  

 Change  Graphics  Mode  

Toggles  the  graphics  mode  between  wire  frame  and  shaded.    Edit  View  Parameters  

Adjust  the  View  Parameters  for  the  active  3D  View.    Rotate  Up  /  Down  /  Left  /  Right  

Change  the  view  direction,  for  the  active  3D  View,  by  the  view  rotation  increment.  

48  

w

 

User  Interface,  Menus  

 

Plan  

Set  the  active  3D  View  to  a  plan  view  (Elevation  =  90°).   Elevation  

Set  the  active  3D  view  to  an  elevation  view  (Elevation  =  0°).   Rotate  90  /  Rotate  -­‐90  

Increase  (or  decrease)  the  view  azimuth  by  90°,  for  the  active  3D  view.    Zoom  In  /  Zoom  Out  

Click  the  zoom  button  to  zoom  in  (decrease  view  size)  or  SHIFT+CLICK  it  to  zoom  out  (increase  view  size).   Reset  to  Default  View  

Set  the  view  parameters  for  the  active  3D  View  to  be  the   default  view  of  the  model.   Set  as  Default  View  

Set  the  default  view  of  the  model  to  be  the  view  parameters  of  the  active  3D  View.   Show  Entire  Model  

Set  the  view  parameters  for  the  active  3D  View  so  that  the  entire  model  will  be  displayed.   Axes  

This  submenu  gives  you  control  of  the  3D  View  Axes  Preferences.   Superimpose  Times  

Allows  model  configurations  for  different  times  of  the  simulation  to  be  superimposed  in  3D  Views.  See  Superimpose   Times.   Current  Position  

Draws  the  model  at  the  latest  time  Ȃ  this  action  is  used  to  cancel  the  Superimpose  Times  view.  

3.5.6

Replay  Menu  

 Edit  Replay  Parameters  

Adjust   the   Replay   Parameters,   such   as   the   period   of   simulation   to   replay,   the   time   interval   between   frames,   the   replay  speed  etc.  For  more  information  see  Replays.    Start  /  Stop  Replay  

Starts  or  stops  the  replay.    Step  Replay  Forwards,  Step  Replay  Backwards  

Step  the  replay  forwards  or   backwards  one   frame  at  a  time.  Click  the  button  to  step   forwards;   CLICK   with   SHIFT   held   down  to  step  backwards.   Replay  Faster  /  Slower  

Increase  or  decrease  the  replay  frame  rate  (replay  speed).   Export  Video  

Exports  the  replay  as  a  video  clip  in  AVI  file  format.  See   Replays  for  more  details.  

3.5.7

Graph  Menu  

Use  Default  Ranges  

Sets  the  graph  axes  to  their  original  ranges  

49  

w

 

User  Interface,  Menus    

Values  

Displays  a  spreadsheet  containing  the  numerical  values  on  which  the  graph  is  based.   Spectral  Density  (only  available  for  time  history  graphs)  

Opens  a  new  spectral  density  graph.   Empirical  Cumulative  Distribution  (only  available  for  time  history  graphs)  

Opens  a  new  empirical  cumulative  distribution  graph.   Rainflow  half-­‐cycle  Empirical  Cumulative  Distribution  (only  available  for  time  history  graphs)  

Opens  a  new  rainflow  half-­‐cycle  empirical  cumulative  distribution  graph.   Properties  

Opens  the  graph  properties  form  (which  can  also  be  opened  by  double  clicking  the  graph).  

3.5.8

Results  Menu  

 Select  Results  

Display   the   results   form   which   allows   you   to   choose   from   the   currently   available   selection   of   graphs   and   results   tables.  Graphs  such  as   Time  Histories,  XY  Graphs  and  Range  Graphs  may   be  created  before  a  simulation  has   been   run,  thus  allowing  you  to  watch  the  variables  during  a  simulation.   Fatigue  Analysis  

Opens  the  Fatigue  Analysis  form.   Modal  Analysis  

Opens  the  Modal  Analysis  form.   Report  Vessel  Response  

Opens  the  Vessel  Response  form.  

3.5.9

Tools  Menu  

 Lock  /  Unlock  Selected  Object  

Locking   an   object   prevents   it   from   being   accidentally   dragged   or   connected   using   the   mouse   on   3D  views,   for   example  if  you  nudge  the  mouse  slightly  while  trying  to  DOUBLE  CLICK.  Lock  /  Unlock  Selected  Object  toggles  the  lock   on   the   currently  selected   object.   If   the   lock   is  on,   it   will   be   switched   off.   If   the   lock   is   off,   then   it   will   be   switched   on.   Locked  Objects  may  still  have  their  positions  edited  in  the  data  Edit  Forms.  The  status  of  the  object  locks  is  shown   by  symbols  in  the  Model  Browser.   Lock  /  Unlock  All  Objects  

Locks  or  unlocks  all  objects  in  the  model.   Set  Thread  Count  

Allows  you  to  change  the  number  of  execution  threads  used  by  OrcaFlex  for  parallel  processing.   Preferences  

Allows   you   to   control   various   program   settings   so   that   you   can   customise   the   program   to   the   way   you   prefer   to   work.  See  Preferences.  

3.5.10

Workspace  Menu  

Open  Workspace  

Opens  a  previously  saved  workspace  file  and  restores  the  window  layout  described  in  that  workspace  file.   Save  Workspace  

Save  the  current  window  layout  to  a  workspace  file.  

50  

w

 

User  Interface,  Menus  

 

Make  default  for  this  file,  Make  default  for  this  folder  

Makes   the   current   window   layout   the   default   workspace   for   the   current   simulation   file   or   for   the   current   folder,   respectively.  The  default  workspace  for  a  simulation  file  will  be  restored  whenever  you  open  that  file.  The  default   workspace  for  a  folder  will  be  restored  whenever  you  open  any  simulation  file  in  that  folder.   If  a  default  workspace  exists  for  a  both  a  file  and  the  folder  containing  the  file,  then  the  default  for  the  file  is  used.   Use  file  default,  Use  folder  default  

Applies   the   default   workspace   to   the   current   model.   This   is  useful   if   you   have  changed   the  window   layout   and   wish   to  restore  the  default  workspace  layout  without  re-­‐loading  the  model.   Remove  file  default,  Remove  folder  default  

Deletes  the  default  workspace.   Most  Recent  Files  

List  of  the   most   recently  saved  workspaces  in   the  directory  which  contains   the   current   model.   Selecting   an   item   on   the  list  causes  the  workspace  to  be  loaded.  

3.5.11

Window  Menu  

 Add  3D  View  

Add  another  3D  View  Window.  Having  multiple  views  on  screen  allows  you  to  watch  different  parts  of  the  system   simultaneously,  or  to  see  different  views  at  the  same  time  (for  example  a  plan  and  an  elevation).   Tile  Vertical,  Tile  Horizontal  

Arranges  all  the  windows  (3D  View,  graph  or  spreadsheet)  so  that  they  fill  the  main  window  area  and  fit  side  by  side   without  overlapping.  The  program  automatically  tiles  windows  every  time  a  new  window  is  created  or  deleted.   Switch  to  Model  Browser  

This  command,  and  the  corresponding  command  on  the   model  browser's  Window  menu,   enable  you  to  switch  focus   between  the  main  form  and  the  model  browser  window.   Statics  Progress  

Displays  the  Statics  Progress  Window.   Window  List  

This  is  a  list  of  all  currently  open  windows.  If  a  window  is  hidden  under  others  it  can  be  selected  easily  from  this  list.  

3.5.12

Help  Menu  

 OrcaFlex  Help  

Opens  the  OrcaFlex  on-­‐line  help  system.   What's  New  

Gives  a  list  of  recent  improvements  and  alterations  to  OrcaFlex.   Tutorial  

Opens  the  help  file  at  the  start  of  the  OrcaFlex  tutorial.   Examples  

Opens  the  help  file  at  the  introduction  to  the  OrcaFlex  Examples  topics.   Keyboard  Shortcuts  

Lists  the  keyboard  shortcuts  used  by  OrcaFlex.   Orcina  Home  Page  

Opens  the  Orcina  homepage  (www.orcina.com).  

51  

w

 

User  Interface,  3D  Views    

About  

Displays   a   window   giving   the   program   version,   details   about   Orcina   Ltd   and   various   other   miscellaneous   information.  

3.6

3D  VIEWS  

3D  Views  are  windows  showing  a  spatial  representation  of  the  model.  Two  distinct  types  of  3D  View  are  available:   wire   frame   shows   an   isometric   projection   of   the   model;   shaded   draws   the   model   as   solid   objects   with   lighting,   shading,  perspective  and  hidden  line  removal.  

  Figure:  

A  wire  frame  3D  View  (left)  alongside  a  shaded  3D  View  (right)  

3D   View   windows  may  be   rotated,  zoomed  and  panned  to   allow   any  aspect   of   the   system  to  be   viewed.  The   view   is   controlled  by  a  number  of  View  parameters  Ȃ  see  View  Parameters  Ȃ  and  the  caption  of  a  3D  View  window  shows   the  current  View  Azimuth  and  View  Elevation  values,  while  a  scale  bar  in  the  view  indicates  the  current  View  Size.   Multiple   view   windows   may   be   placed   side-­‐by-­‐side   so   that   you   can   view   different   parts   of   the   system   simultaneously   or   view  from   different   angles   (for   example   a   plan   and   elevation   view).   This   allows   you   to   build   non-­‐ in-­‐plane   models   on   screen   with   the   mouse.   Further   3D   View   windows   are   added   by   using   the   Window   |   Add   3D   View  menu  item   or   by   CLICKING  on   the  Add   3D   View  button   on   the   tool   bar.   Windows   may  be  arranged  by  dragging   their  borders  or  using  the  Window  |  Tile  Vertical/Horizontal  menu  items.  3D  Views  may  be  closed  by   CLICKING  the   cross  at  the  top  right-­‐hand  corner.   The   objects   in   a   3D   view   are   "live"   in   the   sense   that   you   can   use   the   mouse   pointer   to   select   objects,   drag   them   around  in  the  view  and  make  connections  between  objects.  See   Selecting  Objects,  Creating  and  Destroying  Objects,   Dragging  Objects,  Object  Connections,  for  details.  If  you   DOUBLE  CLICK  on  an  object  then  the  data  form  for  that  object   appears,  so  that  you  can  examine  or  edit  its  data.   Note:  

When   using   the   shaded   view   objects   cannot   be   selected,   dragged   etc.   For   this   reason,   the   wire   frame  view  is  most  useful  when  building  your  model.  

After   running   a   simulation,  or   loading   a   simulation  file,  a   dynamic  replay  (animation)   can  be   shown   in  one   or  more   3D  View  windows.  A  replay  shows  a  sequence  of  snapshots  of  the  model  taken  at  specified  intervals  throughout  part   or  all  of  the  simulation.  Replays  may  be  played  in  just  one  3D  View  window,  or  in  all  of  them  simultaneously  Ȃ  see   Preferences.  

52  

w

 

User  Interface,  3D  Views  

 

Finally,  3D  Views  may  be  printed  by  selecting  the  view  desired  and  using  the  print  menu.  Also,  the  picture  may  be   exported  to  a  file  or  the  windows  clipboard.   Measuring  Tape  Tool  (only  available  in  wire  frame  mode)  

You  can  measure  distance  on  a  3D  view  using  the  measuring  tape  tool.  Hold  down  the   SHIFT  and  CTRL  keys  and  then   drag   a   line   between   any  two   points  Ȃ  the   distance   between   them   is  displayed   on   the   status   bar.   Note   that   this   is  the   distance  in  the  plane  of  the  3D  view.  

3.6.1

View  Parameters  

The  view  shown  in  a  3D  view  window  is  determined  by  the  following  parameters,  which  can  be  adjusted  using  the   view  control  buttons  or  the  Edit  View  Parameters  item  on  the  View  menu.   View  Centre  

Defines  the  3D  global  coordinates  of  the  point  that  is  shown  at  the  centre  of  the  window.   View  Size  

The   diameter  of   the   view   area.   It  equals   the  distance   represented  by  the   smaller   of  the   2   sides   of   the   view   window.   This  parameter  must  be  greater  than  zero.   Example:  

If   the  window   on   screen   is  wider   than   it   is  high,   and   View  Size   =   100.0   then   an   object   100   units   high  would  just  fill  the  height  of  the  window.  

View  Azimuth  and  View  Elevation  

These  determine  the  direction  (from  the  view  centre)  from  which  the  model  is  being  viewed.  The  azimuth  angle  is   measured   from  the  global  X   direction  towards  the   global  Y  direction.  The   elevation  angle  is  then   measured  upwards   (downwards   for   negative   elevation   angles)   from   there.   The   view   shown   is   that   seen   when   looking   from   this   direction  Ȃ  i.e.  by  a  viewer  who  is  in  that  direction  from  the  view  centre.   Example:  

View   Elevation   +90°   means   looking   in   plan   view   from   above,   and   View   Elevation   =   0°,   View   Azimuth  =  270°  (or  -­‐90°)  means  a  standard  elevation  view,  looking  along  the  Y  axis.  

Window  Size  

You  can  adjust  the  size  of  a  3D  view  window   either  by  dragging  the  window  border,   or  by  setting  its  window  size  on   the  view  parameters  form.  The  latter  is  sometimes  useful  when  exporting  a  view  or  exporting  a  replay  video,  since  it   makes  it  easier  to  export  multiple  files  and  produce  videos  with  identical  dimensions.   Graphics  Mode  

Can  be  either  of  the  following  options:   x

Wire  frame:  the  model  is  represented  using  a  wire  frame,  isometric  projection.  

x

Shaded:  the  model  is  represented  as  solid  objects  with  lighting,  shading,  perspective  and  hidden   line  removal.  

Default  View  

Each  model  has  its  own   default  view  parameters  that  are  saved   with  the  model  data.  Whenever  a  new  3D  view  is   created,   it   starts   with   this   default   view.   You   can   set   an   existing   3D   view   to   the   default   view   by   using   the   Reset   to   Default  View  command  (on  the  view  menu  or  pop-­‐up  menu).   To   set   the   default   view   parameters,   first   set   up   a   3D   View   to   the   default   view   that   you   want   and  then   use   the   Set   as   Default  View  command  (on  the  view  menu  or  pop-­‐up  menu).  As  an  alternative  you  can  use  the  calculated  based  on   the  model  extent  option  which  results  in  a  default  view  that  is  sized  so  that  the  entire  model  will  be  displayed.  

3.6.2

View  Control  

You  can  adjust  the  view  in  a  3D  view  window  using  the  view  control  buttons:   Button     +  SHIFT  

 

Menu  Item  

Shortcut  

Action  

View  |  Rotate  Up  

CTRL+Ĺ  

Increase  view  elevation  

View  |  Rotate  Down  

CTRL+Ļ  

Decrease  view  elevation  

View  |  Rotate  Right  

CTRL+ĺ  

Increase  view  azimuth  

53  

w

 

User  Interface,  3D  Views    

Button   +  SHIFT  

  +  SHIFT  

 

Menu  Item  

Shortcut  

Action  

View  |  Rotate  Left  

CTRL+ĸ  

Decrease  view  azimuth  

View  |  Zoom  In  

CTRL+I  

Zoom  in  

View  |  Zoom  Out  

SHIFT+CTRL+I  

Zoom  out  

View  |  Change  Graphics  Mode  

CTRL+G  

Changes  graphics  mode  

CTRL+W   View  |  Edit  View  Parameters   Edit  View  Parameters     You  can  also  use  the  mouse   wheel  button  to  change  view.  Turn  the  wheel  to  scroll  the  3D  view  up  and  down.   Turn  it   with  the  CTRL  key  held  down  to  zoom  in  or  out  on  the  location  at  which  the  mouse  is  currently  pointing.  

For  more  precise  control  you  can  set  the  view  parameters  explicitly  using  the   View  Parameters  form.   Finally,  3D   views  can   also   be   controlled   using   the  View  menu  and   various   shortcut  keys  Ȃ  see  Mouse   and   Keyboard   Actions  and  Navigating  in  3D  Views.  

3.6.3

Navigating  in  3D  Views  

Moving  

Moving  in  3D  Views  can  be  achieved  by  a  variety  of  means:   x

Drag  the  3D  View  with  the  SHIFT  key  held  down.  We  call  this  direct  manipulation  of  the  view  centre  panning.  

x

Use  the  scroll  bars  on  the  3D  View.  

x

Use  the  cursor  keys  ĹĻĸĺ.  Use  these  cursor  keys  with  the  CTRL  key  held  down  to  effect  larger  shifts.  

x

Move  up  and  down  with  the  PGUP  and  PGDN  keys.  

x

Edit  the  View  Centre  in  the  View  Parameters  form.  

Rotating  

Rotating  in  3D  Views  can  be  achieved  by  a  variety  of  means:   x

Drag  the  3D   View  with  the   CTRL  key  held  down.  For  shaded  views  only  you  can  rotate  about  the   viewer  position   (as   opposed   to   rotating   about   the   view   centre)   by   holding   down   the   ALT   key   (as   well   as   the   CTRL   key)   whilst   dragging.  

x

Use  the  rotate  buttons  

x

Use  the  Rotate  Up,  Rotate  Down,  Rotate  Left  or  Rotate  Right  menu  items  or  their  shortcut  keys  CTRL+ALT+  ĹĻĸ ĺ.   For   shaded   views   only   you   can   rotate   about   the   viewer   position   by   holding   the   ALT   key   down   whilst   selecting  these  menu  items  or  shortcuts.  

x

Use   the  Plan,  Elevation,  Rotate   90   or   Rotate  -­‐90  menu  items   or   their   shortcut  keys   CTRL+P,   CTRL+E,   CTRL+Q  and   SHIFT+CTRL+Q.  

x

Edit  the  View  Azimuth  and  View  Elevation  in  the  view  parameters  form.  

.  Pressing  these  with  the  SHIFT  key  held  reverses  the  rotation.  

Zooming  

You   can   zoom   into   and   out  of   3D   Views   by   using   the   zoom  button   ,   the  zoom   menu  items  and  the   shortcut   keys   CTRL+I   and   SHIFT+CTRL+I.   In   addition,   you   can   zoom   in   or   out   using   the   mouse  wheel   button   with   the   CTRL   key   held   down.   The  following  methods  of  zooming  are  only  available  in   wire  frame  3D  Views.   Also  you  can  zoom  in  on  a  particular  region  of  interest  in  a  3D  view  by  defining  a  rectangle  around  it  on  screen  using   the   mouse.   To   do   this,   hold   the   ALT   key   down,   place   the   mouse   in   one   corner   of   the   desired   rectangle   and   press   down  the  left  mouse  button   while  dragging  the  mouse  to  the  opposite  corner.  When   you  release,  the   region  selected   will  be  expanded  to  fill  the  window.   To   zoom   out,   repeat   the   operation   holding   down   the   SHIFT   and   ALT   keys   Ȃ   the   region   shown   in   the   window   will   shrink  to  fit  into  the  rectangle  drawn.  

54  

w

 

User  Interface,  3D  Views  

 

You   can   also   zoom   in   and   out   by   a   fixed   amount,   keeping   the   same   view   centre,   by   using   ALT+CLICK   and   ALT+SHIFT+CLICK.  

3.6.4

Shaded  Graphics  

The   shaded   graphics   mode   renders   the   model   as   solid   objects   with   lighting,   shading,   perspective   and   hidden   line   removal.  

  Figure:  

Shaded  graphics  

Using  the  Shaded  Graphics  mode   To  a  large   extent  there  is   no  extra   work  required  to   build  a  model  for  the  shaded   graphics  mode.  You  are  able  to   build   a   model   or   take   an   existing   model   designed   using   the   wire   frame   mode   and   simply   change   to   the   shaded   graphics  mode  to  see  a  high  quality  shaded   rendering   of  your   model.  There  are  a  number  of  things  you  can   do  to   improve  your  experience  with  the  shaded  graphics  mode  as  described  below.   Translucency  

The  Sea  Surface  and  Seabed  are  drawn  as  textured  surfaces.  If  there  are  objects  on  the  other  side  of  these  surfaces   then  they  can  be  obscured.  These  surfaces  are  drawn  with  a   user-­‐specified  amount  of  translucency  which  allows  you   to  compensate  for  this.   Importing  3D  models  

Objects   like   Lines   are   straightforward   to   draw.   OrcaFlex   uses   the   Line   Type   contact   diameter   to   determine   the   thickness  of  each  segment  of  the  Line.   Objects  like  Vessels  present   more  difficulties.  OrcaFlex  by  default  will  draw  a  solid,  filled-­‐in  shape  based  on  the   wire   frame  data  you  have  specified.  While  this  can  be  sufficient  you  may  prefer  something  less  simplistic.  Alternatively   you   may   import   a   more   detailed   3D   model,   e.g.   the   turret   moored   FPSO   above.   You   can   import   3D   models   for   6D   Buoys,  Wings  and  Shapes  as  well  as  for  Vessels.  

55  

w

 

User  Interface,  3D  Views    

We  have  provided  a  very  basic  selection  of  generic  models  which  you  are  free  to  use.  There  are  models  of  an  FPSO,  a   turret   moored   FPSO,   an   installation   vessel,   a   semisub   and   a   subsea   template.   For   information   on   generating   and   importing  3D  models  specific  to  your  project  please  refer  to   www.orcina.com/Support/ShadedGraphics.   Viewer  Position  

Because  the  shaded  graphics  mode  uses  perspective  it  requires  the  concept  of  the   viewer  position  as  well  as  the   viewer  centre.  The  isometric  wire  frame  view  has  no  such  requirement.  OrcaFlex  defines  the  viewer  position  to  be  in   a  line  in  the  view  direction  (defined  by  the  view  azimuth  and  view  elevation)  at  a  distance  of  view  size  *  1.5  from  the   view  centre.  It  is  possible  to  rotate  the  view  around  both  the  view  centre  and  around  the  viewer  position.   Video  export  

Just   as   for  wire  frame   views   OrcaFlex   can  export   video  files   of   a  replays  in   shaded   views.  When  producing   videos   it   is  very  important  to  use  compression,  otherwise  the  video  file  size  becomes  unreasonably  large.  The  software  that   performs  this  compression  is  called  a  codec.   For   wire   frame   replays   OrcaFlex   uses   a   built-­‐in   codec   called   run-­‐length   encoding.   This   codec   is   not   suitable   for   shaded  replays  and  in  fact  there  is  no  suitable  built-­‐in  codec  in  Windows.  We  would  recommend  using  an  MPEG-­‐4   codec  of  which  many  are  available.  In  our  experience  the  freely  available  XVID  codec  performs  very  well.  The  XVID   codec  can  be  downloaded  from  www.orcina.com/Support/ShadedGraphics.   Should  you  wish  to  use  a  different  codec  you  can  select  this  from  the  Preferences  form.   Hardware  Requirements   The  shaded   graphics   mode   does   require  the  presence  of  a  DirectX  9  compatible  graphics  card.  In  our  experience   the   most  important  factor  to  consider  when  choosing  a  card  to  work  with  shaded  graphics  is  the  amount  of  memory.  We   would  recommend  using  a  card  with  256MB  or  more.   It   is   also   important   to   make   sure   that   your   computer's   graphics   settings   specify   a   colour   mode   of   16   bits   (65536   colours)  or  better.   Notes:  

If  your  machine's  graphics  capabilities  are  insufficient  then  the  shaded  graphics  mode  may  fail  to   function  properly  or  indeed  fail  to  function  at  all.  For  example,  low  quality,  blocky  images  usually   indicate  a  graphics  card  with  insufficient  memory.  This  problem  can  also  manifest  itself  by  failure   to  draw  the  sky  which  appears  plain  white.  

 

For   best   results   you   should   centre   your   model   close   to   the   global   origin.   The  Move   Selected   Objects   facility  can  help  you  do  this.  

3.6.5

How  Objects  are  Drawn  

Each   object   in   the   model   is  drawn   as   a   series   of  lines   using   the   Pen   Colour,   Line   Width   and   Style   (solid,   dashed   etc.)   defined   in   the   drawing   data   for   that   object.   You   can   change   the   pen   colours   etc.   used   at   any   time   by   editing   the   drawing  data  for  that  object.  To  change  the  pen  colour,  select  and   CLICK  the  colour  button  on  the  data  form  and  then   CLICK  on  the  new  colour  wanted.   You   can   also   exclude   (or   include)   individual   objects  from   the   3D   view,  by  opening   the  model  browser,   selecting   the   object  and  then  using  the  Hide  (or  Show)  command  on  the  browser's  Edit  or  pop-­‐up  menu.   Notes:  

In  Windows,  a  line  width  of  zero  does  not  mean  "don't  draw"  Ȃ  it  means  draw  with  the  minimum   line  width.  To  suppress  drawing  either  set  the  line  style  to   null  (the  blank  style  at  the  bottom  of  the   drop  down  list)  or  else  hide  the  object.  

 

On   some   machines   the   display   driver   cannot   draw   the   dashed   or   dotted   pen   styles   and   instead   draws  nothing.  So  on  such  machines  only  the  solid  and  blank  pen  styles  work.  

Wire  Frame  Drawing  

For  wire  frame  views  the  various  objects  are  drawn  as  follows:   x

The   various  coordinate   systems  can   be   drawn   as   small   triplets   of   lines   showing   their   origin   and   the   orientation   of  their   axes.   The   wave,   current   and   wind   directions   can   be   drawn   as   arrows   in   the   top  right   hand   corner   of   3D   views.  You  can  control  both  what  is  drawn  (see  3D  View  Drawing  Preferences)  and  the  drawing  data  used.  

x

The  Seabed  is  drawn  as  a  grid  using  the  seabed  pen.  

56  

w

 

User  Interface,  3D  Views  

 

x

The  Sea  Surface  is  drawn  as  a  grid  or  as  a  single  line.  This  is  controlled  by  the  user's  choice  of  Surface  Type  as   specified  on  the  drawing  page  on  the  Environment  data  form.  If  the  Surface  Type  is  set  to  Single  Line  then  one   line  is  drawn,  aligned  in  the  wave  direction.  If  the  Surface  Type  is  set  to  Grid  then  a  grid  of  lines  is  drawn.  This   line  or  grid  is  drawn  using  the  sea  surface  pen.  

x

Shapes  are  drawn  either  as  wire  frames  (Blocks,  Cylinders  and  Curved  Plates)  or  as  a  grid  (Planes).  As  well  as   controlling  the  pen  colour,  width  and  style,  for   shapes  you  can  also  control  the  number  of  lines  used  to  draw  the   shape.  

x

Vessels  are  drawn  as  a  wire  frame  of  edges  and  vertices  defined  by  the  user  on  the  Vessel  and  Vessel  Types  data   forms.  

x

3D  Buoys  are  drawn  as  a  single  vertical  line  of  length  equal  to  the  height  of  the  buoy.  

x

6D   Buoys   are   drawn   as   a   wire   frame   of   edges   and   vertices.   For   Lumped   Buoys,   the   vertices   and   edges   are   defined   by   the   user   on   the   buoy   data   form.   For   Spar   Buoys   and   Towed   Fish   the   vertices   and   edges   are   automatically   generated   by   OrcaFlex   to   represent   the   stack   of   cylinders   that   make   up   the   buoy.   As   an   option   Spar  Buoys  and  Towed  Fish  can  be  drawn  as  a  stack  of  circular  cylinders   Ȃ  this  is  the  default  setting.  

x

Wings  are  drawn  as  rectangles  in   either  the  6D  Buoy  pen  or  the   Wing  Type  pen  as  determined  in  the  Wing  Type   data.  

x

Lines   are   drawn   as   a   series  of   straight   lines,   one   for  each  segment,   joining   points  drawn   at  each  node.   Separate   pens  are  used  for  the  segments  and  nodes,  so  you  can,  for  example,  increase  the  pen  width  used  for  the  nodes  to   make  them  more  visible.  There  is  also,  on  the  Line  Data  form,  a  choice  of  which  pen  to  use  to  draw  the  segments.  

x

Clumps  are  drawn  as  a  thin  vertical  bar.  

x

Drag   Chains   are   drawn   using   the   colour   and   line   style   specified   on   the  attachment   types   form.   The   hanging   part   of  the   chain   is   drawn   as   a   line,   of   length  equal   to   the   hanging   length   and   at   the   angle   calculated   using   the   above   theory.  The  supported  part  of  the  chain  (if  any  is  supported)  is  separately  drawn  as  a  blob  at  the  seabed,  directly   beneath  the  node.  The  drag  chain  drawing  therefore  directly  reflects  the  way  in  which  the  chain  is  modelled.  

x

Flex  Joints  are  drawn  as  a  circular  blob  using  the  colour  and  line  style  specified  on  the   attachment  types  form.  

x

Links  and  Winches  are  drawn  as  a  straight  line  segments  joining  the  connection  points.  

Lines,  Links  and  Winches  and  Shapes  are  special  slave  objects  that  can  be  connected  to  other  master  objects  Ȃ  see   Connecting  Objects.  To  allow  these  connections  to  be   made,  each  slave  object  has  a  joint  at  each  end  that  you  can   connect  to  a  master  object  or  else  leave  Free.  When  the  program  is  in  Reset  or  Statics  Complete  state  these  joints  are   drawn  as  follows:   The  joint  at  End  A  of  a  line  or  end  1  of  a  Link  or  Winch  is  drawn  as  a  small  triangle.  The  other  joints  are  drawn  as   small  squares.  This  distinguishes  which  end  of  a  Line,  Link  or  Winch  is  which.   If   the   joint   is   connected   to   a   master   object,   then   it   is   drawn   in   the   colour   of   the   master   object   to   which   it   is   connected.  If  the  joint  is  Free,  then  it  is  drawn  in  the  colour  of  the  Line,  Link  or  Winch  to  which  it  belongs.   Shaded  Drawing  

For  shaded  views  the  various  objects  are  drawn  as  follows:   x

View  axes  and  global  axes  are  drawn  as  small  triplets  of  lines  showing  their  origin  and  the  orientation  of  their   axes.  The  wave,  current  and  wind  directions  can  be  drawn  as  arrows  in  the  top  right  hand  corner  of  3D  views.   You  can  control  both  what  is  drawn  (see  3D  View  Drawing  Preferences)  and  the  drawing  data  used.  

x

The   Sea   Surface   and   Seabed   are   drawn   as   textured   surfaces   using   their   respective   pen   colours.   Both   surfaces   can  be  drawn  with  user-­‐specified  levels  of  translucency.  

x

Shapes   are   drawn   as   solid   objects   and   Planes   allow   for   user-­‐specified   levels   of   translucency.   Alternatively   Shapes  can  be  represented  by  an  imported  3D  model.  

x

Vessels   are   drawn   as   a   solid,   filled-­‐in   shape   based   on   the   wire   frame   data.   Alternatively   Vessels   can   be   represented  by  an  imported  3D  model.  

x

3D  Buoys  and  Clumps  are  drawn  as  an  ellipsoid  with  the  specified  volume  and  height.  

x

Lumped   6D   Buoys   are   drawn   as   a   solid,   filled-­‐in   shape   based   on   the   wire   frame   data.   Spar   Buoys   and   Towed   Fish   are   drawn   as   solid   objects   using   the   specified   cylinder   geometry.   Alternatively   6D   Buoys   can   be   represented  by  an  imported  3D  model.  

57  

w

 

User  Interface,  3D  Views    

x

Wings   are   drawn   as   plates   using   their   specified   span   and   chord.   Alternatively   they   can   be   represented   by   an   imported  3D  model.  

x

Lines   are   drawn   as   a   series   of   cylinders,   one   for   each   segment   using   the   contact   diameter   as   specified   on   the   Line  Type  form.  There  is  also,  on  the  Line  Data  form,  a  choice  of  which  pen  to  use  to  draw  the  segments.  

x

Drag  Chains  are  drawn  as  a  chain  with  bar  diameter  derived  from  the  drag  chain's   effective  diameter.  

x

Flex   Joints  are   drawn   as   cylinders  with  radius   2R   and   length  4R   where   R  is  the  radius   of   the  node  to  which  the   flex  joint  is  attached.  

x

Links   and   Winches   are   drawn   as   a   series   of   cylinders   joining   the   connection   points.   The   diameter   of   the   cylinders  can  be  specified  on  the  object's  data  form.  

3.6.6

Selecting  Objects  

A  single  CLICK  on  or  near  an   object  in  a   3D  View  selects  it  ready  for  further  operations.  The  currently  selected  object   is   indicated   in   the   Status   bar.   All   objects   have   a  hot   zone  around   them.   If   several   objects   have   overlapping   hot  zones   at  the  mouse  position,  they  will  be  selected  in  turn  at  subsequent   CLICKS.   To  deselect  the  object  (without  selecting  another  object)   CLICK  on  the  3D  view  away  from  all  objects.   CLICK  on   an   object  to  open  its  data  form.  

3.6.7

Creating  and  Destroying  Objects  

When  the  model  is  in  Reset  or  Statics  Complete  state  then  you  can  create  and  destroy  objects  using  the  mouse.   To   create   a   new  object,   CLICK   on   the   appropriate   new   object  button   on   the   tool   bar  or   select   the   Model  |  New  Object   menu  item.  The  mouse  cursor  changes  to  show  this.  A  new  object  of  that  type  is  created  at  the  position  of  the  next   CLICK  on  a  3D  View.   You   can   also   create   a  new   object   by  copying   an  existing  one.   To   do  this   select   the   object   and   press   CTRL+C  to   take   a   copy  of  it.  You  can  now  press   CTRL+V  (more  than  once  if  you  want  more  than  one  copy)   Ȃ  again  the  mouse  cursor   changes  and  the  copy  object  is  pasted  at  the  position  of   the  next  mouse  CLICK  in  a  3D  view.  This  method  of  creating  a   new  object  is   particularly  useful  if  you   want  an  almost   identical  object   Ȃ  you  can  create  a  copy  of  it  and   then  just   change  the  data  that  you  want  to  differ.   To  destroy  an  object,  simply  select  it  and  then  press  the  DELETE  key.  You  will  be  asked  to  confirm  the  action.  

3.6.8

Dragging  Objects  

An   unlocked   object   may   be   dragged   to   relocate   it   by   pressing   the   mouse   button   down   and   holding   it   down   while   moving  the  mouse.  When  the  mouse  button  is  released,  then  the  object  will  be  positioned  at  the  new  location.  The   current  coordinates  of  the  object  are  shown  in  the   Status  Bar  during  the  drag  operation.   Note:  

Objects   must   be  dragged   a   certain   minimum   distance  (as  specified  in   the  Preferences  form)   before   the  drag  operation  is  started.  This  prevents  accidental  movement  of  objects  when   DOUBLE  CLICKING   etc.  

Objects   may   be   locked   to   prevent   unintended   drag   operations   moving   them   (see   Locking   an   object).   Their   coordinates  may  still  be  edited  on  their  data  form.   Note:  

Slave  objects  that  are  connected  are  moved  relative  to  their  master's  local  origin.  Other  objects  are   moved  in  the  global  coordinate  frame.  

Dragging  is  only  available  in  Reset  or  Statics  Complete  states,  and  when  the  object  is  not  locked.  

3.6.9

Connecting  Objects  

Unlocked  slave   objects  (e.g.   Lines,  Links,   etc.)  can   be  connected   to   master  objects  using  the   mouse  in  a  3D  View  (see   Object  Connections).  First  select  the  end  of  the  slave  that  you  want  to  connect  by   CLICKING  on  or  near  its  end  joint.   Then  hold  down  the   CTRL  key  while   CLICKING  on  the  master  object  Ȃ  the  two  will  then  be  connected  together.  This   operation  is  only  permitted  for  master-­‐slave  object  pairs,  for  example  connecting  a  line  to  a  vessel.  The  connection  is   indicated   in   the   Status   Bar   and   the   joint   connected   is   drawn   in   the   colour   of   the   master   object   to   show   the   connection.   To  Free  a  joint  Ȃ  i.e.  to  disconnect  it  Ȃ  select  it  and  then  CTRL+CLICK  on  the  sea  surface.   To  connect  a  joint  to  a  Fixed  Point,  select  it  and  then  CTRL+CLICK  on  the  global  axes.  

58  

w

 

User  Interface,  Replays  

 

To   connect   an   object   to   an   Anchor   (a   fixed   point   with   a   coordinate   relative   to   the   seabed),   select   it   and   then   CTRL+CLICK  on  the  seabed  grid.  If  the  object  is  close  to  the  seabed  then  the  program  snaps  it  onto  the  seabed.  This   allows   an   object   to   be   placed   exactly   on   the   seabed.   If   you   require   an   anchor   coordinate   close   to,   but   not   on   the   seabed,  connect  it  to  the  seabed  at  a  distance  and  then  drag  it  nearer  or  edit  the  coordinate  in  the   Data  Form.  

3.6.10

Printing,  Copying  and  Exporting  Views  

3D   Views   may   be   printed,   copied   to   the   windows   clipboard,   or   exported   to   a   windows   graphics   metafile,   so   that   the   pictures  may  be  used  in  other  applications  such  as  word  processors  and  graphics  packages.   First  select  the  view  and  adjust  the  viewpoint  as  desired.  Then  to  copy  to  the  clipboard   press   CTRL+C,  or  select  Copy   from  the  pop-­‐up  menu.   The   pop-­‐up  menu  also  has  commands  to  print  or   export  the  3D  view.  If  needed,  you  can   first   adjust  the  printer  setup  using  the  Printer  Setup  command  on  the  pop-­‐up  menu  or  on  the  File  menu.   If  you  are  printing  the  view  on  a  black  and  white  printer  (or  are  transferring  the  view  into  a  document  which  you   intend   to  print   on   a   black  and   white   printer)   then   it   is   often   best   to   first   set   OrcaFlex   to   output   in   monochrome   (use   the  Tools|Preferences|Output  menu  item).  This  avoids  light  colours  appearing  as  faint  shades  of  grey.   After   a   3D   view   has   been   transferred   to   another   application   you   should   be   careful   not   to   change   its   aspect   ratio,   since  this  will  produce  unequal  scaling  in  the  vertical  and  horizontal  directions  and  invalidate  the  scale  bar.  In  Word   you  can  maintain  aspect  ratio  by   dragging  the  corners   of  the  picture,  whereas  if  you  drag  the  centres  of  the  sides   then  the  aspect  ratio  is  changed.  

3.7

REPLAYS  

A  Replay  is  a   sequence   of  3D  views  shown   one   after   another   to   give   an   animation.  A   replay  is  therefore   like   a   short   length  of  film,  with  each  frame  of  the  film  being  a  snapshot  of  a  model  as  it  was  at  a  given  time.   There   are   various   controls   and   parameters   that   allow   you   to   control   a   replay.   You   can   also   view   a   series   of   snapshots  all  superimposed  onto  a  single  view  Ȃ  see  Superimpose  Times.   There  are  two  types  of  replay:   x

Active   Simulation   Replays   show   the   model   as   it   was   at   regularly   spaced   times   during   the   currently   active   simulation.   This   type   of   replay   is   therefore   only   available   when   a   simulation   is   active   and   can   only   cover   the   period  that  has  already  been  simulated.  If  you  have  a  time  history  graph  window  open  when  the  replay  is  run,   then  the  replay  time  is  indicated  on  the  graph.  

x

Custom   Replays   are   replays   where   you   have   complete   control   over   frames   which   make   up   the   replay.   This   means  that,  for   example,   you  are  not  restricted  to   regularly  spaced  times;  you  can  have  frames  from  different   simulation   files   in   the   same   replay;   you   can   include   frames   showing   the   static   configuration   of  a   model;   you   are   able  to  vary  the  view  size,  view  angles  and  view  centre  to  achieve  panning,  rotating  and  zooming  effects.  Custom   replays   were   originally   introduced   to   help   visualise   series   of   static   snapshots,   for   example   during   a   lowering   operation.   However,   the   facility   is   very   powerful   and   you   are   certainly   not   restricted   to   this   application.   See   Custom  Replays  for  details.  

Export  Video  

Replays  can  be  exported  as  a  video  clip  in  AVI  file  format,  using  the  Export  Video  button  on  the  replay  parameters   form.   An   AVI   file   is   generated   containing   the   replay   using   the   most  recently   selected   3D   view  window  and   using   the   same  period,  frame  interval  and  speed  as  the  replay.   When   you   export   a   video   clip   you   will   be   asked   to   select   a   file   name   for   the   video   using   the   standard   Save   File   window.  At  the  bottom  of  this  window  is  a  checkbox  titled   Include  frame  details  in  video.  If  this  is  selected  then   each   frame   in   the   video   has   details   of   that   frame   (e.g.   simulation   time)   written   in   the   top   left   hand   corner   of   the   frame.  There  is  also  a  button  which  provides  a  link  to  the  Video  preferences.   AVI  is  a  standard  video  format,  so  the  file  can  then  be  imported  into  other  applications,  for  example  to  be  shown  in  a   presentation.   The   compression   method   (the   codec)   used   for   the   generating   the   video   file   can   be   set   on   the   Preferences  form.   Note:  

AVI  files  can  be  very  large  if  the  window  size  is  large  or  there  are  a  lot  of  frames  in  the  replay.  Also,   resizing   video   clips   (after   pasting   into   your   presentation)   will   introduce   aliasing   (digitisation   errors),  so  it  is  often  best  to  set  the  3D  View  window  size  to  the  required  size   before  you  export  the   video.  

59  

w

 

User  Interface,  Replays    

3.7.1

Replay  Parameters  

  The   replay   can   be   controlled   by   the   following   parameters   that   can   be   set   in   the   Replay   Parameters   form,   accessed  using  the  Replay  Parameters  button.   Replay  Period  

The   part   of   the   simulation   that   the   replay  covers.   You   can   select   to   replay  the   whole   simulation,   just   one   simulation   stage  (an  asterisk  *  denotes  an  incomplete  stage),  the  latest  wave   period  or   else  a  user  specified  period.  If  you  select   User   Specified   then   you   can   enter   your   own   Start   and   End   Times   for   the   replay   period.   These   can   be   set   to   '~'   which  is  interpreted  as  simulation  start  time  and  simulation  finish  time  respectively.   Interval  

The   simulation   time   step   size   between   frames   of   the   replay.   The   value   '~'   is   interpreted   as   the   actual   sample   interval,  i.e.  the  smallest  possible  interval.   Using   shorter   intervals   means   that   you   see   a   smoother   animation   (though   the   extra   drawing   required   may   slow   the   animation).   Example:   For   a   simulation   with   stages   of   8   seconds   each,   selecting   stage   2   and   a   replay   time   step   of   0.5   seconds   causes  the  replay  to  show  16  frames,  corresponding  to  t=8.0,  8.5,  9.0  ...  15.5.   Target  Speed  

Determines  how  fast  the  replay  is  played.   It  is  specified  as  a  percentage  of  real  time,  so  100%  means  at  real  time,   200%  means  twice  as  fast  etc.  As  a  special  case,  the  fastest  allowable  target  speed  (10000%  at  the  moment)  is  taken   to  mean  "as  fast  as  possible".   Note:  

The   specified   target   speed   is   not   always   achievable   because   the   computer   may   not   be   able   to   draw   each   frame   quickly   enough.   When   this   happens,   the   replay   will   be   played   as   fast   as   possible.   Replays   may   be  slow   if  you   specify   thick   lines   (line   width>1)   for   objects   in   the   model,   since   t his   can   increase  the  drawing  time.  

Continuous  

Continuous   means   replaying   like   an   endless   film   loop,   automatically   cycling   back   to   the   first   frame   after   the   last   frame  has  been  shown;  this  is  suitable  for  replays  of  whole  cycles  of  regular  cyclic  motion.  Non-­‐continuous  means   that  there  will  be  a  pause  at  the  end  of  the  replay,  before  it  starts  again  at  the  beginning;  this  is  more  suitable  for   non-­‐cyclic  motion.   All  Views  

If   this   is   selected,   then   the   replay   is   shown   in   all   3D   Views   simultaneously,   allowing   motion   to   be   viewed   from   several  different  viewpoints.  Otherwise  the  replay  is  played  in  the  currently  selected  view  window  only.   Show  Trails  

If  this  is  selected,  then  when  each  frame  of  the  replay  is  drawn  the  previous  frame  is  first  overdrawn  in  grey  Ȃ  this   results  in  grey  'trails'  showing  the  path  of  each  object.  

3.7.2

Replay  Control  

The   replay   can   be   controlled   from   the   Replay   menu,   by   using   toolbar   buttons   or   with   shortcut   keys.   In   addition,   some  replay  settings  can  only  be  modified  on  the  Replay  Parameters  form.   The  toolbar  has  a  section  dedicated  to  replay  control:     Figure:  

Replay  toolbar  controls  

The  replay  control  buttons,  menu  items  are  listed  in  the  table  below:   Button      

Menu  Item  

Action  

Replay  |  Start  Replay  

Shortcut   CTRL+R  

Replay  |  Stop  Replay  

CTRL+R  

Stop  replay  

60  

Start  replay  

w

 

User  Interface,  Replays  

 

Button  

Action  

Replay  |  Step  Replay  Forwards  

Shortcut   CTRL+A  

Replay  |  Step  Replay  Backwards  

CTRL+B  

Step  to  previous  frame  and  pause  

 

Replay  |  Replay  Faster  

CTRL+F  

Speed  up  replay  

 

Replay  |  Replay  Slower  

SHIFT+CTRL+F   Slow  down  replay  

Replay  |  Replay  Parameters  

CTRL+D  

  +  SHIFT  

 

Menu  Item  

Step  to  next  frame  and  pause  

Edit  replay  parameters  

Replay  Slider  Control  

  The   final   part   of   the   replay   toolbar   is   the   replay   slider.   This   allows   direct   control   of   the   replay   time.  Drag  the  slider  to  the  left  to  move  to  an  earlier  part  of  the  replay  and  to  the  right  to  move  to  a  later  part.  For   fine   grained   adjustment   of   replay   time   you   can   use   the   Replay   |   Step   Replay   Forwards   and   Replay   |   Step   Replay   Backwards  actions  or  alternatively  their  shortcuts,   CTRL+A  and   CTRL+B.  The  replay  time  is  displayed  on  and  can  be   controlled  from  Time  History  graphs.  

3.7.3

Custom  Replays  

Custom  replays  allow  you  to  piece  together  arbitrary  frames  from  different  OrcaFlex  files.  Each  frame  of  the  replay   can  be  either  the  static  configuration,  or  a  snapshot  of  a  specified  time  in  a  dynamic  simulation  file.   Using   frames   of   static   configurations   you   can   string   together   a   series   of   static   snapshots   giving,   for   example,   an   animation   of   an   installation   procedure.   Using   frames   from   dynamic   simulation   files   allows   you   to   create   replays   where   the   frames   are   from   one   or   more   simulations,   and,   if   you   wish,   vary   the   time   intervals   between   frames.   Frames  of   both  static  and   dynamic  configurations  can  be  included  in  the  same  custom  replay.  In  addition  you   are   able  to  vary  the  view  size,  view  angles  and  view  centre  to  achieve  panning,  rotating  and  zooming  effects.   To   use   the   custom   replay   feature   you   must   first   set   the   Replay   Type   data   item   on   the   Replay   Parameters   form   to   Custom  Replay.  Next  you  must  build  the  custom  replay   which  is   most  easily  done  using   the  Custom  Replay   Wizard,   which  can  be  opened  by  clicking  the  Custom  Replay  Wizard  button.   Replay  Specification  

This  is  the  file  containing  the  custom  replay  specification   Ȃ  that  is  the  file  that  is  saved  by  the  Custom  Replay  Wizard.   Custom  Replay  Parameters  

Custom  replays  also  make  use  of  some  of  the  parameters  needed  for  standard  simulation  replays.  These  parameters   are  Target  Speed,  Continuous,  All  Views  and  Show  Trails.  

3.7.4

Custom  Replay  Wizard  

The  Custom  Replay  Wizard  allows  you  to  define  a  series  of  replay  sections.  Each  replay  section  can  show  either:   1.

A  series  of  regularly  spaced  snapshots  from  a  simulation  file.  

2.

The  static  configuration  of  a  model  specified  by  either  a   data  file  or  a  simulation  file.  

Different  files  can  be  used  for  different  replay  sections.   Custom  Replay  Files   When  you  have  built  your  custom  replay  you  must  save  it  using  the  File  menu  or  save  button  on  the  toolbar.  Custom   replay  files  can  be  opened  in  a  similar  way.   We   recommend  that  you  save  your  custom  replay  file  before  you  start  setting  up  the  replay  sections.  This  is  because   once  you  have  saved  the  custom  replay  file  you  will  be  able  to  use  relative  paths  for  the  OrcaFlex  file  names.   Custom  Replay  Data   Custom  replay  specifies  view  parameters  (size,  position,  angles  and  graphics  mode)  

If   this   data   item   is   not   checked   then   the   replay   will   use   the   view   parameters   of   whichever   3D   View   window   it   appears   in.   In   this   mode   of   operation   you   will   be   able   manually   to   pan,   rotate   and   zoom   the   3D   View   using   the   normal  buttons  and  shortcuts.  

61  

w

 

User  Interface,  Replays    

If  this  data  item  is  checked  then  you  will  be  required  to  specify  the  view  parameters  (view  size,  view  centre,  view   azimuth,   view   elevation   and   graphics   mode)   for   each   replay   section.   This   allows   you   to   include   panning,   rotating   and  zooming  effects  in  your  replay.   While  learning  how  custom  replays  work  we  recommend  that  you  do  not  check  this  data  item.   Use  smoothed  panning,  rotating  and  zooming  effects  

This  item  is  only  available  if  the  "Custom  replay  specifies  view  parameters"  option  is  enabled.  If  you  are  panning,   rotating  and  zooming  during  replay  sections  then  the  transition  from  one  section  to  another  sometimes  appears  to   be  disjointed.  If  this  option  is  checked  then  the  transition  between  sections  i s  smoothed.   Frame  interval  in  real  time  

OrcaFlex   needs   to   know  how  fast   to   play   the   replay.   This   data   item   specifies   the   interval,   in   real   time,   between   each   replay   frame,   assuming   a   target   replay   speed   of   100%.   If   the   target   replay   speed   is,   say   200%,   then   the   interval   between  frames  will  be  half  this  value,  and  so  on.   Replay  Sections   You  can  specify  any  number  of  replay  sections.  For  each  replay  section  you  must  also  specify  the  following:   Replay  Section  Name  

This  is  a  descriptive  name  for  the  replay  section.  When  the  replay  is  running  OrcaFlex  displays  a  description  of  the   current  frame  in  the   message  box  on  the  status  bar  Ȃ  this  includes  the  replay  section  name.  This  description  can  also   be  included  in  exported  videos.   OrcaFlex  File  Name  

The  model  to  be  used  for  this  section  Ȃ  either  a  data  file  (.dat  or  .yml)  or  a  simulation  file  (.sim).   Dynamics  

This   setting   determines   whether   the   replay   section   defines   snapshots   from   a   dynamic   simulation   or   a   static   configuration.   If   the   file   is  a   data   file   then   the   replay  section   will   show   the   static   configuration   and   so   this   data   item   cannot  be  edited.   The   custom   replay   displays   static   configurations   for   a   data   file   by   loading   the   file   and   then   performing   the   static   calculation.  This  can  be  time  consuming  Ȃ  static  state  simulation  files  can  be  used  instead  to  avoid  the  overhead  of   performing  statics  each  time  the  replay  is  shown.   Simulation  Time  From,  Simulation  Time  To  

This   specifies   the   period   of   the   dynamic   simulation   covered   by   the   replay   section.   These   are   OrcaFlex   simulation   times  for  the  specified  simulation  file  of  this  replay  section.   If  the  replay  section  is  a  static  snapshot  then  these  data  items  are  not  editable.   Number  of  Frames  

This  is  the  total  number  of  frames  in  the  replay  section.  If  your  custom  replay  is  a  series  of  static  snapshots  then  you   would  usually  set  this  value  to  1.   Included  in  Replay  

This  allows   you   to   exclude   certain   sections   from   the  replay.  This  may  be   useful   while   developing  the   custom   replay   because  it  allows  you  to  concentrate  on  particular  replay  sections.   PowerPoint  slide  number  

Custom   replays   can   be   used   to   control   PowerPoint   slideshows.   To   make   use   of   this   you   need   to   be   showing   a   PowerPoint  slideshow  while  the  custom  replay  is  running.  At  the  start  of  each  replay  section  OrcaFlex  will  change   the  PowerPoint  slide  to  the  slide  number  specified  here.   If  you  do  not  wish  to  use  this  feature  you  should  leave  this  data  item  at  its  default  value  of  '~'.   View  Parameter  data   The  following  data  items  are  only  available  when  the  specifies  view  parameters  option  is  checked.  

62  

w

 

User  Interface,  Data  Forms  

 

From  View  Parameters,  To  View  Parameters  

The  view  size,  view  centre,  view  azimuth  and  view  elevation  for  the  first  and  last  frames  of  the  replay  section.  These   view  parameters  are  varied  between  these  values  for  the  other  frames  in  the  replay  section.   Hint:  

These  values  can  be  copied  from  OrcaFlex's  View  Parameters  form  using  the  clipboard.  

Graphics  Mode  

Specifies  either  the  Wire  frame  or  Shaded  graphics  mode  for  the  replay  section.  

3.7.5

Superimpose  Times  

Allows   model   configurations   for   different   times   of   the   simulation   to   be   superimposed   in   3D   Views.   Use   View  |  Current  Position  to  return  to  the  normal  view.   The  data  items  are:   List  of  Times  

The  simulation  times  which  will  be  superimposed.   All  Views  

If  this  box  is  checked  then  the  superimposed  view  is  drawn  in  all  3D  View  windows.  If  not  then  it  is  drawn  in  the   selected  3D  View.  

3.8

DATA  FORMS  

Each  object   in   the   model   has   data   items   that   define   its  properties.   The   data   are  examined   and   edited  in   the   object's   Data  Form,  which  can  be  accessed  by  various  methods:   x

use  the  Model  Browser  

x

DOUBLE  CLICK  the  object  in  a  3D  view  

x

RIGHT  CLICK  the  object  in  a  3D  view  and  use  the  pop-­‐up  menu.  

If  a  simulation  is  active  then  most  data  items  cannot  be  changed  since  they  affect  the  calculation,  but  you  can  change   things  like  the  object's  colour.   Control  Buttons   Ok  

Accepts  the  data  changes  made  and  then  closes  the  form.   Cancel  

Cancels  the  data  changes  made  and  then  closes  the  form.   Next  

Accepts   the   data   changes   made   and   then   displays   the   next   form   in   sequence.   Holding   the   SHIFT   key   down   while   CLICKING  the  Next  button  accepts  the  changes  and  then  displays  the  previous  data  form  in  sequence.  You  can  also   use  the  keyboard  shortcuts  ]]F6  for  next  and  SHIFT+F6  for  previous.   Pop-­‐up  Menu   The  pop-­‐up  menu  on  a  data  form  provides  various  facilities,  including:   x

The  data  form  can  be  printed,  copied  to  the  clipboard  or  exported  to  a  file.  The  data  for  the  whole  model  may  be   printed  using  the  File  |  Print  menu  item.  

x

Access  to  the  next  and  previous  data  form  and  to  the  Variable  Data  form.  

x

The  batch  script  names  for  the  currently-­‐selected  block  of  data  items.  

x

Data   forms   for   3D   Buoys,   6D   Buoys,   Vessels   and   Lines   provide   a   Connections   Report.   This   is   a   spreadsheet   listing  information   about   other   objects   connected   to   it.   Note   that   the   same   information,   but   for   all   objects   in   the   model,  can  be  displayed  using  the  Model  |  Show  Connections  Report  menu  item.  

63  

w

 

User  Interface,  Data  Forms    

x

On  data  forms  of  some  model  objects,  a  report  of  the  properties  of  that  object.  The  report  displays  properties   like   weight   in   air,  displacement,   weight   in   water   etc.   These   reports   are   currently  available   for   General   Data,  3D   Buoys,  6D  Buoys,  Vessels,  Lines,  Line  Types  and  Clump  Types.  

Calculator   A  simple  calculator  is  available  from  any  OrcaFlex  data  form.  It  can  be  opened  from  the  popup  menu  or  alternatively   by   pressing   F12.   Numbers   can   be   transferred   to   and   from   it   with   standard   Windows   copy   ( CTRL+C)   and   paste   (CTRL+V).   The   calculator   can   also   be   closed   by   pressing   F12   Ȃ   if   you   do   this   then   the   value   in   the   calculator   is   transferred  to  the  active  edit  cell.  

3.8.1

Data  Fields  

Data   items   on   each   Data   Form   are   displayed   in   Fields,   generally   with   related   fields   organised   into   Groups   or   Tables.   You  can  select  a  field  with  the  mouse,  or  use  the  keyboard  to  navigate  around  the  form.   TAB  moves  from  group  to   group,  and  the  arrow  keys  move  across  the  fields  in  a  group.   The  following  types  of  fields  are  used:   Text  

A  general  string  of  text,  used  for  example  for  titles  and  comments.   Name  

Each  object  is  given  a  name,  which  you  can  edit.  Object  names  must  be  unique  Ȃ  you  cannot  have  two  objects  with   the   same   name.   Certain   names   are   reserved   for   special   purposes:   Fixed,   Anchored   and   Free   (see   Connecting   Objects).   Numeric  

Numbers   can   be  entered   in  a   number   of   formats   such  as   3,   3.0,   0.3,   .3   or   3.0e6   or   3.0E6.   It   is  possible   to  enter  more   digits  than  those  shown  in  the  field,  but  beware  that  it  will  not  be  possible  to  see  them  again  without  editing  again   and  using  the  arrow  keys  to  examine  the  rest  of  field.   For   some   numeric   data   items   the   value   '~'   is   permitted.   For   example   this   is   sometimes   used   to   mean   'default   value'.   Details  are  given  in  the  descriptions  of  the  relevant  data  items.   Spin  Buttons  

These   are   small   buttons   with   up   and   down   arrows,   used   for   incrementing   and   decrementing   the   associated   field   (such   as   the   number   of   entries   in   a   table).   Using   the   mouse,   CLICK   on   the   upper   or   lower   parts   of   the   button   to   increment  or  decrement  the  associated  counter.   Multi-­‐choice  Buttons    

These  are  used  when  a  number  of  options  are  available.  Activate  the  button  to  step  on  to  the  next  available  option.   Check  Boxes  

These  show  a  tick,  meaning  selected,  or  are  blank,  meaning  not  selected.  CLICK  or  press  RETURN  to  change.   Colour  Selection  

These   show  as   a   block   of   colour.   DOUBLE  CLICK   or   press   RETURN   to   open   the   Colour   Selection   dialogue   window.   The   desired  colour  may  now  be  selected.   List  Boxes  

These   show   the   current   selection,   such  as   the   name   of   another   object   that   this   object   is  connected   to.   DOUBLE  CLICK   or  press  RETURN  to  show  a  List  Box,  and  then  select  another  item  and  RETURN  to  accept  the  new  choice.  

3.8.2

Data  Form  Editing  

The  TAB,  SHIFT+TAB,  HOME,  END  and  ARROW  keys  and  the  mouse  can  be  used  to  navigate  around  the  Edit  Form.   Editing  mode  is  entered  by   DOUBLE  CLICKING  a  cell  with  the  mouse,  or  by  starting  to  type  alphanumeric  characters,   which  are  entered  into  the  field  as  they  are  typed.  The  characters  that  have  been  typed  can  be  edited  by  using  the   arrow  keys  to  move  around  (now  within  the  field)  and  the  BACKSPACE  and  DELETE  keys.   Editing  mode  is  ended,  and  the  new  value  takes  effect,  when  you  press   RETURN  or   select  another  field  or  button  on   the  form.  To  end  editing  mode  but  reject  the  edit  (and  so  keep  the  old  value)  press   ESC.  

64  

w

 

User  Interface,  Results  

 

Many   numeric   fields   have   limits   on   the   range   of   values   that   can   be   entered,   for   example   an   object's   mass   must   always  be  greater  than  zero.  Warnings  are  given  if  invalid  values  are  typed.   Input  can  also  be   from  the   Windows  clipboard.   CTRL+C  copies  the  selected   field  or  block  of  fields  to   the  clipboard   whilst   CTRL+V   pastes   from   the   clipboard   into   the   selected   field.   In   this   way   data   can   be   easily   transferred   to   and   from  Spreadsheets,  Word  Processors,  etc.   Mouse  Actions   CLICK  

Select  Field  

CLICK+DRAG,   SHIFT+CLICK  

Select  a  block  of  fields  

DOUBLE  CLICK  

Start  Edit  Mode  in  this  field  (please  also  see  Data  Fields)  

SECONDARY   BUTTON  CLICK  

Context  sensitive  pop-­‐up  menu  for  copying,  exporting  and  printing  the  form  and,  for  some   model  objects,  viewing  additional  properties  

Group  Movement   TAB   Next  Group   SHIFT+TAB   Previous  Group   ALT+...  

Move  to  the  group  with  this  letter  underlined  in  its  heading  

Field  Movement   ĸĹĻĺ  

Go  to  adjacent  row  or  column  

HOME  

Go  to  leftmost  column  

END  

Go  to  rightmost  column  

PAGE  UP  

Go  to  top  row  

PAGE  DOWN  

Go  to  bottom  row  

Table  Editing   INSERT,  DELETE  

Insert  or  delete  rows  

Start  Editing   0..9,  A..Z  

Edit  (replace)  

During  Editing   ĸĺ,  HOME,  END  

Move  within  field  

End  Editing   ESC  

Cancel  edit  

ĹĻ  

Accept  edit  and  move  to  previous/next  row  

RETURN  

Accept  edit  

Copy  /  Paste   CTRL+C  

Copy  selected  field/block  to  clipboard  

CTRL+V  

Paste  from  clipboard  into  selected  field  

CTRL+D  

Fill  selection  from  top  (copy  top  cell  down)  

CTRL+R  

Fill  selection  from  left  (copy  leftmost  cell  to  right)  

CTRL+U   SHIFT+CTRL+D  

Fill  selection  from  bottom  (copy  bottom  cell  up)  

CTRL+L   SHIFT+CTRL+R  

Fill  selection  from  right  (copy  rightmost  cell  to  left)  

3.9

RESULTS  

3.9.1

Producing  Results  

You   can   access   results   by   either   CLICKING   on   the   Results   button   menu  item;  the  Select  Results  form  then  appears.  

65  

  on   the   toolbar   or   by   using   the   Select   Results  

w

 

User  Interface,  Results    

There   is   a   Keep   Open   switch   on   the   form's   context   menu,   which   allows   you   to   choose   whether   the   form   automatically  closes  when  you  select  a  result,  or  alternatively  stays  open  (and  on  top)  until  you  explicitly  close  it.   Graphs   and   Tables   can   be   sent   straight   to   the   printer   by   CLICKING   the   Print   button.   If   the   values   of   a   graph   are   required  in  text  form  then  CLICK  the  Values  button  Ȃ  this  give  the  values  in  a  Spreadsheet  window,  which  can  handle   multiple  variables  if  desired.   The  Select  Results  form  allows  you  to  select  the  results  you  want  by  specifying:   Result  Type  

This  option  allows  you  to  select  which  of  the  various  types  of  results  output  you  require.  Results  are  available  as  text   tables   (summary   results,   full   results,   offset   tables,   statistics,   linked   statistics,   extreme   statistics   or   line   clashing   reports)  or  as  graphs  (time   histories,  range  graphs,  XY  graphs,  offset  graphs   or  spectral  response  graphs).  The  types   of  results  available  depend  on  the  current   model  state.   Object  

The   object   for   which   you   want   results   (selected   in   the   same   way   as   in   the   Model   Browser)   and   for   some   objects   which  point  in  the  object.   x

For  the  Environment  you  must  specify  the  global  X,Y,Z  coordinates  of  the  point  for  which  you  want  results.  

x

For  6D  Buoys  that  have  wings  attached,  results  for  the  buoy  and  for  each  wing  are  available  separately.  

x

For  6D  Buoys  and  Vessels  the  position,  velocity  and  acceleration  results  are  reported  at  a  user  specified  point  on   the  object.  This  point  is  specified  in  object  local  coordinates.  

x

For  lines  you  must  specify  the  arc  length  along  the  line  Ȃ  see  Line  Results.  

Period  

For  time  histories,  XY  graphs  and  range  graphs  you  must  specify  the  period  of  the   simulation  to  be  included.  This   can   be   one   of  the   stages   of  the   simulation,   the  Whole   Simulation,   Specified   Period   or   Latest   Wave   (only   available   if  the   wave  is  regular).  The  Specified  Period  values  can  be  set  to  '~'  which  is  interpreted  as  simulation  start  time   and   simulation  finish  time  respectively.   For   Range   Graphs   the   period   can   also   be   Static   State   or   Instantaneous   Value.   The   Static   State   period   is   only   available   after   a   statics   calculation   and   the   graph   shows   a   curve   of   the   values   in   the   static   configuration.   The   Instantaneous   Value   period   is   available   when   a   simulation   has   been   run.   It   shows   a   curve   of   the   values   at   the   instantaneous   simulation   time.   This   is   normally   the   latest   simulated   time.   However,   if   a   replay   is   active   then   the   graph   shows   a   curve   of   values   at   the   active   replay   time.   This   allows   you   to   see,   for   an   entire   line,   how   a   results   variable  evolves  over  a  simulation.   Variable  

The  desired  variable(s).   Definitions  of  the  results  variables  can  be  obtained  by  selecting  them  in  the  Variable  list  box  and  then  pressing   F1.   Logging  for  results   The  summary  and  full  results  are  taken  directly  from  the  current  state  of  the  model.  All  the  other  results  are  derived   from   the   simulation   log   file   which   OrcaFlex   creates   automatically   when   a   simulation   is   run.   As   the   simulation   progresses,   OrcaFlex   samples   the   variables   for   each  object   at   regular   intervals   and   stores   the   sampled   values   in   the   log  file.  All  time  histories,  statistics  and  range  graphs  are  derived  from  the  simulation  log  file.   You  can  control  the  time  resolution  of  the  results  by  setting  the  Target  Sample  Interval  data  item  on  the  general  data   form.   This   must   be   done   before   the   simulation   is   started.   Decreasing   the   sample   interval   will   improve   the   time   resolution  of  the  results  (and  increase  the  number  of  samples  taken).  However,  because   more  samples  are  taken  this   will  also  increase  the  size  of  the  simulation  file  that  is  created.   Spike  Logging   A   special   algorithm   is   used   for   logging   results   that   tend   to   vary   rapidly   to   ensure   that   any   spikes   that   may   occur   between  samples  are  recorded.  We  refer  to  this  algorithm  as  spike  logging.   Line  Results  

OrcaFlex   spike   logs   Effective   Tension,   Torque,   Clash   Force,   Clash   Energy,   Solid   Contact   Force,   End   Force   results   a nd   Vortex  Force  results.  In  addition  other  results  which  are  derived  from  these  quantities  are  effectively  spike  logged  

66  

w

 

User  Interface,  Results  

 

by  association.  Such  variables  include  Wall  Tension,  Normalised  Tension,  Direct  Tensile  Strain,  ZZ  Strain,  Worst  ZZ   Strain,  Direct  Tensile  Stress,  von  Mises  Stress,  Max  von  Mises  Stress  and  ZZ  Stress.   Link  and  Winch  Results  

OrcaFlex  spike  logs  Tension  and  Velocity.   Solid  Results  

OrcaFlex  spike  logs  contact  force  magnitude.   General  Results  

OrcaFlex  spike  logs  Implicit  solver  iteration  count  and  Implicit  solver  time  step.   Inadequate  segmentation  warning   If  any  lines  have,  during  the  simulation,  gone  into  greater  compression  than  their  segment  Euler  load  then  a  warning   note   is   added   to   the   Results   form.   Such   lines   are   marked   with   the   symbol   §   in   the   Model   Browser.   Usually   this   means  that  finer  segmentation  is  needed  in  some  sections  of  these  lines  in  order  to  model  compression  adequately.   Offset  warning   If  any  of  the  multiple  statics  calculations  have  failed  then  a  warning  note  is  added  to  the  Results  form.  

3.9.2

Selecting  Variables  

Each  object  has  associated  with  it:   x

A  currently  selected  variable  that  will  be  used  for  graphs.  

x

A  set  of  statistics  variables  that  will  be  included  in  statistics  reports.  

For  the  currently  selected  object,  the  currently  selected  variables  are  shown  in  a  list  on  the  results  selection  form.   If  Statistics  results   are   selected,   then   the   list  shows  the  set   of   variables   that  will   be  included   in   the   statistics   report   and  you  can  add  or  remove  variables  by  CLICKING  on  them  in  the  list.   If  a  Time  History  is  selected,  the  list  shows  the  (single)  currently  selected  variable  and  you  can  select  a  different   variable  by  CLICKING  on  it  in  the  list.   You  can  also  multi-­‐select  variables,  using:   CLICK  

select  one  variable  

DRAG  

select  a  range  of  variables  

SHIFT+CLICK   select  a  range  of  variables   CTRL+CLICK  

add  /  remove  one  variable  

CTRL+DRAG  

add  /  remove  range  of  variables  

If   more   than   one   variable   is   selected,   then   the   Values   button   will   give   a   single   Spreadsheet   Window   with   a   time   history   column   for   each   selected   variable,   and   the   Graph   button   will   give   a   separate   Graph   Window   for   each   variable.   New  columns  can  be  appended  to  existing  time  history  spreadsheet  windows,  as  follows:   x

Select  the  spreadsheet  window  to  which  you  want  to  append,  by  clicking  on  it.  

x

Then  open  the  Select  Results  form  and  select  the  variables  that  you  want  to  append.  

x

Then  hold  the  CTRL  button  down  and  click  the  Values  button.  

x

Provided   that   the   selected   spreadsheet   window  is   a   time  history   values   table   and   that   the   time   periods   for   both   sets  of  histories  match,  then  the  new  time  histories  will  be  appended  to  the  active  window.  This  allows  you  to   have  a  single  window  containing  results  from  different  objects.  

3.9.3

Summary  and  Full  Results  

These   spreadsheet   windows   give   the   current   state   of   an   object   or   of   the   whole   model.   For   example,   in   Statics  Complete  state  the  full  results  tables  show  the  positions  of  objects  in  their  static  position.  If  a  simulation  is   active,  then  they  show  the  positions  of  objects  at  the  latest  time  calculated.   To  obtain  one  of  these  results  tables:  

67  

w

 

User  Interface,  Results    

x

Select  Summary  Results  or  Full  Results  on  the  Results  form.  

x

Select  the  object  required.  

x

Click  the  Table  button.  

The  summary  results  are  simply  an  abbreviated  form  of  the  full  results,  in  which  the  results  for  lines  only  include   the  end  nodes,  not  all  of  the  intermediate  nodes.   When  the  model  is  in  Statics  Complete  state  the  summary  and  full  results  include  estimates  of  the  shortest  natural   periods   of   objects   or   of   the   whole   model.   These   can   be   used   to   determine   suitable   simulation   time   steps.   The   simulation  inner  time  step  should  normally  be  no  more  than  1/10th  of  the  shortest   natural  period  of  the  model   Ȃ   this   is  given   at   the   top   of   the   summary  results   or   full   results   report   for  All   Objects.  In   addition   the   full   results   table   for  a  line  contains  detailed  reports  of  the  shortest  natural  periods.  

3.9.4

Statistics  

The  Statistics  report  provides,  for  each  statistics  variable:   x

The  minimum  and  maximum  values  and  the  simulation  times  when  they  occurred.  

x

The  mean  and  standard  deviation  (i.e.  the  root  mean  square  about  the  mean).  

These   statistics   are   reported   for   each   of   a   number   of   periods   of   the   simulation.   If   Statistics   by   Wave   Period   is   selected  then  these  periods  are  successive  wave  periods;  otherwise  they  are  the  stages  of  the  simulation.   To  obtain  a  Statistics  report:   x

Select  Statistics.  

x

Select  the  object  and  the  variables  of  interest  (see  Selecting  Variables).  

x

CLICK  the  Table  button.  

The  report  is  presented  in  a  spreadsheet.   Note:  

3.9.5

Be  careful  when  interpreting   statistics   of  Line   Clearance  and  Seabed   Clearance,   since  these   results   are  already  minima   Ȃ  the  shortest  distance  to  any  other  line  and  to  any  point  on  the  seabed.  For   example,   the   maximum   of   Line   Contact   Clearance   will   be   the   maximum   value   that   the   smallest   clearance  took  during  the  period.  

Linked  Statistics  

The  Linked  Statistics   table   relates  a  group  of  variables   for  a  given  object.  For  a  specified  group  of  variables  and  a   specified   period   of   simulation,   OrcaFlex   finds   the   minimum   and   maximum   of   each   variable   and   reports   these   extreme  values,  the  times  they  occurred  and  the  values  that  all  the  other  variables  took  at  those  times.   The  report  also  includes:   Ɋ  

mean,  

ɐ  

standard  deviation,  

Tz  

mean  up-­‐crossing  period,  estimated  as  the  average  time  between  successive  up-­‐…”‘••‹‰•‘ˆ–Š‡‡ƒ˜ƒŽ—‡Ɋǡ  

Tc  

mean  crest  period,  estimated  as  the  average  time  between  successive  local  maxima,  

m0  

zeroth  spectral  momentǡ‡•–‹ƒ–‡†ƒ•ɐ2,  

m2  

second  spectral  moment,  estimated  as  m0/Tz2,  

m4  

fourth  spectral  moment  estimated  as  m2/Tc2,  

ɂ  

spectral  bandwidth  parameter,  estimated  as  (1-­‐Tc2/Tz2)½,  

To  obtain  a  Linked  Statistics  report:  

x

Select  Linked  Statistics.  

x

Select  the  required  object  and  period.  

x

Select  the  variables  of  interest  (see  Selecting  Variables).  

x

CLICK  the  OK  button.  

The  report  is  presented  in  a  spreadsheet.  

68  

w

 

User  Interface,  Results  

 

Note:  

3.9.6

Be  careful  when  interpreting   statistics   of  Line   Clearance  and  Seabed   Clearance,   since  these   results   are  already  minima   Ȃ  the  shortest  distance  to  any  other  line  and  to  any  point  on  the  seabed.  For   example,   the   maximum   of   Line   Contact   Clearance   will   be   the   maximum   value   that   the   smallest   clearance  took  during  the  period.  

Offset  Tables  

These   Text   Windows   are   available   only   after   multiple   statics   calculations   and   only   for   vessels.   For   a   given   offset   direction  they  report  the  total  load  on  the  vessel  and  show  how  it  varies  with  offset  distance.  The  worst  tension  in   any  segment  of  any  line  connected  to  the  vessel  is  also  reported  for  each  offset.   To  obtain  an  Offset  Table:  

x

Select  Offset  Table  on  the  Results  form.  

x

Select  the  offset  vessel.  

x

Select  the  offset  direction  required.  

x

CLICK  the  Table  button.  

The  report  is  presented  in  a  spreadsheet.  

3.9.7

Line  Clashing  Report  

The  Line  Clashing  Report  produces  a  detailed  tabular  report  about  the  line  clashing  events  during  a  simulation.   To  obtain  a  Line  Clashing  Report:  

x

Select  Line  Clashing  Report  on  the  Results  form.  

x

Select  a  line.  

x

Select  the  period  required.  

x

CLICK  the  Table  button.  

The  report  is  presented  in  a  spreadsheet.   Contents  of  the  Line  Clashing  Report   The  report  lists  a  summary  table  followed  by  a  detailed  table  as  described  below.   Summary  table   The   summary  table   lists   all   clash  events   for   segments   on   the   selected   line.   A  clash   event  is  deemed  to   start  when   a   segment  from  the  selected  line  first  comes  into  contact  with  another  line  segment.   We  shall  refer  to  the  selected  line   as  L1  and  to  the   particular  segment  on   this  line  as  S1.  The  clash  event  ends  when  S1  is  no  longer   in  contact  with  any   other  line  segments.   Note:  

During  the  course  of  a  clash  event  the  segment  S1  may  be  in  contact  with  a  number  of  different  line   segments   from   other   lines,   e.g.   if   the   clash   is   a   sliding   contact.   This   is   counted   as   a   single   clash   event  from  the  perspective  of  S1.  

For  each  clash  event  the  following  results  are  reported:   Event  number  

A  number  of  clash  events  may  occur  during  the  simulation.  Each  event  is  given  a  number  to  identify  it.  This  is  useful   when  relating  the  summary  results  of  a  clash  event  to  the  detailed  results.   Segment  number  and  segment  arc  length  

This  identifies  the  segment  S1  on  the  selected  line.   Start  Time,  End  Time  and  Duration  

The  simulation  time  of  the  start  and  end  of  the  clash  event  together  with  its  duration.   Total  Impulse  

The  total  impulse  of  the  clash  event.  

69  

w

 

User  Interface,  Results    

Peak  Clash  Force  

A  scalar  value  reporting  the  greatest  value  of  clash  force  achieved  during  the  clash  event.   The  clash  force  vector  is   monitored  during  each  clash  event  and  the  greatest  magnitude  of  this  vector  is  reported.   Peak  Clash  Energy  

A  scalar  value  reporting  the  greatest  value  of  clash  energy  achieved  during  the  clash  event.   Max  Penetration  

At   each   time   step   we   calculate   the   depth   of   penetration   between   the   outer   surfaces   of   segment   S 1   and   all   other   segments.  Let  S2  be  a  segment  on  another  line.   Let   the   radii  of   the   two   segments   be   r1   and   r2   (as   defined   by   the   line   type  contact   diameter).   OrcaFlex   calculates   the   shortest  separation  distance,  d,  between  the  centrelines  of  the  two  segments.  The  penetration  of  these  two  segments   is  defined  to  be  (r1  +  r2)  Ȃ  d.  The  value  reported  as  Max  Penetration  is  the  maximum  value  of  penetration  between   segment  S1  and  any  other  segment  over  the  duration  of  the  clash  event.   Detailed  table   The  detailed  table  reports  information  about  each  individual  contact  between  segment  S 1  and  another  segment.  If   during  the  course  of  a  clash  event  segment  S1  is  in  contact  with  a  number  of  segments  on  other  lines  then  the  start   time,  end  time  and  duration  of  each  of  those  individual  contacts  is  reported.   Contact  velocity  

The   detailed   table   also   includes   the   contact   velocity   for   each   individual   contact.   This   is   defined   to   be   the   normal   relative  velocity  of  the  two  contact  points  at  the  instant  in  time  when  the  clash  event  started.  

3.9.8

Time  History  and  XY  Graphs  

Time   History   graphs   are   of   a   single   variable   against   time.   XY   graphs   are   of   one   time   dependent   variable   against   another.   The  period  of  simulation  covered  by  the  graph  is  chosen  from  a  list.   To  obtain  a  Time  History  or  XY  Graph:  

1.

Select  Time  History  or  XY  Graph  on  the  Results  form.  

2.

Select  the  object  required.  

3.

Select  the  variable  required  (see  Selecting  Variables).  More  than  one  variable  can  be  selected  for  time  histories.  

For   XY   graphs   the   steps   2   and   3   need   to  be   done   for   both   axes.   Do   this   by   CLICKING   on   one   of  the   options   labelled  X-­‐ axis  or  Y-­‐axis,  which  are  located  at  the  bottom  of  the  results  form,  and  then  repeating  steps  2  and  3.   x

Select  the  period  required.  

x

CLICK  the  Graph  button.  

Time   history   and   XY   graphs   are   displayed   in   Graph   Windows   and   they   are   "live"   Ȃ   i.e.   they   are   regularly   updated   during  the  simulation.  You  can  therefore  set  up  one  or  more  graph  windows  at  the  start  of  a  simulation  and  watch   the  graphs  develop  as  the  simulation  progresses.  If  you  reset  the  simulation  then  t he  curves  will  be  removed  but  the   graphs   will   remain,   so   you   can   adjust   the   model   and   re-­‐run   the   simulation   and   the   graphs   will   then   be   redrawn.   Graphs  are  automatically  deleted  if  the  object  that  they  refer  to  is  removed,  for  example  by  loading  a  new  model.   Range  Jump  Suppression  

For  time  histories  of  angles  OrcaFlex  chooses  the  angle's  range  so  that  the  time  history  is  continuous.   For  example  consider   vessel  heading,  which  is  normally  reported  in  the  range   -­‐180°  to  +180°.  If  the  vessel's  heading   passes  through  180°  then  without  range  jump  suppression  the  time  history  would  be:   ..,  179°,  180°,  -­‐179°,  ..   i.e.  with   a   360°   jump.   To   avoid   this   jump  OrcaFlex   adds   or   subtracts   multiples   of   360°  to   give   the   best   continuation   of  the  previous  value.  So  in  this  example  it  adds  360°  to  the  -­‐179°  value  and  hence  reports:   ..,  179°,  180°,  181°,  ..   This  addition  is  valid  since  181°  and  -­‐179°  are  of  course  identical  headings.  

70  

w

 

User  Interface,  Results  

 

Note  that  this  means  that  angle  time  history  results  can  go  outside  the  range  -­‐360°  to  +360°.   Spectral  Density  

From  any  time  history  graph  you  can  use  the  pop-­‐up  menu  to  obtain  the  spectral  density  graph  for  that  time  history.   The   curve   shown   on   the   graph   is   the   one-­‐sided   power   spectral   density   (PSD)   per   unit   time   of   the   sampled   time   history,  obtained  using  the  Fourier  Transform.   Notes:  

Using   the   Fourier   Transform   to   estimate   the   PSD   inevitably   introduces   'noise'   or   'leakage'   to   the   spectrum.   To   reduce   the   leakage   the   time   history   is   partitioned   into   a   number   of   overlapping   periods.  The  PSDs  are  calculated  for  each  period  and  then  averaged  to  give  the  reported  PSD  which   has  the  effect  of  smoothing  the  resulting  PSD.  

 

This  smoothing  technique  is  only  applied  if  there  is  more  than  200s  of  data  in  the  time  history.  

Empirical  Cumulative  Distribution  

From  any  time  history  graph  you  can  use  the  pop-­‐up  menu  to  obtain  the  empirical  cumulative  distribution  graph  for   that  time  history.  This  graph  shows  what  proportion  of  the  samples  in  the  time  history  are  less  than  or  equal  to  a   given  value.   These   graphs   are   sometimes   referred   to   as   Exceedence   Plots   since   they   can   sometimes   be   used   to   estimate   the   probability  that  the  variable  will  exceed  a  given  value.   Warning:  

The   samples   in   a   time   history   are   not   independent.   They   have   what   is   called   'serial   correlation',   which  often  affects  the  accuracy  of  statistical  results  based  on  them.  

Rainflow  half-­‐cycle  Empirical  Cumulative  Distribution  

From  any   time  history  graph  you  can  use   the   pop-­‐up  menu   to  obtain   the   rainflow   half-­‐cycle  empirical  cumulative   distribution  graph  for  that  time  history.  The  curve  on  this  graph  is  produced  in  the  following  way:   1.

The  time  history  is  analysed  using  the   rainflow  cycle-­‐counting  algorithm.  For  details  of  this  algorithm  see  the   paper  by  Rychlik.  

2.

The   rainflow   algorithm   produces   a   list   of  half-­‐cycles   associated   with   the   time   history.   The   empirical   cumulative   distribution  of  these  half-­‐cycles  is  then  plotted.  

3.9.9

Range  Graphs  

Range   graphs   are   only   available   for   a   selection   of   variables   and   they   are   only   available   for   Lines.   They   show   the   values   the   variable   took,   during   a   specified   part   of   the   simulation,   as   a   function   of   arc   length   along   the   Line.   In   particular:   x

Range  graphs  show  the  minimum,   mean  and  maximum  values  that  the  variable  took   during  the  specified  part  of   the  simulation  with  the  exception  that  the  Line  Clearance  range  graphs  only  show  the  minimum  value.  

x

Effective   tension   range   graphs   have   extra   curves   showing   the   segment   Euler   load   and   the   Allowable   Tension   value  (as  specified  on  the  Line  Types  data  form).  

x

Bend   Moment   range   graphs   have   an   extra   curve   showing   the   maximum   permitted   bend   moment   (EI  /  Minimum  Bend  Radius  specified  on  the  Line  Types  data  form).  

x

Curvature  range  graphs  have  an  extra  curve  showing  the  maximum  permitted  curvature  (the  reciprocal  of  the   Minimum  Bend  Radius  specified  on  the  Line  Types  data  form).  

x

Stress  range  graphs  show  the  Allowable  Stress  (as  specified  on  the  Line  Types  data  form).  

x

A   Standard   Deviation   curve   can   also   be   added   to   a   range   graph   Ȃ   to   do   this   edit   the   graph's   properties   (by   double   clicking   on   the   graph)   and   set   the   Standard   Deviation   curve's   visible   property   (by   default   the   curves   are   not   visible).   Two   curves   are   then   drawn,   at   Mean  ±  šɐǡ ™Š‡”‡ š ‹• ƒ —•‡” …Š‘•‡ ˜ƒŽ—‡ ƒ† ɐ ‹• –Š‡ •–ƒ†ƒ”† deviation.   The   standard   deviation   is  calculated   from   all   the   samples   that   lie   in  the   simulation   period   chosen   for   the  graph.   Warning:  

Be  careful  not  to  assume  that  95%  of  the  data  lie  in  the  interval  Mean  ±  ͸ɐǤŠ‹•…‘‘‰—‹†‡Ž‹‡ is  based  on  the  assumption  that  the  data  are  sampled  from  a  N ormal  (i.e.  Gaussian)  distribution.  

To  obtain  a  Range  Graph:  

x

Select  Range  Graph  on  the  Results  form.  

71  

w

 

User  Interface,  Results    

x

Select  the  object  required.  

x

Select  the  arc  lengths  required.  This  can  be  the  entire  line,  a  selected  arc  length  range,  or  a  selected  line  section.  

x

Select  the  variable  required  (see  Selecting  Variables).  

x

Select  the  period  required.  

x

CLICK  the  Graph  button.  

Range   graphs   are   displayed   in   Graph   Windows   and   they   are   "live"   Ȃ   i.e.   they   are   regularly   updated   during   the   simulation.  You  can  therefore  set  up  one  or  more  graph  windows  at  the  start  of  a  simulation  and  watch  the  graphs   develop  as  the  simulation  progresses.  If  you  reset  the  simulation  then  the  curves  will  be  removed  but  the  graphs  will   remain,   so   you   can   adjust   the   model   and   re-­‐run   the   simulation   and   the   graphs   will   then   be   redrawn.   Graphs   are   automatically  deleted  if  the  object  that  they  refer  to  is  removed,  for  example  by  loading  a  new  model.   Range  Jump  Suppression  

Just  as  it  does  for  Time  History  and  XY  Graphs,  OrcaFlex  applies  range  jump  suppression  for  range  graphs  of  angles.  

3.9.10

Offset  Graphs  

These   graphs   are   available   only  after   a  multiple   statics  calculation   has   been   done   and   only  for   the   offset   vessel.   The   following  variables  are  plotted  against  offset  distance:   Restoring  Force  

The  magnitude  of  the  horizontal  component  of  the  total  force  applied  to  the  vessel  by  the  attached  Lines  or  other   objects.  Note  that  this  force  is  not  necessarily  in  the  offset  direction.   Vertical  Force  

The  vertically  downwards  component  of  the  total  force  applied  to  the  vessel  by  the  attached  Lines  or  other  objects.   Yaw  Moment  

The  moment,  about  the  vertical,  applied  to  the  vessel  by  the  attached  Lines  or  other  objects.   Worst  Tension  

The  largest  tension  in  any  segment  of  any  Line  connected  to  the  vessel.   To  obtain  an  Offset  Graph:  

x

Select  Offset  Graph  on  the  Results  form.  

x

Select  the  offset  vessel.  

x

Select  the  offset  direction  required.  

x

Select  the  variable  required.  

x

CLICK  the  Graph  button.  

3.9.11

Spectral  Response  Graphs  

These  graphs  are  available  only  if  you  have  run  a   response  calculation  wave.  The   graph  is  only  available  once  the   simulation  has  been  completed.   The  graph  plots  the  calculated  RAO  for  the  selected  variable  on  the  Y  axis  and  wave  frequency  on  the  X  axis.   To  obtain  a  Spectral  Response  Graph:  

x

Select  Spectral  Response  Graph  on  the  Results  form.  

x

Select  the  object  required.  

x

Select  the  variable  required  (see  Selecting  Variables).  More  than  one  variable  can  be  selected.  

x

CLICK  the  Graph  button.  

3.9.12

Extreme  Statistics  Results  

There   is  often   a   requirement   to   predict   the   extreme   responses   of   a   system,   for  example   to   determine   the   likelihood   of  a  load  exceeding  a  critical  value  that  may  lead  to  failure.  Such  values  are  needed  when  using  standards  such  as   DNV-­‐OS-­‐F201  and  API  RP  2SK.  

72  

w

 

User  Interface,  Results  

 

OrcaFlex  can   estimate   extreme  values  for  any  given   result  variable   by  analysing  the  simulated  time  history   of   the   variable  using  extreme  value  statistical  methods.  You  may,  for  instance,  perform  a  mooring  analysis  in  an  irregular   sea-­‐state  and  then  estimate  the  maximum  mooring  line  tension  for  a  3-­‐hour  storm.   The   statistical   theory  for   this   estimation   is  well-­‐established   and   is   described   in   the  theory   section.   The   procedure   is   essentially  this:   x

You  select  the  statistical  distribution  to  be  used  to  model  the  distribution  of  extremes.  See  Data  below.  

x

OrcaFlex  estimates  the  distribution  model  parameters  that  best  fit  the  simulation  time  history  of  the  variable.  

x

OrcaFlex   uses   the   fitted   distribution   to   estimate   and   report   the   required   extreme   statistic   (e.g.   return   level),   for   a  specified  period  of  exposure.  See  Results  below.  

x

OrcaFlex  provides  diagnostic  graphs  that  you  should  use  to  judge  the  reliability  of  the  results.  

The  Extreme  Statistics  Results  form  is  designed  to  lead  you  through  this  process.   When  you  open  the  Extreme  Statistics  Results  form,  for  a  selected  results  variable,  you  will  come  first  to  the   Data   page,   where  you   will  select  the  distribution.  Moving  then  to   either  of  the   other   pages   (Results  or   Diagnostic  Graphs)   will   cause   OrcaFlex   to   carry   out   the   estimation   part   of   the   procedure.   The   Diagnostic   Graphs   assist   in   testing   the   model.  The  Results  page  reports  the  estimated  statistics,  e.g.  the  return  value  for  the  specified  period,  the  estimation   uncertainty  inherent  in  that  value  etc.   Data   For  convenience,  the  time  history  result  graph  is  reproduced  on  the  Data  page.  The  data  required  for  the  fitting  of   the  model  are  entered  on  this  page,  and  are  as  follows.   Distributions  

These   fall   into   two   groups,   according   to   the   statistical   method   with   which   they   are   applied.   For   details   see   the   Extreme  Statistics  Theory  section.   x

Rayleigh  distribution.  This  method  assumes  that  the  variable  is  a  stationary  Gaussian  process.  This  is  perhaps  a   reasonable   assumption   for   waves,   particularly   in   deep   water,   and   for   responses   which   are   approximately   linear   with  respect  to  wave  height.  However,  for  many  other  variables  of  interest,  the  Gaussian  assumption  is  invalid   and  leads  to  poor  estimates  of  extreme  values.  

x

Weibull  and  Generalised  Pareto  (GPD)  distributions.  These  distributions  are  both  fitted  using  the  maximum   likelihood   method.   Historically,   the   Weibull   distribution   has   often   been   used   for   marine   systems,   but   the   Generalised   Pareto   is   preferred   by   the   extreme   statistics   community   because   of   its   sound   mathematical   foundations.  

Extremes  to  analyse  

Specifies  whether  maxima  (upper  tail)  or  minima  (lower  tail)  are  to  be  analysed.   Threshold  and  Decluster  Period  

These  data  are  only  required  when  using  the  Weibull  and  GPD  distributions,  which  are  fitted  to  extremes  of  the  time   history  and  those  extremes  are  selected  using  the   peaks-­‐over-­‐threshold  method  with  (optional)  declustering.   The   threshold   controls   the   peaks-­‐over-­‐threshold   method.   This   allows   you   to   control   the   extent   to   which   the   analysis  is  based  on  only  the  extreme  values  in  the  data  (the  tail  of  the  distribution).   The  decluster   period  controls   the  declustering.   This  helps  avoid   or   reduce   any  statistical   dependence   between   the   extreme  data  values  used  in  the  analysis.  It  can  be  set  to  one  of  the  following:   x

Zero,  in  which  case  no  declustering  will  be  done,  and  all   values  above  the  specified  threshold  will  be  included.   This  is  generally  not  recommended  since  the  values  are  unlikely  to  be  independent.  

x

A   positive   value.   In   this  case   OrcaFlex  will   break   the   sequence   of   time   history   values   into   clusters  of   successive   values   that   stay  above   the   threshold.   It  will   then  decluster  by  merging   successive   clusters   that   are   separated  by   periods   (during   which   the   variable   is   less   than   the   threshold)   that   last   no   longer   than   the   specified   decluster   period.  The  most  extreme  value  of  each  of  the  resulting  merged  clusters  will  then  be  included  in  the  analysis.  

x

'~'.   This   special   value   may   be   used   to   tell   OrcaFlex   to   take   the   clusters   to   be   the   groups   of   values   between   successive  up-­‐crossings  of  the  mean  value  (or  down-­‐crossings  if  analysing  lower  tail).  The  most  extreme  value   of  each  such  cluster  will  then  be  included  in  the  analysis,  but  ignoring  any  that  do  not  exceed  the  threshold.  

73  

w

 

User  Interface,  Results    

The  threshold  is  drawn  on  the  time  history  graph,  to  help  visualise  its  value  relative   to  the   extremes  o f  the  data.  The   number   of   data   points   that   will   be   included   in   the   analysis  (after   the   threshold   and  declustering   have  been   done)   is   also  displayed.  This  helps  with  setting  the  threshold  and  decluster  period.   The   best   value   for   the   threshold   is   one   that   strikes   a   balance   between   a   not-­‐extreme-­‐enough   value   (which   will   increase  the  number  of  data  points  fitted  but  may  give  biased  fitting  by  allowing  less  extreme  values  to  influence  the   fitting  too   much),  and  a  too-­‐extreme  value  (which   will   fit  to  only  the  more   relevant  extreme   data  points,   but   may   give  very  wide  confidence  intervals  if  there  are  too  few  such  extremes  in  the  data).   Results   The  following  data  items,  found  on  the  Results  page,  do   not  affect  the  fitting  of  the  statistical  model.  Rather,  they   are   applied  to  the  fitted  model  to  obtain  the  reported  results.   Rayleigh  

Storm  duration  is  the   return   period   for   which  the   return   level   is  reported.   The   length  of   the   simulation,  relative   to   this  duration,  will  determine  the  accuracy  of  the  estimate  for  the  return  level.   Risk   factor   is   the   probability   of   exceeding   (or   falling   below,   for   lower   tail)   the   estimated   extreme   value.   For   example,  you  may  ask  for  the  3-­‐hour  extreme  value  that  is  exceeded  with  a  probability  of  0.01  (i.e.  a  risk  factor  of   1%).   Weibull  and  GPD  

Storm  duration  is  defined  as  for  the  Rayleigh  distribution.   The   maximum   likelihood   fitting   procedure   used   for   these   distributions   allows   the   estimation   of   a   confidence   interval  for  the  return  level,  for  a  specified   confidence  level.  OrcaFlex  reports  this  estimated  confidence  interval  in   addition  to  the  estimated  return  level.   The  reported  return  level  is  defined  to  be  the  level  whose  expected  number  of  exceedences  in  the  specified  storm   duration  is  one.  The  fitted  values  of  the  model  parameters  and  corresponding  standard  errors  are  also  reported.   Note:  

For  some  values  of  storm  duration  (usually  small  values)  it  might  not  be  possible  to  calculate  the   return   level.   This   is   indicated   by   the   value   'N/A'   (meaning   'not   available').   Similarly,   for   some   combinations   of   storm   duration   and   confidence   level,   the   calculation   may   fail   to   determine   the   confidence  limits,  and  again  these  are  then  denoted  by  'N/A'.  

Diagnostic  Graphs   The  diagnostic  graphs   will  help  you  to  assess  the  goodness-­‐of-­‐fit  of  the  model,  and  how  appropriate  or  not  the  fitted   distribution  is.  They  should  be  interpreted  together,  not  in  isolation,  as  follows.   x

The   Quantile   Plot   displays   quantiles   of   the   empirical   data   plotted   against   model   quantiles.   If   the   model   is   a   good  fit,  then  the  points  should  lie  close  to  the  superimposed  45°  diagonal  line,  and  any  significant  departure   from   this   (especially   a   systematic   one,   for   example   an   obvious   trend   away   from   the   diagonal)   indicates   poor   model   fit.   The   vertical   lines,   drawn   through   each   point,   are   pointwise   95%   tolerance   intervals   and   may   be   used   as   a   guide   to   deciding   whether   any   departure   from   the   diagonal   is   significant.   If   all   the   vertical   lines   intersect  the  diagonal  line,  then   the   modelled  values  are  probably  sufficiently  close  to  the   empirical  value  not  to   be  of  concern.  If,  however,  a   number  of  the  vertical  lines   fail  to  reach  the   diagonal,  then  that  may  raise  concerns   about  the  validity  of  the  fitted  model.  

x

The   Return   Level   Plot   shows   return   level   against   return   period   (i.e.   storm   duration),   with   the   latter   on   a   logarithmic   scale   to   highlight   the   effect   of   extrapolation.   The   central   line   on   the   graph   is   the   return   level   for   the   fitted   model,   and   the   pair  of   outer   lines   the   corresponding   pointwise   95%  confidence   limits.  The  points   are   the   empirical  return  levels,  based  upon  the  data,  and  should  lie  between  the  confidence  limits  if  the  model  fits  the   data   well.   As   with   the   quantile   plot,   a   significant   number   of   points   contravening   these   limits   indicates   poor   model  fit.  Again,  OrcaFlex  may  sometimes   be  unable  to  determine  the  confidence  limits  for  some  return   periods   Ȃ  this  may  result  in  gaps  in  the  confidence  limit  lines,  or  even  in  their  not  appearing  at  all.  

An  example  of  diagnostics  graphs  indicating  a  good  model  fit  is  shown  below:  

74  

w

 

User  Interface,  Results  

 

  Figure:  

Diagnostics  graphs  for  a  good  model  fit  

If  either  of  these  graphs  indicates  a  poor  model  fit,  then  you  should  reconsider  the  entries  on  the  data  page:   x

Distribution.   The   distribution   may   be   inappropriate   Ȃ   the   data   may   simply   not   conform   to   the   selected   distribution.  

x

Threshold.   The   threshold   may   be   too   low,   hence   including   too   many   points   which   are   not   in   the   tail   of   the   distribution;  or  too  high,  resulting  in  too  few  data  points  for  the  analysis  and  consequent  large  variation  in  the   results.  

x

Decluster  period.  This  may  be  too  long  (so  too  few  data  points),  or  too  short  (so  successive  data  points  might   not  be  independent).  

3.9.13

Presenting  OrcaFlex  Results  

OrcaFlex   users   often   wish   to   show   their   OrcaFlex   results   in   a   slide   presentation   prepared   using   a   presentation   program  such  as  Microsoft  PowerPoint.  Here  are  some  tips  on  how  this  can  be  done.   Graphs  

Graphs  can  be  transferred  from  OrcaFlex  to  presentation  programs  by  simple  copy  +  paste.   Note:  

In   PowerPoint,   instead   of   using   Paste,   it   is   better   to   use   Paste   Special   (from   the   Edit   menu)   and   then  select  the  Enhanced  Metafile.  This  gives  better  resolution.  

Replays  

Replays  can  be  transferred  by  exporting  to  an  AVI  file  and  then  importing  that  video  clip  file  into  the  presentation   program.   An  XVID  encoded  AVI  file  (and  possibly  other  codecs)  added  to  Microsoft  PowerPoint  slides  as  a  Movie  Object  may   not  play  correctly  (displaying  a  blank  screen  on  replay,  or  the  video  only  appearing  in  full  screen  mode).  To  avoid   these  problems,  an  XVID  AVI  file  needs  to  be  inserted  as  a  Video  Clip  Object.  This  can  be  done  in  two  ways:   1.

Drag  and  Drop  the  AVI  file  onto  the  PowerPoint  slide,  or  

2.

From   the   PowerPoint   menu,   choose   Insert   |   Object.   Select   'Create   from   file'   and   Browse   to   your   file   (do   not   select  the  'Link'  option).  

To   set   options   such   as   auto   repeat,   right-­‐click   on   the   image   in   the   slide,   then   select   Video   Clip   Object   |   Open,   this   displays  the  video  player  window  and  menus.  

75  

w

 

User  Interface,  Graphs    

The  Video  Clip  Object  links  to  the  AVI   file  (it  is  not  embedded  within  PowerPoint)  so  the  file  location  needs  to  be   accessible  when  running  the  presentation.  The  computer  running  the  presentation  must  also  have  the  XVID  codec   installed.   Note:  

Resizing   video   clips   (after   pasting   into   your   presentation)   will   introduce   aliasing   (re-­‐digitisation   errors)  so   it   is   best   to   set   the   OrcaFlex   3D   View   window   to   the   required   size  before  you   export   the   video.  

Video  Clips  of  OrcaFlex  in  Use  

Your   presentation   can   even   show  video   clips   of   OrcaFlex   in   use,   illustrating   how  the   program   i s   used.   However,   it   is   rather   harder   to   generate   the   required   video   files.   We   recommend   using   software   called   Camtasia   (www.techsmith.com)  to  record  video  clips  showing  OrcaFlex  in  use.  

3.10

GRAPHS  

When   you   request   results   in   graphical   form,   they   are   presented   in   Graph   Windows.   You   can   open   several   simultaneous   graph   windows,   showing   different   results,   and   tile   them   on   the   screen   together   with   3D   Views   and   text  results  windows.  To  adjust  a  graph's  properties  (range  of  axes,  colours,  etc.)  see   Modifying  Graphs.   Graphs  have  a  pop-­‐up  menu  that  provides  the  following  facilities.   x

Use  Default  Ranges.  

x

Copy  copies  the  graph  to  the  clipboard,  from  where  you  can  paste  it  into  other  applications.  

x

Values.  

x

Spectral  Density.  

x

Empirical  Cumulative  Distribution.  

x

Rainflow  half-­‐cycle  Empirical  Cumulative  Distribution.  

x

Export  enables  you  to  export  the  graph  to  a  metafile  or  bitmap  file.  

x

Print  facilities  and  the  Monochrome  Output  preference.  

x

Properties.  

Graphs   of  simulation   results   are   updated   automatically   as   the   simulation   progresses.   Also,   they   are   kept   even   if  you   reset   the   simulation,   so   once   you   have   set   up   a   set   of   interesting   graphs   you   can   edit   the   model   and   re-­‐run   the   simulation  to  see  the  effect  of  changing  the  model.   You   can   also   set   up   results   graphs   when   in   reset   state,   prior   to   running   a   simulation   Ȃ   the   graph   will   be   empty   initially   and   will   grow   as   the   simulation   progresses.   Note   that   we   do   not   recommend   this   for   graphs   of   line   clearance,  however,  since  updating  them  can  significantly  slow  down  the  simulation.   The  workspace  feature  provides  a  very  powerful  way  of  managing  collections  of  related  graphs.   When  a  replay  is  in  progress  the  replay  time  is  indicated  on  both   Time  History  and  XY  graphs.  

76  

w

 

User  Interface,  Graphs  

 

  Figure:  

Replay   time   indicator   on   a   Time   History   Graph   (vertical   line   at   Time=16s)   and   on   an   XY   Graph  (grey  cross  in  bottom  right  of  the  graph).  

The   replay  time   indicator   on   a   Time  History  graph   can  be   directly  manipulated   using   the  mouse.  With   the   CTRL  key   pressed   you   simply  click   on  a   Time  History  graph   and   the   indicator  moves   to   where  you   have   clicked.   Any  open  3D   Views  are  updated  to  show  the  new  replay  time.  Similarly,  with  the   CTRL  key  pressed  you  can  click  and  then  drag  the   indicator.   This   direct   manipulation   of   the   replay   time   indicator   is   designed   to   help   understand   and   visualise   how   your  model  is  behaving  at  key  moments  of  the  simulation.   Printing  Graphs  

To  print  a  graph,  use  the  File  |  Print  menu  item.  When  printing  to  a  monochrome  printer  you  will  get  the  best  results   by  setting  the  Monochrome  Output  preference  Ȃ  this  is  set  by  default  when  the  program  is  first  installed.   Copy  and  Paste  with  graphs  

You  can  also  copy  a  graph  to  the  clipboard  Ȃ  simply  select  the  graph  window  by   CLICKING  on  it  and  then  using  the   Edit  |  Copy  menu  item.  From  the  clipboard  you  can  then  paste  it  into  another  application,  for  instance  into  a  word   processor   document.   Graphs   can   also   be  exported   as   Windows   metafiles,   use   the  File   |   Export  menu  item.   Metafiles   can  be  imported  into  many  Windows  programs,  such  as  word  processors,  spreadsheets,  graphics  packages  etc.   Note:  

3.10.1

When  copying  a  graph  to  the  clipboard,  the  size  of  the  graph  window  you  copy  from  has  an  effect   on  how  the  text  label  fonts  appear  when  the  graph  is  pasted  into  another  application.  For  example,   if  you  are  copying  a  graph  to  a  Word  Processor  and  want  the  graph  to  be  full  page  size,  then  the   graph  window  should  be  made  large  on  screen  (e.g.  maximised).  If  you  want  a  number  of  graphs  on   one  page  of  a  document  then  the  graph  should  be  smaller  on  screen   Ȃ  try  tiling  or  cascading  the   windows   (see  the  Window   menu).   By   experimenting   with   various   differently   sized   graphs   it   should   be  possible  to  arrange  for  the  fonts  to  appear  as  you  wish.  

Modifying  Graphs  

You  can  zoom  into  a  graph  by  holding  down  the  ALT  key  and  dragging  a  box  around  the  area  that  you  want  the  graph   to   display.  When   you   release   the   mouse   button   the   region   selected   will   be  expanded   to   fill   the   graph.   If   you   want   to   reverse  this  process  then  right  click  the  mouse  and  choose  Use  Default  Ranges  from  the  pop-­‐up  menu.   You   can   also   change   the   appearance   of   a   graph   by  double   clicking   on   the   graph   or   by  selecting  Properties  from   the   graph's  pop-­‐up  menu.  A  form  is  then  shown  which  allows  you  to  change  various  aspects  of  the  graph,  as  follows:  

77  

w

 

User  Interface,  Spreadsheets    

Axes  

You  can  set  the   range,  the   tick  spacing  and  the   number   of  small  ticks.  The  Use  Default  Tick  Spacing  button  sets  the   tick  spacing  and  the  number  of  small  ticks  to  sensible  default  values  based  on  the  range.  This  is  useful  if  you  want  to   set  the  range  to  a  specific  value  and  want  the  tick  spacing  to  be  set  automatically.   Labels  

You  can  alter  the  text  and  fonts  of  the  axis  and  tick  labels.   Curves  

You  can  control  the  line  properties  and  visibility  for  each  curve  on  the  graph.   Legend  

The  legend  is  a  key  showing  which  curve  is  which.  It  only  appears  on  graphs  that  have  multiple  curves,  e.g.  range   graphs.   You   can   control   whether   the   legend   is   shown   and   if   so   where   and   using   what   font.   Note   that   the   legend   includes  all  the  curves,  even  if  some  of  them  may  not  be  visible  at  the  time.   Intercepts  

Intercepts   are   lines,   like   the  axes,   that   go   right   across   the   graph.   In   fact   the  X   and  Y   axes   themselves   are   considered   to  be  intercepts.   You  can  add  more  intercepts,  for   example  to  mark  things  like  stage  start  times,  and  you  can  control   their  position  and  style.   Save  As  Default  

Changes   to   a   graph's   properties  normally  only  apply  to   that   graph.   But  for   general   settings   (fonts  etc.)   you   can   also   click  the  Save  As  Default  button.  OrcaFlex  then  remembers  the  current  settings  for  use  with  future  graphs.  

3.11

SPREADSHEETS  

Some   numerical   results   (e.g.   obtained   with   the   Values   button   on   the   Results   form)   appear   in   an   Excel   compatible   spreadsheet.   The   spreadsheet   is   read-­‐only.   If   you   wish   to   modify   or   extend   it   you   must   first   save   it   as   described   below.   Printing,  Copying  and  Exporting  Spreadsheets  

To   print   the   spreadsheet  right  click   and   select  Print,   but   remember   that   OrcaFlex   time   histories   are   normally  quite   long  and  will  therefore  produce  many  pages.  If  necessary,  you  can  first  adjust  the  printer  setup  using   File  |  Printer   Setup.   You  can  also  easily  transfer  the  results  to  other  applications  by  either:   x

Copy  and  paste  via  the  Windows  clipboard.  Select  the  block  to  be  transferred  and  press  CTRL+C.  

x

Saving  to  file.  Choose  Export  from  the  popup  menu  to  save  as  Excel  format  (.xls),  comma  separated  values  (.csv)   or  as  tab  delimited  text  (.txt).  

3.12

TEXT  WINDOWS  

Simple  text  windows  are  used  for  some  reports  Ȃ  see  below.  To  print  a  text  window,  use  the  File  |  Print  menu.  You   can   also   copy   text   to   the   clipboard  Ȃ  simply  select   a   region   of   text   and   then   use   the  Edit  |  Copy  menu  item   (or   press   CTRL+C).   From   the   clipboard   you   can   then   paste   it   into   another   application,   for   instance   into   a   word   processor   document.   Alternatively,   you   can   export   the   text   to   a   file   by   using   the  File  |  Export  menu   item.   The   resulting   text  file   can  then  be  imported  into  your  word  processor.   Statics  Progress  Window   During   a   Statics  Calculation,   the   progress   of  the   calculation   is   shown   in   the   message   box   on   t he   status   bar.   However   the   messages   are   also   sent   to   a   text   window   that   is  normally   minimised.   This  window   may   be   viewed   by   clicking   on   the  message  box  during  statics,  or  by  selecting  the   Window  |  Statics  Progress  menu  item  if  you  wish  to  watch  the   process  more  closely.  Like  other  text  windows  it  may  be  printed,  copied  or  exported,  as  described  above.  

3.13

WORKSPACES  

It   is   common   to   have   many  windows   (3D   View,   graph   or   spreadsheet)   open   within   OrcaFlex.   The   workspace   facility   is  designed  to  help  manage  these  windows.  

78  

w

 

User  Interface,  Comparing  Data  

 

Workspace  files  

A  collection  of  view,  graph  or  spreadsheet  windows  can  be  saved  using  the  Workspace  |  Save  Workspace  menu  item.   This   creates   a   text   file   with   the   .wrk   file   extension   containing   a   specification   of   the   current   window   layout.   The   workspace  can  be  restored  at  any  time  with  the  Workspace  |  Open  Workspace  menu  item.  This  can  give  significant   time  savings  if  you  wish  to  look  at  a  number  of  different  results  windows  for  a  large  number  of   OrcaFlex  models.   Note   that   the   contents   of   the  windows   are   not   saved   to  the  workspace   file,   just   a   logical   description  of   the   window.   For  example,  suppose  you  saved  a  workspace  containing  a  graph  of  Effective  Tension  of  a  Line  called  Riser.  If  you   then  loaded  a  different  simulation  file  and  open  that  workspace  then  you  would  see  the  Effective  Tension  of  the  Line   called  Riser  in  the  new  simulation  file  and   not  the  simulation  filed  open  when  the  workspace  was  saved.  This  means   that  you  can  look  at  the  same  collection  of  results  for  any  number  of  simulation  files.   Workspace  files  are  not  limited  to  simulation  files  Ȃ  static  results  and  multiple  statics  results  can  also  be  saved.   Default  workspaces  

As   an   alternative   to  loading   a   workspace   by   using   the  Workspace   menu   items   you   can   associate   default   workspaces   with  either  individual  simulation  files  or  with  entire  directories.   x

If   you   define   a   default   workspace   for   a  simulation   file   then   the   workspace   is  restored   whenever   you   open   that   same  simulation  file.  

x

If   you   define   a   default   workspace   for   a   directory   then   the   workspace   is   restored   whenever   you   open   any   simulation  file  in  that  directory.  

Getting  the  most  out  of  workspaces  

We   recommend   that   you   save   your   workspace   files   in   the   same   directory   as   the   OrcaFlex   files.   If   you   do   so   then   the   workspace  file  will  appear  in  the  Most  Recent  Files  list  on  the  Workspace  menu.   Workspace  files  can  be  very  useful  if  you  are  sending  simulation  files  to  another  person.  By  including  a  workspace   file  with  the  results  of  interest  you  can  be  sure  that  they  will  view  the  correct  information.  This  can  be  particularly   valuable   when   sending   files   to   someone   who   is   not   an   experienced   OrcaFlex   user.   This   can   even   be   useful   when   sending   files   to   Orcina   for   software   support   because   they   contain   a   precise   specification   of   the   results   you   are   interested  in.  

3.14

COMPARING  DATA  

The   Compare   Data   menu   item   opens   the   Compare   Data   form,   which   allows   you   to   find   differences   between   the   data   in  two  OrcaFlex  files.   The   comparison   is   done   using   a   user-­‐provided   compare   program,   so   when   you   first   use   this   facility   you   need   to   configure  OrcaFlex  to  tell  it  which  compare  program  that  you  want  to  use;  see  Configuration  below.   You  can  then  compare  files  as  follows:   x

On  the  Files  page,  specify  the  two  files  that  you  wish  to  compare.  These  can  be  data  or  simulation  files.  

x

Click  the  Compare  button.  

x

OrcaFlex  then  saves  the  data  from  the  two  files  to  temporary  text  files  and  then  runs  the  user-­‐specified  compare   program  to  compare  those  text  files.  

As   an   alternative   to   comparing   two   data   files   on   disk   you   can   optionally   choose   to   compare   the   currently   loaded   model  with  a  single  file  on  disk.   Configuration   On   the   Configuration   page   you   need   to  tell   OrcaFlex   the   text   file   compare   program   that   you   want   to   use,   and   how  to   use   it.   The   compare   program   must   be   a   program   that   can   compare   text   files   passed   to   it   through   the   command   line.   Various  such  programs  are  available  on  the  web;  examples  are  WinDiff,  Compare  It!  and  Araxis  Merge.   Compare  Program  

This  is  the  compare  program's  executable  file  name.  You  can  specify  either  the  full  path,  or  just  the  file  name  if  the   executable  file  resides  in  a  directory  which  is  on  your  system  path.   A  basic  compare  program  called  WinDiff  is  freely  available  (you  can  find  it  by  searching  the  Internet)  and  is  quite   sufficient  for  this  purpose.  

79  

w

 

User  Interface,  Preferences    

Command  Line  Parameters  

This  defines  the  command  line  parameters  that  are  passed  to  the  compare  program.  OrcaFlex  replaces  the  special   strings  %1  and  %2  with  the  file  names  of  the  temporary  text  files.  For  most  compare  programs  the  default  setting  of   "%1  %2"  will  be  sufficient.  Otherwise  you  will  need  to  consult  the  documentation  of  your  compare  program.  

3.15

PREFERENCES  

OrcaFlex  has  a  number  of  settings  that  can  be  customised  to  suit  the  way  that  you  work.  The  majority  of  settings  can   be  adjusted  in  the  Preferences  form,  which  is  accessed  by  using  the  Tools  |  Preferences  menu  item.   3D  View  Preferences   Minimum  Drag  Distance  

Object  positions  are  not  updated  until  the  mouse  has  been  dragged  at  least  this  distance  (in  pixels).  This  prevents   accidental   changes   to   object   positions.   To   make   a   small   movement,   drag   away   and   then   back   again,   or   edit   the   coordinate  directly  in  the  object's  Edit  Form.   View  Rotation  Increment  

Each  CLICK  on  a  Rotate  View  button  increments  or  decrements  View  Azimuth  or  Elevation  by  this  amount.   Refresh  Rate  

During   a   simulation   calculation   all   3D   View   and   Graph   windows   are   updated   at   this   rate.   Selecting   a   faster   rate   allows  you  to  see  the   behaviour  of  the  simulation  more  clearly  at  the   expense  of   performance.   Set  a  slow  Refresh   Rate  to  give  the  numerical  calculation  more  processor  time.   Background  Colour  

This  sets  the  background  colour  of  all  3D  View  windows.   Locate  Object  Method  

Can   be   either   Flash   object   or   Hide   other   objects.   It   determines   what   method   the   Locate   action   in   the   model   browser  uses.   x

When  the  Flash  object  preference  is  set  then  the  Locate  action  repeatedly  draws  and  hides  the  object  on  the  3D   View,  like  a  blinking  cursor.  

x

When  the  Hide  other  objects  preference  is  set  then  the  Locate  action  temporarily  hides  all  other  objects.  

Normally  the  default  setting  of  Flash  object  is  sufficient  to  locate  objects.  However,  if  the  model  you  are  searching   for   is  obscured  by  other   objects   then   this  method  may  not   help   you   to   locate   the   object.   In   this   case   you   should  use   the  Hide  other  objects  preference.   3D  View  Axes  Preferences   View  Axes  

The   view   axes   show   the   same   directions   as   the  global   axes,   but  are   drawn   in   the   top   right   hand   corner   of   3D   views,   rather  than  at  the  global  origin.  Can  also  be  set  from  the  View  menu.   Scale  Bar  

Determines  whether  a  scale  bar  is  drawn  in  3D  views.  Can  also  be  set  from  the  View  menu.   Note:  

The   Scale   Bar   is   not   drawn   for   shaded   graphics   views   because   it   would   be   meaningless   due   to   perspective.  

Global  Axes  

Determines  whether  the   global  axes  are  drawn,  at  the  model's  global  origin  (0,0,0).  Can  also  be  set  from  the  View   menu.   Environment  Axes  

Determines  whether  the  wave,  current  and  wind  directions  are  drawn  in  the  3D  view.  When   multiple  wave  trains   are  present  the   first   wave  train  is   taken   to   be  the  dominant  one  and  is  drawn  using   sea  surface  pen,  whereas   the   other  wave  trains'  directions  are  drawn  in  the  secondary  wave  direction  pen.  Can  also  be  set  from  the  View  menu.  

80  

w

 

User  Interface,  Preferences  

 

Local  Axes  

Determines   whether   the   local   axes   for   vessels,   buoys   and   line   ends   are   shown.   Drawing   the   local   axes   on   the   3D   view  helps  you  check  the  orientations  of  these  objects.  This  preference  can  also  be  set  from  the  View  menu.   Note:  

Local  Axes  are  not  drawn  for  shaded  graphics  views.  

Node  Axes  

Determines  whether  axes  for  line  nodes  are  shown.  This  preference  can  also  be  set  from  the  View  menu.   Out  of  Balance  Forces  

If   selected,   then   in   the   static   analysis   (not   during   the   simulation)   there   are   extra   lines   drawn   on   the   3D   view,   representing   the   out   of  balance   force   acting   on   each  vessel   and  buoy.  This  preference   is  sometimes   useful  for   static   analysis,  since  it  enables  you  to  see  how  far  a  buoy  or  vessel  is  from  being  in  equilibrium.  This  preference  can  also   be  set  from  the  View  menu.   The  force  is  drawn  as  a  line,  starting  at  the  force's  effective  point  of  application,  and  whose  length  represents  the   size  of  the  force.  The  scaling  is  piecewise  linear  and  based  on  the  View  Size  of  the  3D  view.  Lines  up  to  ViewSize/2   long   mean   forces   up   to   10   force   units   and   lines   from   ViewSize/2   to   ViewSize   mean   forces   from   10   to   1000   force   units.   Note:  

Out  of  Balance  Forces  are  not  drawn  for  shaded  graphics  views.  

Video  Preferences   The  video   preferences  allow  you  to  control  the  compression  algorithm  used  for  exported  video.  The  software  which   performs   this   compression   is   called   a   codec.   Because   the   different   graphics   modes   produce   very   different   images   they  require  different  types  of  codec.   Shaded  Graphics  Codec  

The   run-­‐length   encoding   which   works   well   for   wire   frame   graphics   is   not   suitable   for   shaded   replays   and   in   fact   there   is   no   suitable   built-­‐in   codec   in   Windows.   We   would   recommend   using   an   MPEG-­‐4   codec   of   which   many   are   available.   In   our   experience   the   freely   available   (licensed   under   the   GPL)   XVID   codec   performs   very   well.   The   Shaded  Graphics  topic  has  more  information  about  the  XVID  codec.   Another  reasonable   choice   is   the  Windows  Media   Video  9   codec,   which  is  identified  by  the   code   WMV3.   This  codec   produces  lower  quality  videos  than  XVID  for  the  same  video  file  size,  but  does  have  the  advantage  that  the  videos   should  work  on  almost  all  Windows  machines  without  the  need  for  codec  installation.  Details  on  how  to  download   this  codec  can  be  found  at:  www.orcina.com/Support/ShadedGraphics.   You   can   choose   to  use   other   codecs   that   are   installed   on   your   machine.   Should   you   do   so   then   you   must   also   specify   the  following  information:   x

Codec  4  character  code:  Codecs  are  identified  by  unique  codes,  4  characters  long.  Good  alternatives  to  XVID   and  WMV3  include  DIVX,  the  3ivx  codec  (character  code  3IV2).  

x

Padding:   MPEG-­‐4   codecs   commonly   require   round   number   frame   sizes   (width   and   height   in   pixels).   For   example   XVID   requires   frame   sizes   to   be   multiples   of   8.   When   OrcaFlex   exports   the   video   it   ensures   that   the   frame   sizes   are   a   multiple   of   this   number.   If   you   are   unsure   of   what   number   to   use   for   your   codec   then   we   recommend  trying  8  which  usually  works.  

x

Colour  depth:  Some  MPEG-­‐4  codecs  require  a  specific  colour  depth.  Again,  if  you  are  unsure  of  what  value  to   use  then  we  recommend  trying  32  bit  or  16  bit  colour  depth.  

Wire  Frame  Graphics  Codec  

Run-­‐length   encoding   is   the   default   setting   and   is   usually   the   best   choice.   This   codec   offers   good  compression   rates   for  OrcaFlex  wire  frame  video.  The  AVI  files  produced  using  this  codec  can  be  played  on  most  Windows  PCs.   If  you  choose  Uncompressed  then  each  frame  of  the  video  is  stored  as  an  uncompressed  bitmap.  This  means  that   the  AVI  file  produced  can  be  extremely  large.   Output  Preferences   Printer  Margins  

These  set  the  Left  and  Top  margins  used  on  printouts.  

81  

w

 

User  Interface,  Printing  and  Exporting    

Monochrome  Output  

If  this  is  checked  then   external  output  (copying  to  the  clipboard,  exporting  metafiles  and  printing)  is  in   black  and   white.  This  is  useful  with  black  and  white  printers,  since  otherwise  pale  colours  may  be  drawn  in  very  light  grey  and   may  be  hard  to  see.   Miscellaneous  Preferences   Show  Splash  Screen  

Determines  whether  OrcaFlex  displays  its  splash  screen  when  the  program  starts.   Batch  Auto  Save  

If   this   is   enabled   then   simulations   run   in   batch   mode   are   automatically   stored   to   simulation   files   at   the   specified   regular   Auto   Save   Interval.   This   is   useful   if   your   computer   is   prone   to   failure   (for   example   because   of   overnight   power   failures)   since   the   part-­‐run   simulation   file   can   be   loaded   and   continued,   rather   than   having   to   re-­‐run   the   whole  simulation  from  scratch.  The  Auto   Save  Interval  should  be   neither  too  short,  since  then  the   program   will  then   waste  a  lot  of  time  repeatedly  storing  away  the  results,  nor  too  long,  since  then  a  lot  of  simulation  work  will  be  lost  if   a  failure  occurs.  

3.16

PRINTING  AND  EXPORTING  

The  Print  /  Export  form  is  accessed  using  either  the  File  |  Print  or  the  File  |  Export  menu  item  and  allows  you  to   choose  one  or  more  of  the  following  items  to  be  printed  or  saved  to  file:   x

The   model   data.   Vessel   Types   often   have   very   large   amounts   of   data,   much   of   which   may   not   apply   to   the   current  model,  so  OrcaFlex  offers  you  the  option  of  printing  all  the  vessel  type  data  or  only  the  data  that  is  in   use.  

x

Any  3D  Views,  Graphs,  Spreadsheets  and  Text  Windows  currently  on  display.   Note:  

Graphs  are  printed  as  large  as  possible  whilst  maintaining  aspect  ratio.  

82  

w

 

Automation,  Introduction  

 

4

AUTOMATION  

4.1

INTRODUCTION  

OrcaFlex  provides  several  important  facilities  for  automating  and  post-­‐processing  work:   x

OrcaFlex   is   supplied   with   a   special   Excel   spreadsheet   which   enables   you   to   automate   the   extraction   of   simulation  results  into  your  own  spreadsheet.  You  can  then  use  the  normal  Excel  calculation  facilities  to  do  your   own  customised  post-­‐processing  and  graphing.  

x

The   Batch   Processing   facility   enables   you   to   run   a   set   of   simulations   in   unattended   mode,   for   example   as   an   overnight  job.   The  simulations  can   either  be   of  pre-­‐prepared  data  files,  or   else  can   be  specified  by  a  batch  script   file  that  specifies  the  simulation  as  variations  on  a   base  data  file.  The   OrcaFlex   Spreadsheet   mentioned  above   also  has  facilities  for  automating  the  production  of  batch  script  files  and  text  data  files.  

x

OrcaFlex   includes   a   well-­‐documented   programming   interface   called   OrcFxAPI   (short   for   OrcaFlex   Application   Program   Interface).   See   the   OrxFxAPI   help   file   for   details.   OrcFxAPI   is   a   Windows   dynamic   link   library   (DLL)   that   is   installed   when   you   install   OrcaFlex,   and   which   provides   facilities   for   setting   data,   calculating   static   positions  and  extracting  results  from  those  calculations  or  from  pre-­‐run  simulation  files.  For  example  you  can   write  programs  to  automate  post-­‐processing  or  that  use  OrcaFlex  as  a  'statics  calculation  engine'.  One  important   example   application   of   this   is   for   real-­‐time   monitoring   of   pipes,   moorings   etc.   For   further   information   or   to   discuss  possible  applications  of  OrcFxAPI,  please  contact  Orcina.  

4.2

BATCH  PROCESSING  

4.2.1

Introduction  

Simulations,   script   files,   post   processing   spreadsheets   and   fatigue   analyses   can   all  be   run   in   unattended   mode,   by   using  the  Calculation  |  Batch  Processing  menu  item.  This  command  opens  a  form  that  allows  you  to  set  up  a  list  of   jobs  that  are  to  be  run.  The  list  can  include  any  number  and  mixture  of  the  following  types  of  job:   1.

Static  analysis  of  pre-­‐prepared  OrcaFlex  data  files  (.dat  or  .yml).  OrcaFlex  opens  the  data  file,  performs  the  static   analysis   and   then   saves   the   results   in   a   simulation   file   with   the   same   name   as   the   data   file,   but   with   a   .sim   extension.  

2.

Dynamic  analysis  of  pre-­‐prepared  OrcaFlex  data  files  (.dat  or  .yml).  OrcaFlex  opens  the  data  file,  performs  the   static  analysis,  runs  the  dynamic  simulation  and  then  saves  the  results  in  a   simulation  file  with  the  same  name   as  the  data  file,  but  with  a  .sim  extension.  

3.

Partially-­‐run   OrcaFlex   simulation   files   (.sim).   OrcaFlex   opens   the   simulation   file,   finishes   the   dynamic   simulation  and  then  saves  the  completed  simulation,  overwriting  the  original  file.  

4.

A  batch  script  file  (.txt).  This  is  a  text  file  which  contains  OrcaFlex  script  commands.  OrcaFlex  opens  the  script   file  and  obeys  the  commands  in  turn.  The  most  common  use  of  script  files  is  to  perform  a  series  of  systematic   variations  on  a  base  data  file.  

5.

A  fatigue   analysis  file  (.ftg).  OrcaFlex   performs   the   fatigue   analysis  and   saves   the   results   to   an   Excel   compatible   spreadsheet  of  the  same  name  but  extension  .xls.  

6.

An  OrcaFlex  Spreadsheet  (.xls  or  .xlsx).  OrcaFlex  will  process  all   Instructions  sheets  in  the  Excel  workbook.  Note   that  if  the  spreadsheet's  "Contains  Dependencies"  options  is  checked  (or  the  spreadsheet  is  pre  OrcaFlex  v9.4)   then   the   workbook   will   be   processes   as   a   single   job   us‹‰ ƒ •‹‰Ž‡ –Š”‡ƒ†Ǥ ˆ ‹– ‹•ǯ– …Š‡…‡†ǡ –Š‡ ‡ƒ…Š instructions  sheet  will  be  broken  down  into  multiple  load  cases  which  are  individually  added  to  the  batch  and   may  be  processed  simultaneously.     Note:  

If   you  wish   to  use  Excel   for  any   reason  while   OrcaFlex  is   processing   spreadsheets   within   a  batch   it   is   important   that   you   open  Excel   first,   then   open   the   file   you   wish   to   work   on.   The   reason   for   this   is   that   when   you   double   click   an   Excel   file,   Windows   will   try   to   use   the   copy   of   Excel   OrcaFlex   has   claimed,  resulting  in  unpredictable  failures.  

When  adding   data  files  (.dat  or  .yml)  you  need   to  specify  whether   static  or  dynamic  analysis  is  to  be  performed.   This  choice  is  made  from  the  Add  Files  file  dialogue  window,  or  from  the  popup  menu.   OrcaFlex   can  auto-­‐save  partial   completed  dynamic   simulations   to  file   at   regular   intervals   during   the  batch  job.  This   is  useful  if  your  computer  is  prone  to  failure  (for  example  because  of  overnight  power  failures)  since  the  part-­‐run   simulation  file  can  be  loaded  and  continued,  rather  than  having  to  re-­‐run  the  whole  simulation  from  scratch.  

83  

w

 

Automation,  Batch  Processing    

Multi-­‐threading   The   batch  processing  functionality  can  make  use  of   multiple  processor  cores.  So,  for  example,   if  you  have  a  quad-­‐ core  machine  then  4  simulation  files  can  be  run  concurrently.   Since   some   batch   tasks   can   depend   on   the   output   of   other   tasks   OrcaFlex   processes   tasks   in   a   very   particular   order,   as  follows:   x

The   batch   script   files   are   all   processed   first.   Because   it   is   common   to   write   scripts   that   output   data   files   it   is   important  to  complete  all  batch  scripts  before  processing  the  data  files.  

x

Any  data  or  simulation  files  are  processed  next.  

x

Fatigue  files  are  processed  next.  These  use  simulation  files  as  input  and  so  should  not  be  started  until  all  data  or   simulation  files  have  been  processed.  

x

Finally   any   OrcaFlex   spreadsheet   files   or   load   cases   are   processed.   These   also   cannot   be  started   until   all   data   or   simulation  files  have  been  processed.  

The   commands   in   batch   script   files   are   processed   sequentially.   Consequently   any   simulations   that   are   processed   with  Run  commands  cannot  be  performed  in  parallel.  Because  of  this  it  is  advisable  to  use  the   SaveData  command   rather  than  the  Run  command  when  creating   batch  scripts.  Such  a  script   would  create  a  number  of   OrcaFlex  data   files  which  you  could  then  process  in  the  batch  form  using  all  available  processor  cores.   Batch  Form  User  Interface   Close  

Dismisses  the  batch  form.   Add  Files  

Adds  jobs   to  the  list.  The  standard   file  dialogue   window   is  displayed,   where  you  select  one  or  more  files  to  be  added   to  the  list.   Files  can  also  be  added  by  drag  and  drop.  That  is  if  you  are  browsing  your  file  system  then  you  can  highlight  files   and  drag  them  onto  the  jobs  list.   Remove  Files  

Removes  any  files  highlighted  in  the  jobs  list.   Check  Files  

OrcaFlex  opens  each  file  in  the  jobs  list,  checks  that  they  contain  valid  OrcaFlex   data  or  script  commands  and  reports   any  errors.  When  checking  OrcaFlex  spreadsheet  or  fatigue  files  it  simply  confirms  the  file  exists.   Run  Batch  

Processes  the  list  of  jobs.  If  a  job  fails  then  it  is  abandoned  but  other  jobs  are  still  attempted.  Any  errors  are  reported   once  all  jobs  have  been  processed.   Pause  Batch  

Pauses   the   currently   running   batch   jobs.   This   can   be   useful   if   you   temporarily   want   another   process   on   your   machine  to  have  the  processor  resource  that  OrcaFlex  is  using.   Stop  Batch  

Terminate  processing  of  batch  jobs.   Warnings  

Displays   a   window  allowing   you   to  review   all   warnings   generated   by   OrcaFlex   during   a   calculation.   These   warnings   are   suppressed   when   you   are   operating   in   batch   mode   and   this   button   allows   you   to   review   them   once   the   simulation  has  completed.   Close  program  when  Batch  completes  

If   checked   then   OrcaFlex   will   close   once   the   processing   of   jobs   completes.   This   feature   is   intended   principally   for   users  with  networked  licences.  It  allows  you  to  release  your  claim  on  an  OrcaFlex  licence  as  soon  as  the  batch  of  jobs   is  complete.  

84  

w

 

Automation,  Batch  Processing  

 

4.2.2

Script  Files  

OrcaFlex   provides   special   facilities   for   running   a   series   of   variations   on   a   base   data   file,   using   a   script   file.   This   contains   a   sequence   of   commands   to   read   a   data   file,   make   modifications   to   it,   and   run   the   modified   file,   storing   the   results  for  later  processing.  The  file  can  also  include  comments.  The   syntax  for  the  instructions  is  described  in  the   next  topic.   Script   files   can   be   written   using   any   text   editor.   Alternatively,   there   are   facilities   in   the   OrcaFlex   spreadsheet   for   automatically  generating  script  files  for  regular  sets  of  cases.  

4.2.3

Script  Syntax  

An   OrcaFlex   batch   script   is   made   up   of   commands,   which   are   obeyed   sequentially,   and   comments,   which   are   ignored.  A  comment  is  a  line  that  is  either  blank  or  on   which  the  first  non-­‐blank  characters  are  "//".  A  command  can   be:   1.

A   directive   followed   by   one   or   more   arguments,   optionally   separated   by   white   space   (one   or   more   spaces   or   tabs).   For   example:  load  c:\temp\test.dat  where  load  is  the   directive   and  c:\temp\test.dat  is  the   argument.  

2.

An  assignment  of  the  form  VariableName=value,  again  with  optional  white  space  separators.  For  example:   Length  =  55.0.  

Note  that:   x

Directives,  variable  names,  and  model  object  names  are  all  case  independent.  

x

If  your  script  includes  a  relative  file  name  then  it  is  taken  to  be  relative  to  the  directory  from  which  the  script   was  loaded.  

x

File   names,   arguments,   variables   or   values   containing   spaces   or   non-­‐alphanumeric   characters   must   be   enclosed   in  single  or  double  quotes  and  they  must  not  contain  the  same  quote  character  as  is  used  to  enclose  them.  For   example  '6"  pipe'  and  "200'  riser"  are  valid,  but  the  following  are  not  valid:   6  inch  pipe  Ȃ  contains  spaces,  so  needs  to  be  enclosed  in  quotes;   6"pipe  Ȃ  contains  a  double  quote,  so  needs  to  be  enclosed  in  single  quotes;   '6'  pipe'  Ȃ  contains  a  single  quote,  so  needs  to  be  enclosed  in  double  quotes  instead  of  single.  

4.2.4

Script  Commands  

The   following   batch   script   commands   are   currently   available.   You   need   to   put   quotes   round   file   names   or   other   parameters  that  include  spaces  or  non-­‐alphanumeric  characters.   Load    

Opens  the  OrcaFlex  file  named  .  The  file  can  be  either  a   data  file  or  a  simulation  file.   LoadData    

Opens  the  data  from  the  OrcaFlex  data  file  named  .   RunStatics    

Perform   statics   for   the   current   model   and   save   the   resulting   simulation   to   .   After   the   file   is   saved   the   model  is  reset.   RunDynamics    

Run   dynamics   for   the   current   model   and   save   the   resulting   simulation   to   .   After   the   file   is   saved   the   model  is  reset.   Run    

Identical  to  RunDynamics.   Save    

Save  the  current  model  to  .  

85  

w

 

Automation,  Batch  Processing    

If   calculation   results   (either  statics   or   dynamics)   are   available   then   a  simulation   file  will   be   saved.   Otherwise   a  data   file  will  be  saved.  When  saving  data,  if  the  file   extension  is  .yml  then  a  text  data  file  will  be  saved;  otherwise  a   binary   data  file  will  be  saved.   SaveData    

Save  the  data  from  the  current  model  to  .   If  the  file  extension  is  .yml  then  a  text  data  file  will  be  saved;  otherwise  a  binary  data  file  will  be  saved.   Note:  

In   the   Load/LoadData,   Save/SaveData   and   RunStatics/RunDynamics/Run   commands,   if    is  a  relative  path  then  it  is  taken  to  be  relative  to  the  directory  from  which  the  script   file  was  loaded.  

ExtendSimulation    

Adds   a   new   stage   of   length   .   This   command   is   equivalent   to   the   Calculation   |   Extend   Dynamic   Simulation  menu  item.  You  would  normally  follow  this  command  with  a  Run  command.   Reset  

Resets  the  current  model.  This  command  is  equivalent  to  the  Calculation  |  Reset  menu  item.   NewModel  

Deletes  all  objects  from  the   current  model  and  resets  data  to  default  values.  This  command  is  equivalent  to  the  File  |   New  menu  item.   Create    []  

Creates  a  new  object  of  type  .  The  new  object  is  automatically  selected  which  means  that  subsequent   assignment  commands  apply  to  this  new  object.   The     parameter   can   be   "Line   Type",   "Vessel   Type",   "Line",   "Winch"   etc.   Select   Edit   |   Add   from   the   Model  Browser  menu  to  see  a  list  of  possible  values  for  this  parameter.   Alternatively   variable   data   sources   can   be   created   by   setting   the     parameter   to   "Bending   Stiffness",   "Drag  Coefficient"  etc.  This  list  of  possible  variable  data  source  object  types  can  be  found  in  the  Data  Source  Type   tree  on  the  variable  data  form.   If  the  optional    parameter  is  included  then  the  new  object  will  be  given  that  name.   Delete    

Deletes  the  object  called  .   Select  []    

Specify  the  model  object  to  which  subsequent  assignment  commands  will  apply.   The     parameter   is   optional,   and   by   default   is   'object',   meaning   select   the   named   model   object.     must   then   be   either   the   name   of   an   object   that   exists   in   the   current   model   or   one   of   the   reserved   names  'General'  (for  the  General  data  form)  or  'Environment'  (for  the  Environment  data  form).   Some  examples  of  the  select  and  assignment  commands  are  given  in   Examples  of  setting  data.   Other    values  only  need  to  be  specified  in  the  following  special  cases.   If   the   Environment   has   been   selected   and   there   is   more   than   one  wave   train,   then   before   you   can   specify   any   wave   train   data   you   must   give   another   select   command   to   select   the   wave   train.   This   second   select   command   has   the   form:   Select  WaveTrain    

So,  for  example:   Select  Environment      Select  WaveTrain  Primary          WaveDirection  =  30.0  

Similarly,  if  the  Environment  has  been  selected  and  there  is  more  than  one   current  data  set,  then  you  must  select   one  of  them  before  specifying  any  current  data.  For  example:  

86  

w

 

Automation,  Batch  Processing  

 

Select  Environment      Select  Current  Current2          RefCurrentDirection  =  270.0  

Note  that  this  is  not  the  same  as  setting  the   Active  Current.  In  fact,  you  should  avoid  setting  up  multiple  current  data   in  batch  scripts  if  possible:  this  is  best  done  interactively  on  the  Environment  form.   If   a   vessel   type   has   been   selected   and   it   has   more   than  one  draught,   then  before   specifying   any  draught-­‐dependent   data  you  must  give  another  select  command  that  selects  the  draught.  This  second  select  command  has  the  form:   Select  Draught    

Before  specifying  data  for  RAOs  you  need  to  specify   the   type  of  RAOs  Ȃ  this  can  be  either   Displacement,  WaveLoad   or  QTF.  This  is  done  with  a  command  of  the  form:   Select  RAOs    

Similarly,   before   specifying   vessel   type   data   for   a   given   wave   direction   you   must   give   another   select   command   to   select  that  direction.  This  takes  the  form:   Select  Direction    

So,  for  example:   Select  "Vessel  Type1"      Select  Draught  Transit          Select  RAOs  Displacement              RAOOriginX  =  10              RAOOriginY  =  0              RAOOriginZ  =  2              Select  Direction  45                  RAOSurgeAmplitude[2]  =  0.1              Select  Direction  90                  RAOSurgeAmplitude[2]  =  0.16  

Note:  

Indentation  with  spaces  or  tabs  is  optional,  but  makes  scripts  more  readable.  

Assignment  

Assignment  commands  take  the  form   VariableName  =  Value  

The   VariableName   on   the   left   hand   side   must   be   one   of  the   recognised  variable   names   and   the   named   variable   must   exist  in  the  currently  selected  model  object.  The  Value  on  the  right  hand  side  must  be  in  the  appropriate  form  for   that  variable  (i.e.  numeric  or  text)  and  it  must  be  given  in  the  same  units  as  used  in  the  current  model.   For  example:   Select  Vessel1      Length  =  110      Draught  =  "Operating  draught"  

If  VariableName  is  the  name  of  a  variable  that  appears  in  a  check  box  in  OrcaFlex  then  the  Value  should  be  True  or   False.  For  example:   Select  Environment      CurrentRamp  =  True  

If  VariableName  is  the  name  of  a  variable  that  appears  in  a  table  in  OrcaFlex,  then  its  row  number   must  be  given.   The  row  number  is  given  as  an  index  enclosed  by  either  square  or  round  brackets  (don't  mix  them  on  the  same  line),   and  is  always  1-­‐based,  i.e.  [1]  is  the  first  row  of  the  table.  Note  that  this  sometimes  requires  care,  since  in  OrcaFlex   the  table  might  not  be  1-­‐based.  For  example,  when  setting  the  prescribed  motion  for  a  vessel,  the  command   PrescribedMotionVelocity[2]  =  4.8  

sets   the   velocity  in   the  2nd  row  of   the   table,  but  in   this  case   the   first   row   of  the   table   is  stage   0   (the  build-­‐up   stage)   so  this  command  (slightly  confusingly)  sets  the  velocity  for  stage  1.   More  examples  of  the  select  and  assignment  commands  are  given  in  Examples  of  setting  data.  

87  

w

 

Automation,  Batch  Processing    

InvokeWizard  

Sets  the  data  for  the  selected  object  using  either  the  Line  Type  Wizard  or  the  Plasticity  Wizard.  The  selected  object   must   be  either   a   line   type  or   a   bend   stiffness   variable  data   source.   The   input  data   for   the  Wizard   should   first  be   set   using  data  assignment  commands.   An  example  of  how  to  use  this  command  is  given  in  Examples  of  setting  data.   WaveSearch    

Exports   the   wave   search   spreadsheet   to  the   specified   file.   The   file   can   be   an   Excel   spreadsheet   (.xls),   a   tab   delimited   file   (.txt)   or   a   comma   separated   file   (.csv).   The   decision   is   taken   based   on   the   file   extension   that   you   specify.   The   input  data  for  the  wave  search  should  first  be  set  using  data  assignment  commands.   DisplacementRAOsReport    []   SpectralResponseReport    []  

Exports  the  vessel  response   report  spreadsheets  to  the  specified  file  for  the  specified  vessel.  The  file  can   be  an  Excel   spreadsheet   (.xls),   a   tab  delimited   file   (.txt)  or   a   comma  separated   file   (.csv).   The  decision   is  taken   based   on   the  file   extension  that  you  specify.  If  no  vessel  is  specified,  and  there  is  only  one  vessel  in  the  model,  then  that  vessel  will  be   used.  The  input  data  for  the  response  reports  should  first  be  set  using  data  assignment  commands.   SHEAR7DataFile      

Exports  to    a  SHEAR7  data  file  for  the  line  named  .   SHEAR7MdsFile      [  ]  

Exports  to    a  SHEAR7  Mds  file  for  the  line  named  .   The    and     parameters  are  optional.  If  they  are  specified  then  mode  numbers  in  the   range    to    inclusive  are  exported.  If  these  parameters  are  omitted  then  all  modes  are  exported.   Only  the  Transverse  and  Mostly  Transverse  modes  are  included  in  the  exported  file.  If  you  have  specified  first  and   last  modes  to  export  then  these  mode  numbers  refer  to  the  transverse  and  mostly  transverse  modes.  The  program   takes  the  following  steps:   1.

Calculate  all  modes.  

2.

Sort  the  modes  into  order  of  decreasing  period  /  increasing  frequency.  

3.

Remove  all  modes  which  are  not  Transverse  or  Mostly  Transverse.  

4.

Export  the  modes  in  the  range    to    inclusive.  

SHEAR7OutFile      

Exports  to    the  SHEAR7  .out  file  for  the  line  named  .  This  command  is  only  available  if  the   direct  SHEAR7  interface  is  in  use.   SHEAR7PltFile      

Exports  to    the  SHEAR7  .plt  file  for  the  line  named  .  This  command  is  only  available  if  the   direct  SHEAR7  interface  is  in  use.  

4.2.5

Examples  of  setting  data  

The   Select   command   is   probably   the   most   difficult   script   command   to   use.   To   help   understand   how   it   works   we   present  some  examples  of  its  use  below:   Simple  examples   For  many  objects  the  script  commands  for  setting  data  take  the  form:   1.

Select  the  object  using  the  command  Select  .  

2.

Set  the  data  using  one  or  more  commands  of  the  form  VariableName  =  Value.  

The   object   name   is   most   easily   found   on   the   Model   Browser.   The   variable   name   is   found   by   opening   the   relevant   data  form,  selecting  the  required  data  item  and  pressing   F7.   Some  examples  of  this  procedure  follow:   Select  Link1      UnstretchedLength  =  50  

88  

w

 

Automation,  Batch  Processing  

 

  Select  "3D  Buoy1"      Mass  =  4      Volume  =  8      Height  =  7.5     Select  Line1      IncludeTorsion  =  Yes  

Note:  

The   name   "3D   Buoy"   needs   to   be   enclosed   in   quotes   because   it   contains   a   space.   If   the   name   contains  a  double  quote  and  spaces  then  it  should  be  enclosed  with  single  quotes.  

Data  in  tables  and  indices   Some   data   in   OrcaFlex   appears   in   tables.   For   example   consider   the   Structure   page   of   the   Line   Data   form   which   specifies  how  a  Line  is  made  up  of  a  number  of  sections.  Each  section  is  specified  by  its  Line  Type,  length  etc.  The   following  example  sets  the  number  of  sections  of  the  line  to  be  2  and  then  sets  data  for  both  sections  in  turn.   Select  Line1      NumberOfSections  =  2        LineType[1]  =  Riser      Length[1]  =  75      TargetSegmentLength[1]  =  4        LineType[2]  =  Rope      Length[2]  =  200      TargetSegmentLength[2]  =  20  

Note  that  we  use  blank  lines  to  lay  out  the  script.  This  is  not  essential  but  makes  reading  the  script  easier.   Data   which   appears   in   tables   is   always   set   using   the   indexed   notation   used   above.   Having   stated   this   rule,   we   immediately  break  it  in  the  section  below!   Line  Type,  Clump  Type  and  Flex  Joint  Type  data   These  data  are  set  by  first  selecting  the  type  by  name  and  then  assigning  the  data  as  illustrated  below:   Select  "Line  Type1"      OuterDiameter  =  0.28      InnerDiameter  =  0.21  

On   the   Line   Types   Data   form   there   is   an   option   to   view   the   data   for   all   Line   Types   at   once   or   to   view  it   by   individual   Line  Type.  When  the  data  is  being  viewed  for  all  Line  Types  at  once  the  data  appears  in  tables  with  one  row  per  Line   Type.  However,  the  data  must  be  set   by   first  selecting  the  type  by  name  and  then  assigning  the   data.  You  cannot   set   Line  Type  using  index  notation.   Similar  rules  apply  to  Clump  Type  data  and  to  Flex  Joint  Type  data.   Drag  Chain  Type  and  Wing  Type  data   These  data  are  also  set  by  first  selecting  the  type  by  name  and  then  assigning  the  data.  For  example:   Select  "Drag  Chain  Type1"      Length  =  12     Select  "Wing  Type1"      NumberOfAngles  =  12      Angle[2]  =  -­80      Lift[2]  =  0.2      Drag[2]  =  0.15      Moment[2]  =  0.5  

Data  found  on  the  General  Data  form   Data  found  on  the  General  Data  form  can  be  set  as  follows:   Select  General      InnerTimeStep  =  0.01      OuterTimeStep  =  0.1  

89  

w

 

Automation,  Batch  Processing    

Data  found  on  the  Environment  Data  form   Data  found  on  the  Environment  Data  form  can  be  set  as  follows:   Select  Environment      SeaBedStiffness  =  3000      SeaBedDamping  =  80  

For  data  specific  to  a  particular  wave  train  you  must  first  select  the  Environment  and  then  select  the  particular  wave   train.  This  makes  use  of  the  alternative  syntax  for  Select  which  is   Select    .  For  a  wave   train  you  replace    with  WaveTrain  and  replace    with  the  name  of  the  wave  train  as   defined  on  the  Environment  Data  form.  For  example:   Select  Environment      Select  WaveTrain  "Swell  from  SW"          WaveTrainDirection  =  135          WaveTrainType  =  "Dean  Stream"          WaveTrainHeight  =  2.5          WaveTrainPeriod  =  18      Select  WaveTrain  "Local  Wind  Sea""          WaveTrainDirection  =  40          WaveTrainType  =  JONSWAP          WaveTrainHs  =  5.7          WaveTrainPeriod  =  9  

Data  for  Current  data  sets   Multiple  Current  data  sets  can  be  defined.  Again  this  requires  the  alternative  syntax  for  Select  as  shown  below:   Select  Environment      MultipleCurrentDataCanBeDefined  =  Yes          NumberOfCurrentDataSets  =  2          CurrentName[1]  =  "120deg"          CurrentName[2]  =  "150deg"          Select  Current  "120deg"              RefCurrentDirection  =  120          Select  Current  "150deg"              RefCurrentDirection  =  150      ActiveCurrent  =  "150deg"  

Vessel  Type  Data   Some  Vessel  Type  data  is  set  in  a  straightforward  manner  as  follows:   Select  "Vessel  Type1"      Length  =  120      PenWidth  =  3      Symmetry  =  "XZ  plane"  

Note:  

Because  the  symmetry  value  (XZ  plane)  contains  a  space  it  must  be  enclosed  in  quotes.  

However,   the   majority   of   Vessel   Type   data   requires   that   you   also   specify   which   draught   the   data   applies   to.   For   example:   Select  "Vessel  Type1"      Select  Draught  "Transit  Draught"          CurrentCoeffSurgeArea  =  1200          CurrentCoeffSwayArea  =  1100          CurrentCoeffYawAreaMoment  =  120E3  

To  set  data  for  displacement  RAOs,  wave  load  RAOs  and  wave  drift  QTFs  you  must  also  specify  which  type  of  RAO   the  data  applies  to.  For  example:   Select  "Vessel  Type1"      Select  Draught  "Survival  Draught"            Select  RAOs  Displacement              RAOOriginX  =  10              RAOOriginY  =  0              RAOOriginZ  =  2            Select  RAOs  WaveLoad  

90  

w

 

Automation,  Batch  Processing  

 

           RAOOriginX  =  0              RAOOriginY  =  0              RAOOriginZ  =  0            Select  RAOs  QTF              RAOOriginX  =  -­3              RAOOriginY  =  0              RAOOriginZ  =  4  

Note   that   the   variable   names   are   the   same   but   different   data   is   set   depending   on   which   type   of   RAOs   has   been   selected.   In  addition,  when  setting  RAO  table  data  (for  displacement  RAOs,  wave  load  RAOs  and  wave  drift  QTFs)  you  must   specify  which  direction  the  data  applies  to.  For  example:   Select  "Vessel  Type1"      Select  Draught  "Survival  Draught"          Select  RAOs  Displacement              Select  Direction  22.5                  RAOYawAmplitude[2]  =  0.13              Select  Direction  45                  RAOYawAmplitude[2]  =  0.18  

However,  it  is  worth  pointing  out  that  situations  where  you  would  wish  to  specify  RAO  table  data  in  a  batch  script   are  rare.  It  is  much  more  likely  that  you  would  import  this  data  into  OrcaFlex  from  some  external  source  and  then   save  it  as  part  of  the  base  case  data  file.   Variable  Data  sources   Data  for  Variable  Data  sources  can  be  set  from  the  batch  script,  although  once  again  we   feel  it  is  unlikely  that  you   would  need  to  do  this  often.  The  procedure  for  setting  variable  data  sources  is  illustrated  below:   Select  Stiffness1      NumberOfRows  =  3        IndependentValue[1]  =  0      DependentValue[1]  =  0        IndependentValue[2]  =  0.2      DependentValue[2]  =  1000        IndependentValue[3]  =  0.4      DependentValue[3]  =  5000  

Note   that   IndependentValue   and  DependentValue   are   the   variable   names   for   the   X   and   Y   columns   of   the   variable   data   source.   That   is   if   you   are   setting   data   for   a   bending   stiffness   data   source   then   IndependentValue   denotes   curvature  and  DependentValue  denotes  bend  moment.   Line  Type  Wizard   The  Line  Type  Wizard  can  be  used  from  batch  script.  First  of  all  you  must  select  the  Line   Type  and  set  its  Wizard   data.  Once  this  is  complete  the  Wizard  is  invoked  using  the  InvokeWizard  command  as  illustrated  below:   Select  "Line  Type1"      WizardCalculation  =  "Homogeneous  Pipe"      PipeMaterial  =  Steel      PipeOuterDiameter  =  0.082      PipeWallThickness  =  0.005   InvokeWizard     Select  "Line  Type2"      WizardCalculation  =  "Line  with  Floats"      FloatBaseLineType  =  "Line  Type3"      FloatDiameter  =  0.80      FloatLength  =  1.2      FloatPitch  =  5.5   InvokeWizard  

91  

w

 

Automation,  Batch  Processing    

Plasticity  Wizard   The   Plasticity   Wizard   can   be   used   from   batch   script.   First   of   all   you   must   select   the   Bend   Stiffness   variable   data   source   and   set   its   Wizard   data.   Once   this   is   complete   the   Wizard   is   invoked   using   the   InvokeWizard   command   as   illustrated  below:   Select  Stiffness1      StressOD  =  0.30      StressID  =  0.27      Type  =  "Ramberg-­Osgood  curve"      E  =  230.0e6      RefStress  =  385.0e3   InvokeWizard  

Polar  Coordinates  data  on  the  All  Objects  form   The  All  Objects  data  form  allows  end  connection  data  to   be  specified  as  polar  coordinates  and  this  polar  coordinates   data  is  only  accessible  from  this  form.   The   data   appears   in   a   table  containing   all   Line,   Winch  and   Link   connections.   However,   the   data   still   belongs   to   each   individual   object   and  the   appearance   of   a   table   of  data   is  purely  presentational.   This   means   that   to   set   the   data  you   must  first  select  the  individual  Line,  Link  or  Winch  and  then  set  the  data,  as  illustrated  below:   Select  Line1      PolarR[1]  =  20.0      PolarTheta[1]  =  45.0      PolarR[2]  =  340.0      PolarTheta[2]  =  45.0   Select  Line2      PolarR[1]  =  20.0      PolarTheta[1]  =  90.0      PolarR[2]  =  340.0      PolarTheta[2]  =  90.0     Select  Winch1      PolarR[3]  =  0.0      PolarTheta[3]  =  90.0      PolarR[4]  =  10.0      PolarTheta[4]  =  90.0  

For  lines  and  links  an  index  of  1  means  End  A  and  an  index  of  2  means  End  B.  For  winches  the  index  identifies  the   winch  connection  point.   Colour  data   Drawing  colour  data  items  can  be  set  through  batch  script  in  a  variety  of  ways.  The  simplest  is  to  use  the  pre-­‐defined   colours  as  follows:   Select  "Line  Type1"      PenColour  =  Red   Select  "Line  Type2"      PenColour  =  Green  

The  full  list  of  pre-­‐defined  colours  is:  Black,  Maroon,  Green,   Olive,  Navy,  Purple,  Teal,  Gray,  Silver,  Red,  Lime,  Yellow,   Blue,  Fuchsia,  Aqua,  MoneyGreen,  SkyBlue,  Cream,  MedGray  and  White.   If  you  want  more  control  over  the  colour  then  you  can  specify  an  RGB  value  as  an  integer.  The  following  example  has   the  same  effect  as  the  previous  one:   Select  "Line  Type1"      PenColour  =  255   Select  "Line  Type2"      PenColour  =  65280  

Using  decimal  values  for  RGB  value  is  impractical.  Instead  a  neat  trick  is  to  specify  the  colour  as  a  hexadecimal  value   by  prefixing  it  with  a  $  sign  as  follows:   Select  "Line  Type1"      PenColour  =  $0000FF   Select  "Line  Type2"      PenColour  =  $00FF00  

92  

w

 

Automation,  Batch  Processing  

 

Select  "Line  Type3"      PenColour  =  $FF0000  

This  sets  the  colours  to  red,  green  and  blue  respectively.  Each  pair  of  hex  digits  controls  the  amount  of  red,  green   and   blue.   So   white   is  $FFFFFF   and   black   is   $000000.   A   value   of   $C0C0C0   gives   a   light  grey  and   $808080   produces   a   darker  grey.  

4.2.6

Handling  Script  Errors  

As   with  other   computer   programs,   OrcaFlex   batch  script   files   can   easily  contain   errors.  It   is  therefore   wise   to   check   your  script  file  for  errors  before  running  it  as  a  batch  job.  To  check  for  errors  in  your  scripts,  use  the  "Check  Files"   button   on   the   OrcaFlex   batch   form.   This   will   read   and   obey   all   the   commands   in   the   script   files   except   those   that   perform   calculations   or   write   files.   It   will   then   report   any   errors   it   finds,   including   the   line   number   on   which   the   error   occurs.   You   can   then   correct   the   problem   before   running   the   script.   Example3.txt   is   an   example   containing   various  errors,  so  to  see  how  errors  are  reported  use  the  "Check  Files"  button  with  this  example.   Warning:  

4.2.7

If  you   misspell  a  variable  name   in  an  assignment  statement  then  "Check  Files"  will  report  an  error.   But   if   you   incorrectly   specify   a   variable   name   which   is   nevertheless   valid   then   OrcaFlex   cannot   detect  the  error.  So  you  need  to  be  very  careful  that  you  use  the  correct  variable  names  for  the  data   items  that  you  want  to  change.  

Obtaining  Variable  Names  

Each  OrcaFlex  data  item  has  its  own  name  that  is  used  to  specify  it   in  a  script  file.  The  names  of  the  data  items  are   based   on   the   corresponding   labels   used   on   the   data   form.   To   find   out   the   name   of   a   data   item,   open   the   appropriate   data   form,   select   the   data   item,   and   then   open   (e.g.   by   right   click)   the   pop-­‐up   menu   and   select   the   Batch   Script   Names   command   (or   press   F7).   This   displays   the   variable   name   of   the   selected   data   item   and   you   can   select   and   copy+paste  the  name  directly  into  your  batch  script.   If   the   data   item   is  in   a   table   (or   group)   of  data   items   then   the   Batch  Script   Names   form   displays   the   names   of  all   the   data  items  in  the  table.  The  different  columns  in  the  table  each  have  their  own  names;  you  then  need  to  add  an  index   to  specify  which  row  you  want.  The   exceptions  to  this  are  the  'Connections'   data  controls  f or  Lines  and  Links,  which   consist  of  two  rows,  one  for  End  A  and  one  for  End  B.  For  these,  the  Batch  Script  Names  form  lists  the  names  for  End   A  only:  those  for  End  B  may  be  obtained  by  simply  replacing  'EndA'  with  'EndB'  in  the  name.   Finally,   note   that   Batch   Script   Names   are   not   available   for   an   empty   table   Ȃ   e.g.   if   you   want   the   names   for   the   attachments   table   on   the   line   data   form,  but  there   are   currently  no   attachments.   In  this   case   you   must   add   a   row   to   the  table  before  you  can  use  the  Batch  Script  Names  form.  

4.2.8

Automating  Script  Generation  

The  OrcaFlex  Spreadsheet  has  facilities  for  automating  the  generation  of  a   script  file  for  a  regular  set  of  cases.  To  use   this  facility  select  the  Pre-­‐processing  worksheet,  then  select  the   Script  Table  cell  and  then  click  the  Create  Batch   Scripts  command  which  can  be  found  on  the  OrcaFlex  menu  in  Excel.   The   batch   script   filename   is   specified   in   the   cell   next   to   the   Script   Table   cell.   It   is   relative   to   the   directory   containing   this   spreadsheet,   so   if   you   don't   specify   the   folder   name   then   it   will   be   created   in   the   folder   containing   this   spreadsheet.   Below  the  Script  Table  cell  is  a  table  defining  the  script,  consisting  of  3  sections:   x

First   is   one   or   more   title   rows   (shown   with   a   green   background   in   the   example   below).   Only   the   first   table   column  is   used   in   these   rows,   and   the   contents   of   those   cells   are   simply   copied   to   the   script.   The   other   columns   are   ignored.   The   title   rows   can   therefore   contain   any   comments   or   other   script   commands   that   you   want   to   appear  at  the  start  of  the  script.  The  title  rows  end  at  the  first  row  with  a  blank  cell  in  the  table's  first  column.  

x

Next   is   a   series   of   header   rows   for   the   cases   (shown   with   a   blue   background   in   the   example   below).   The   last   header  row  is  deemed  to  be  the  next  row  with  a  comment  command  (i.e.  starting  with  "//")  in  the  first  column.  

x

Finally  there  are  a  series  of  rows,  one  row  for   each  case.  The  cells  in  this  section  are   processed  from  left  to  right   on  each  row,  and  then  down  to  the  next  row,  and  each  cell  generates  script  commands  as  follows.  If  the  cell  is   empty   then   no   script   commands   are   generated.   But   if   the   cell   is   not   empty   then   all   the   (non-­‐empty)   script   commands  in  the  header   rows  in  that  column  are   written  to  the  script,  with  the  cell's  value  appended   to  the  last   header  row  command.  This  allows  different  columns  to  set  data  values  for  the  various  cases,  and  a  blank  value   leaves   the   data   item   as   it  was   in   the   previous   case.   The   cases   (and   the   whole   table)   end   at   the   next   row   that   has   a  blank  cell  in  the  table's  first  column.  

93  

w

 

Automation,  Batch  Processing    

x

Note  that  you  can  add  extra  columns  to  the  table  or  indeed  remove  columns  from  the  table.  

x

The  table  can  be  arrange  with  rows  and  columns  transposed.  To  do  this  you  must  use  the  keyword  Script  Table   Row.  An  example  of  this  alternative  approach  can  be  found  in  the  default  OrcaFlex  spreadsheet  template.  

An  example  is  shown  below:  

  Figure:  

Example  table  for  automatic  batch  script  generation  

The  script  generated  by  this  table  loads  a  base  case  from  a  file  called  "Base  Case.dat".  Because  no  path  is  specified   then  this  file  is  located  in  the  same  directory  as  the  spreadsheet.  Four  cases  are  produced  based  on  this  data  file  with   current  values  of  0.5  and  0.8  and  line  lengths  of  100  and  120.   Note:  

The   cell   containing   the   base   case   data   file   name   has   the   file   name   surrounded   by   quotes.   This   is   because  the  file  name  contains  a  space.  However,  the  quotes  are  not  needed  for  the  file  names  in   the  last  column  because  they  do  not  contain  spaces.  

The   script   is   generated   as   follows.   First   select   the   cell   containing   the   text   Script   Table.   Then   drop   down   the   OrcaFlex   menu   contained   in   the   main   Excel   menu   and   click   Create   Batch   Scripts.   When   you   do   this   you   are   presented  with  the  following  window:  

94  

w

 

Automation,  Batch  Processing  

 

  Figure:  

Automatically  generated  batch  script  

The   script   file   has   not   been   saved   yet.   You   should   check   that   the   automatically   generated   script   is   as   intended.   Should  you  wish  to,  you  can  modify  the  script  file  name  at  this  point.   If  there  is  a  problem  with  the  script  you  can  click  the  Close  button  and  correct  the  script  table.   Save  button  

Saves  the  script  file.   Save  and  Run  button  

Saves  the  script  file  and  then  processes  it.   If  the  script  has  any   Run  commands  then   OrcaFlex  is  loaded  and  the  script  is  processed  by  the  standard  OrcaFlex   batch  form.  Otherwise  the  script  is  processed  from  within  Excel  Ȃ  progress  is  reported  on  the  Excel  status  bar.   Save,  Run  and  Submit  button  

Saves   the   script   file   and   then   processes   it   within   Excel.   Each   data   file   saved   by   the   script   is   then   submitted   to   Distributed  OrcaFlex  which  runs  and  saves  the  simulation  file.   Note:  

The   Save,   Run   and   Submit   button   is   only   available   if   Distributed   OrcaFlex   is   installed   on   your   machine.  In  addition,  it  cannot  be  used  if  the  script  contains  any   Run  commands.  

95  

w

 

Automation,  Batch  Processing    

Multiple  tables   You   can   have   multiple   script   tables   w ithin   a   workbook.   To   create   all   the   batch   scripts   in   one   operation   select   all   the   script  tables  and  then  click  Create  Batch  Scripts  in  the  OrcaFlex  menu.  

4.2.9

Automating  Text  Data  File  Generation  

The  OrcaFlex   Spreadsheet  has   facilities   for   automating   the   generation   of  text   data   files  for   a   regular   set   of   cases.  To   use  this  facility  select  the  Pre-­‐processing  worksheet,  then  select  the  Text  Data  Files  cell  and  then  click  the  Create   Text  Data  Files  command  which  can  be  found  on  the  OrcaFlex  menu  in  Excel.   The   basic   idea   is   very   similar   to   the   facility   for   automating   generation   of   batch   script   files.   An   example   table   is   shown  below:  

  Figure:  

Example  table  for  automatic  text  data  file  generation  

The   cell   containing   "Text   Data   Files",   highlighted   in   yellow,   is   known   as   the   anchor   cell.   The   text   data   files   are   generated   based   on   the   template   file   specified   the   in   cell   immediately   to   the   right   of   the   anchor   cell.   In   this   example   the  template  file  might  look  like  this:   BaseFile:  Base  Case.dat   Environment:      WaveDirection:  %direction      WaveHeight:  %height   Line1:      Length[1]:  %length   The  row  immediately  below  the  anchor  cell,  highlighted  in  blue,  contains   variable   names.  You  are  free  to  choose   these  names  as  you  please.  We  have  adopted  a  convention  that  the  variable  names  begin  with  a  percentage  sign  (%).   Although   you   do   not   need   to   follow   this   convention,   doing   so   will   have   the   benefit   of   making   the   variable   names   stand  out.   The  rows  beneath  the  variable  names  row  are  known  as  the   value  rows.  Each  row  defines  a  single  text  data  file.  The   text  data  file  is  generated  by  starting  from  the  template  file  and  then  replacing  each  variable  name,  in  turn,  by  the   value  specified  in  the  table.   The  %filename  variable   name  is  compulsory,  that  is   it   must  be  included  as  one  of   the   variable   names.   It  specifies   the   name   of   the   generated   text   data   file.  Relative   paths  can   be   used  for   the  template  file   name  and  the  output  file  names.   The   extent   of   the   table   is   determined   by   the   presence   of   empty   cells.   So   the   variable   names   row   ends   at   the   first   empty  cell.  Likewise  the  value  rows  end  at  the  first  empty  cell  in  the  column  beneath  the  anchor  cell.   The   example  above   produces  8  text  data  files  as  its  output,  named  Case01.yml,  Case02.yml,   etc.  This  first  of  these   looks  like  this:   BaseFile:  Base  Case.dat   Environment:      WaveDirection:  0      WaveHeight:  8   Line1:      Length[1]:  100   Choosing  variable  names  

It  is  clearly  important  that  you  choose  unique  variable  names.  However,  there  is  a  further  subtlety  which  can  arise   when   one   variable   name   is   a   sub-­‐string   of  another.   For   example,   consider   the   variable   names   %x1   and   %x10.   When  

96  

w

 

Automation,  Post-­‐processing  

 

occurrences   of  %x1  in   the   template   file   are   replaced   by   their   actual   values,   the   first   3   characters   of  any  occurrences   of  %x10  will  also  be  detected.   Such   ambiguities   seldom   arise,   but   if   you   are   affected   then   you   can   extend   the   naming   convention   to   include   a   trailing   %  sign.  In  the  example  given  above,  the  variable  names   become  %x1%  and   %x10%  and  clearly  the  problem   does  not  arise.   Multiple  tables  

You   can   have   multiple   tables   within   a   workbook.   To   process   all   the   tables   in   one   operation,   select   all   the   tables   and   then  click  the  Create  Text  Data  Files  command.   Benefits  over  script  tables  

The  text  data  file  approach  to  load  case  file  generation  described  above  is  very  similar  to  the  approach  using   batch   script  files.  The  choice  of  which  to  use  is  largely  one  of  personal  preference.  If  you  are  already  familiar  with  batch   script  and  not  yet  familiar  with  text  data  files  then  it  may  prove  easier  to  continue  using  batch  script.   There  is  one  significant  advantage  of  using  text  data  files  which  is  that  it  avoids  duplication  of  the  OrcaFlex  model   data.   Consider   the   following   typical   sequence   of   actions   when   using   batch   script   where   we   assume   that   the   basic   model  and  scripts  are  already  in  place:   1.

Modify  the  single  base  model,  represented  by  an  OrcaFlex  data  file.  

2.

Run  the  batch  script  (or  scripts)  that  generate  all  the  load  case  data  files.  

3.

Run  the  simulations  for  all  load  case  data  files.  

If  text  data  files  are  used,  as  described  above,  then  step  2  is  not  required.  This  is  because  the  load  case  text  data  files   contain  a  reference  to  the  base  model  rather  than  containing  a  copy  as  is  the  case  when  using  batch  script.   This   is   a   relatively   minor   advantage   but   it   does   reduce   the   likelihood   of   mistakenly   forgetting   to   carry   out   step   2   when   using   batch   script   files.   In   addition,   more   complex   analyses   can   lead   to   your   load   cases   being   defined   by   multiple  script  files  which  have  to  be  executed  in  a  particular  order.  Using  the  text  data  file  approach  means  that  this   complexity   is   dealt   with   just   once   when   setting   up   the   text   data   files,   as   opposed   to   every   time   a   modification   to  the   base  model  is  made.  

4.3

POST-­‐PROCESSING  

4.3.1

Introduction  

OrcaFlex   users   often   use   spreadsheets   to   post-­‐process   their   OrcaFlex   results.   This   can   be   done   manually   by   transferring   the   results   from   OrcaFlex   into  the   spreadsheet   using   copy   +   paste.   However,   this   is   laborious   and   error   prone  if  a  lot  of  results  need  transferring,  so  we  have  developed  special  facilities  to  automate  the  process.   This  automation  is  done  using  an  Excel  spreadsheet  that  has  facilities  for  automatic  extraction  of  specified  results   from  one  or  more  OrcaFlex  files  into  nominated  cells  in  the  spreadsheet.  You  can  then  use  the  normal  spreadsheet   facilities  to  calculate  other  post-­‐processed  results  from  those  OrcaFlex  results.   Note:  

The  OrcaFlex  spreadsheet  works  with  Excel  2000  or  later  and  requires  OrcaFlex  to  be  installed  on   the  machine.  

Creating  OrcaFlex  Spreadsheets   You  can  create  OrcaFlex  spreadsheets  from  Excel  templates  that  are  supplied  with  OrcaFlex.  You  should  base  your   own   OrcaFlex   spreadsheets   on   this   template,   which   is   installed   in   the   OrcaFlex   installation   directory   when   you   install  OrcaFlex  on  your  machine.  To  create  your  own  OrcaFlex  spreadsheet,  open  the  Windows  Start  menu,  select   Programs  |  Orcina  Software  and  then  select  New  OrcaFlex  Spreadsheet.  This  shortcut  creates  a  new  spreadsheet   based  on  the  template.   Note:  

Different  versions  of  Excel  require  different  versions  of  the  OrcaFlex  Spreadsheet.  When  installing   OrcaFlex   you   are   asked   which   version   of   Excel   you   are   using.   Both   spreadsheet   templates   are   installed  and  a  single  Start  Menu  shortcut  to  the  appropriate  one  is  created.  

Before   you   try   to   use   the   new   spreadsheet   you   need   to   save   it   to   a   file;   it   can   be   given   any   valid   file   name.   It   is   usually   most   convenient   to   save   it   to   the   directory   containing   the   OrcaFlex   files   from   which   you   want   to   extract   results.  You  can  then  specify  the   names   of  those   files  in  the  spreadsheet  using   relative  paths.  Using   relative  paths   makes  it  easier  to  rename  the  directory  or  move  the  spreadsheet  and  OrcaFlex  files  to  some  other  directory.  

97  

w

 

System  Modelling:  Data  and  Results,  Vessels    

vertex 3

edge joining 3 to 5

vertex 5

z (heave)

yaw

y (sway) pitch V

x (surge)

roll

  Figure:  

Vessel  Model  

Each   vessel   has   a   Vessel   Type   that   determines   its   RAO   and   drawing   data.   To   illustrate   this,   consider   a   model   of   a   pipe  being  towed   by   two  identical  tugs.  This  is  modelled   by  creating  a  vessel  type  called  'Tug'  and   then  creating  two   vessels,   each  of  type  'Tug'.   The  drawing  data  (defining   the  tug  outline)  are  data   of   the   Tug  vessel   type,  since  they   apply  to  both  tugs.  Similarly,  the  RAOs  are  data  of  the  vessel  type,  since  again  they  are  the  same  for  both  tugs.  On  the   other   hand   the   two   tugs   differ   in   their   positions   and   prescribed   motion,   so   these   are   properties   of   the   individual   vessel  objects.   Note:  

The  vessel  also  has  extra  drawing  data  Ȃ  this  is  to  allow  you  to  set  up  vessel-­‐specific  drawing.  For   example   the   lead   tug   may   have   a   special   tow-­‐point   fitting   that   you  want   to   draw.   When   the  vessel   is   drawn,   OrcaFlex   first   draws   the   vessel   type   wire   frame   and   then   draws   the   vessel   wire   frame.   These   two   wire   frames   can   have   different   colours,   so   you   can   highlight   application-­‐specific   drawing.  

The  vessel  is  defined  relative  to  a  right-­‐handed  system  of  local  vessel  axes  Vxyz,  where:   x

V  is  the  vessel  origin  for  this  vessel  type.  This  is  chosen  by  the  user  when  the  vessel  type  is  set  up.  However  note   that   if   you   specify   that   the   vessel   type   has   symmetry   then   the   vessel   origin   must   be   placed   on   the   plane   of   symmetry  or  at  the  centre  of  circular  symmetry;  see  Vessel  Types:  Conventions  for  details.  

x

Vx,  Vy   and   Vz   must   be   the   directions   of  surge,   sway   and   heave,   respectively,   for   this   vessel   type.   Note   that   these   directions  must  therefore  be  the  directions  to  which  the  RAOs  apply.  

Points  on  the  vessel,  for  example  where  cables  or  risers  are  connected,  are  then  defined  relative  to  these  vessel  axes.   These  points   then  move  with  those  axes  as  the  vessel   moves  and   rotates  relative   to  the  global  axes,  and  OrcaFlex   calculates  these  motions  automatically.   The  vessel  is  drawn,  in  3D  views  of  the  model,  as  a  "wire  frame"  of  user-­‐specified  vertices  and  edges.  This  allows  a   simple  visual  check  that  amplitudes,  phases  etc.  are  consistent  with  the  applied  wave.  The  vessel  wire  frame  can  also   be   used   to  do  a   visual   check  for   interference   between   lines   and   vessel   structure.   As   with   all   points   on   the   vessel,   the   vertices  are  defined  relative  to  the  vessel  axes  Vxyz.  

6.7.1

Vessel  Data  

Name  

Used  to  refer  to  the  Vessel.  

230  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

Type  

Specifies  the  Vessel  Type.  The  Vessel  Types  button  allows  you  to  view  and  edit  the  Vessel  Type  Data.   Draught  

Specifies  which  set  of  RAOs  to  use  from  the  specified  vessel  type.  See  Draughts.   Length  

Specifies  the  length  of  this  vessel.  The  default  value  '~'  means  that  this  vessel  is  the  same  length  as  the  vessel  type.  If   you   specify   a   length   that   differs   from   the   vessel   type   length,   then   OrcaFlex   will   scale   all   the   vessel   type's   data   to   allow   for   the   scaling   factor   VesselLength/VesselTypeLength.   This   is   useful   if   you   have   data   for   a   70m   ship,   for   example,  but  want  to  use  a  50m  ship  that  is  otherwise  very  similar.   The  scaling  is  done  using  Froude  scaling  (see   Rawson  and  Tupper).  Froude  scaling  scales  all  items  of  data  by  a  factor   that   depends   on   the   units   of   that   data.   If   R   =   ratio   of   vessel   length   to   vessel   type   length,   then   the   scaling   factor   applied  is  as  follows:   x

All  lengths  are  scaled  by  R.  

x

All  masses  (and  added  masses)  are  scaled  by  R3.  

x

All  times  are  scaled  by  R0.5.  

Data   items   with   other   units   are   scaled   by   writing   the   units   in   terms   of   the   fundamental   units   of   mass,   length   and   time,   and   then   applying   the   above   factors.   For   example   force   data   has   units   equivalent   to   Mass*Length/Time2,   so   force   data   is   scaled   by   (R3).(R)   /   ((R0.5)2)   =   R3.   Dimensionless   items   such   as   translational   RAOs   (surge,   sway,   heave),   QTFs  and   phase  angles  are  unchanged  by  the  scaling  process,  but   note  that  Froude  scaling  does  apply  to  the   periods   (or   frequencies)   specified  for   RAO   and  QTF   data.   Note   also   that   rotational   RAOs   (roll,   pitch,  yaw)   are   often   given  in   dimensional  terms  (degrees  per  metre)  and  these  also  scale.   These  scaling  rules  are  the  same  as  those  used  in  deriving  full  scale  ship  performance  from  physical  model  tests,  and   are  correct  if  the  vessel  is  a  perfect  scaled  replica  of  the  vessel  type  in  all  respects.   Warnings:   If  the  vessel  does  not  scale  uniformly  in  all  dimensions,  then  this  type  of  scaling  introduces  errors   and  should  not  normally  be  used.  Instead,  accurate  data  specific  to  this  vessel  should  be  obtained.    

However,  for  ships  in  head  and  stern  seas  the  RAO  scaling  errors  may  be  acceptable,  since  the  RAOs   for  these  wave  directions  depend  mainly  on  vessel  length.  For  other  cases  the  RAO  scaling  is  likely   to  be  poor,  so  OrcaFlex  issues  a  warning  if  scaling  is  used  and  the  wave  direction  is  not  close  to  a   head  or  stern  sea.  

Initial  Position  and  Orientation  

These  specify  the  vessel's  static  position  relative  to  the  global  axes.  The  Initial  Position  defines  the  position  of  the   vessel   origin  V.   The  Initial   Orientation  defines   the   orientation   of   the   vessel   axes   Vxyz   as   three   rotations,  Heading,   Trim  and  Heel.  The  static  orientation  of  Vxyz  is  that  which  results  from  starting  with  Vxyz  aligned  with  the  global   axes  and  applying  the  Heading  rotation  about  Vz,  then  the  Trim  rotation  about  Vy  and  finally  the  Heel  rotation  about   Vx.   If   the   vessel   is   not   included   in   the   static   analysis   then   this   Initial   Position   is   taken   to   be   the   static   position   of   the   vessel.   If   the   vessel   is   included   in   the   static   analysis,   then   this   Initial   Position   is   used   as   an   initial   estimate   of   the   vessel   position   and   the   statics   calculation   will   move   the   vessel   from   this   position   iteratively   until   an   equilibrium   position  is  found.   Note:  

The  vessel  Z  coordinate  can  only  be  changed  by  editing  on  the  vessel  data  form.  Dragging  in  the  Z   direction  with  the  mouse  is  prevented.  

Warning:  

If   you   have  included   any   harmonic  motion  on   the   vessel   (see  Harmonic   Motion)   then   the   phases   of   the   harmonic   motions   will   normally   depend   on   the   vessel   Initial   Position,   so   if   you   change   the   Initial  Position  you  may  need  to  change  the  harmonic  motion  phases  accordingly.  

Calculation  Data   The   following   settings   (on   the   Calculation   page   on   the   vessel   data   form)   control   how  the   vessel's   static   position   and   dynamic  motion  are  determined.  

231  

w

 

System  Modelling:  Data  and  Results,  Vessels    

Included  in  Static  Analysis   You   can   control   whether   the   OrcaFlex   static   analysis   calculates   the   static   equilibrium   position   of   the   vessel,   or   simply  places  the  vessel  in  the  user-­‐specified  initial  position.   OrcaFlex   first   places   the   vessel   at   the   initial   position   and   orientation   specified   by   the   user.   If   Included   in   Static   Analysis   is   set   to   None   then   OrcaFlex   leaves   the   vessel   in   this   user-­‐specified   position.   This   is   not   necessarily   an   equilibrium  position.   If  Included   in   Static  Analysis  is  set   to  3   DOF  then   OrcaFlex   starts   from   the   user-­‐specified  position   and   adjusts   the   vessel's   X,   Y   and   Heading   until   an   equilibrium   position   is   reached.   Note   that   only   these   3   free   degrees   of   freedom   of   the   vessel   (X,   Y   and   Heading)   are   included   in   the   calculation.   The   other   three   degrees   of  freedom   (Z,   Heel   and   Trim)   are  assumed  to  be  constrained  and  so  are  left  at  the  values  specified  by  the  user.  This  means  that  the  Z  component  of   resultant   force   may   be   non-­‐zero   in   the   equilibrium   position.   Likewise   the   resultant   Heel   and   Trim   moments   may   be   non-­‐zero  in  the  equilibrium  position.   If  Included  in  Static  Analysis  is  set  to  6  DOF  then  OrcaFlex  starts  from  the  user-­‐specified  position  and  adjusts  all   six  degrees  of  freedom  until  an  equilibrium  position  is  reached.  In  this  case,  all  of  the  X,  Y,  Z  forces  and  Heel,  Trim,   Heading  moments  will  be  zero  at  the  equilibrium  position.   The   loads   on   the   vessel   to   be   taken   into   account   in   calculating   the   equilibrium   position   are   determined   by   the   Included  Effects,  with  the  provisos  that,  in  statics,  first  order  wave  load  RAOs  will  have  no  effect  and  second  order   wave  drift  load  will  only  have  an  effect  through  the  mean  wave  drift  component.   Note:  

If  multiple  statics  are  being  performed  on  the  vessel  then  no  equilibrium  calculation  is  performed   on   the   vessel   and   its   placement   is   determined   by   the   multiple   statics   data.   Other   vessels   in   the   model  are  included  in  the  static  analysis  as  specified  by  their  own  data.  

Dynamic  Analysis   The   motion   of   a   vessel   during   the   dynamic   analysis  can   be   specified   in   a   variety   of  ways.   OrcaFlex   allows   the   vessel   motion  to  be  made  up  of  two  parts,  called  the   Primary  motion  and  the  Superimposed  motion.  Broadly,  the  Primary   motion  is  aimed  at  modelling  the  steady  or  low  frequency  motion  of  the  vessel,  whereas  the  Superimposed  motion  is   aimed  at  modelling  the  higher  frequency  motion,  such  as  that  generated  by  waves.   As  an  example,  consider  a  ship  being   driven  under  power  along  a  specified  course.  In  the  absence  of  waves  it  moves   steadily   along   its   course   and   this   would   be   modelled   by   the   Primary   motion.   But   when   waves   are   present   the   primary   motion   is   augmented   by   wave-­‐generated   motion   that   would   often   be   modelled   in   OrcaFlex   as   Superimposed  motion  specified  by  RAOs.  OrcaFlex  superimposes  this  latter  motion  on  the  primary  motion  to  give   the  total  combined  motion  of  the  vessel.   You  can  specify  the  Primary  and  Superimposed  motions  in  a  number  of  ways,  as  follows.  See  examples  below  of  how   these  options  can  be  used.   Primary  Motion  

The  Primary  motion  determines  what  OrcaFlex  refers  to  as  the  primary  position  of  the  vessel.  It  can  be  one  of  the   following  options.   x

None.   In   this   option   there   is   no   primary   motion   and   the   primary   position   of   the   vessel   remains   fixed   at   the   position  determined  by  the  static  analysis.  

x

Prescribed.  This  option  allows  you  to  drive  the  vessel  around  the  sea  surface,  for  example  to  model  the  vessel   moving   station   during   the   simulation.   The   vessel's   speed   and   course   is   specified   using   the   data   on   the   Prescribed  Motion  page  on  the  vessel  data  form.  

x

Calculated   (3   DOF).   In   this  option  OrcaFlex   calculates   the   vessel   primary  motion   in   only  3   degrees   of   freedom   (surge,  sway,  yaw)  based  on  the  included  loads  plus  loads  from  any  lines  or  other  objects  that  are  attached  to   the  vessel.  There  is  no  primary  motion  in  the  other  3  degrees  of  freedom  (heave,  roll,  pitch).  The   added  mass   and  damping  matrices  of  the  vessel  type  must  be  specified,  plus  the  data  for  all  the   included  loads.  

x

Calculated  (6  DOF).  In  this  option  OrcaFlex  calculates  the  vessel  motion  in  all  6  degrees  of  freedom,  based  on   the  included  loads,  plus  loads  from  any  lines  or  other  objects  that  are  attached  to  the  vessel.  The   added  mass,   damping,   stiffness   and   equilibrium   position   of   the   vessel   type   must   be   specified,   plus   the   data   for   all   the   included  loads.  

232  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

x

Time   History.   In   this   option   the   user   specifies   the   primary   motion   in   a   time   history   file   that   defines,   as   a   function   of   time,   the   vessel   Primary   X,   Primary   Y,   Primary   Z,   Primary   Rotation   1,   Primary   Rotation   2   and   Primary  Rotation  3.  See  the  vessel's  Time  History  data.  

Superimposed  Motion  

The  Superimposed  motion   is  applied  as  an  offset  from  the  position  given  by   the  primary  motion.  It  can   be  one  of  the   following  options.   x

None.  In  this  option  there  is  no  offset  and  the  vessel  position  is  equal  to  the  primary  position  at  all  times.  

x

Displacement  RAOs  +  Harmonic.  In  this  option  the  vessel's  position  oscillates  harmonically  about  the  primary   position.   The   harmonically  varying   offset   comes   from   two   sources.   Firstly,  if   waves   are   present   and   you   specify   non-­‐zero   displacement   RAOs   for   the   vessel   type,   then   the   offset   will   include   the   wave-­‐generated   harmonic   motions   specified   by   those   RAOs.   Secondly,   the   vessel's   superimposed   offset   also   includes   any   harmonic   motions  that  you  specify  on  the  Harmonic  Motions  page  on  the  vessel  data  form.  

x

Time   History.   In   this   option   the   user   specifies   the   offset   in   a   time   history   file   that   defines   the   vessel   Surge,   Sway,  Heave,  Roll,  Pitch  and  Yaw  as  a  function  of  time.  See  the  vessel's  Time  History  data.  

Typical  Examples  of  Primary  and  Superimposed  Motion  

Here   are   some   typical   examples   of   how   primary   and   superimposed   motion   can   be   combined   to   achieve   various   things.   x

Primary   motion   set   to  None   and   superimposed   motion   set   to   RAOs   +   Harmonic   can   be   used   to   model   a   simple   steady  (e.g.  moored)  vessel  whose  motion  is  due  to  the  1st  order  wave  effects  only.  

x

Prescribed   or   Time   History   primary   motion   can   be   used   to   model   a   vessel   being   driven   over   a   predetermined   course,  and  the  superimposed  motion  set  to  RAOs  +  Harmonic  to  model  the  1st  order  wave  induced  motion.  

x

Calculated   (3   DOF)   primary   motion   can   be   used   to   model   the   low   frequency   slow   drift   of   a   vessel   in   the   horizontal   plane   (due   to   2nd   order  wave   loads,  wind   and   current   drag,   etc.),   with  the  superimposed   motion  set   to  RAOs  +  Harmonic  to  model  the  superimposed  1st  order  wave  induced  motion.  

x

Calculated  (6  DOF)  primary  motion  can  be  used  to  model  the  whole  motion  of  a  vessel,   optionally  including  1st   and   2nd   order   wave   loads,   current   and   wind   drag   loads,   applied   loads,   and   always   including   loads   from   any   attached   lines   or   other   objects.   In  this   case   the  superimposed   motion  would   normally   be   None,   since   all   motion   has  already  been  accounted  for  in  the  primary  motion,  and  the  displacement  RAO  data  would  therefore  not  be   used.  

x

Time   History   primary   motion   (or   Time   History   superimposed   motion   but   not   usually   both)   can   be   used   to   completely  specify  the  motion  of  a  vessel.  

Included  Effects   You  can  choose  which  vessel  loads  are  included  and  which  are  ignored  by  ticking  the  corresponding  checkboxes.  If  a   box  is  checked  then  that  load  will  be  calculated  and  applied  to  the  vessel,  and  its  value  will  be  available  as  a  result.  If   a  box  is  not  checked,  then  that  load  will  not  be  calculated  and  it  will  be  taken  as  zero.   Note:  

Loads  which  are   included  will   only   affect   the  vessel   static  equilibrium   position  if  Included   in   Static   Analysis  is  set  to  other  than  None,  and  will  only  affect  the  vessel  motion  if  the   primary  motion  is  set   to  one  of  the  Calculated  options.  

Prescribed  Motion   The  prescribed  motion  data  only  applies  if  the  vessel's  Primary  Motion  is  set  to  Prescribed.  It  enables  you  to  drive   the  vessel  around  the  sea  surface  along  a  predetermined  path,  by  specifying  how  the  vessel's  primary  position  and   heading  change  during  the  simulation.   The  vessel  is  driven   by  specifying,  for   each  stage   of  the  simulation,  the  velocity  (speed  and  direction)  of  the   primary   position  and  the  rate  of  change  of  the  heading.   Warning:  

The  Prescribed  Motion  facility  can  cause  discontinuities  of  velocity  at  stage  boundaries  and  these   may  cause  transients  in  the  system.  OrcaFlex  issues  warnings  at  the  start  of  the  simulation  if  the   data  specifies  a  discontinuous  velocity.  

233  

w

 

System  Modelling:  Data  and  Results,  Vessels    

Speed  of  Motion  

For  each  simulation  stage  the  speed  of  the  motion  can  be  specified  as  either  a   Constant  Speed  or  a  Speed  Change.   If  Constant  Speed  is  selected  then  that  speed  applies  throughout  the  stage.   If  Speed  Change  is  selected  then  the  speed  is  incremented  linearly  by  the  specified  amount  during  the  stage.  For   example,   an   increment  of   1m/s   during   a   stage  of   length   10   seconds   causes   an   acceleration  of   0.1m/s 2  to  be   applied   throughout  the  stage,  so  that  if  the  vessel  starts  at  rest  then  at  the  start  of  the  next  stage  the  vessel  is  travelling  at   1m/s.   Note:  

Negative  speeds  can  be  specified.  

Direction  of  Motion  

For   each   simulation   stage   the   direction   of  the   motion   can   be   specified   as   either   relative   to   the  Global   X  direction   or   relative  to  the  Vessel  heading.   The   direction   value   is  interpreted   as   an  azimuth.   So,   a   value   of   90°   relative   to   the   Global   X   direction   would   result  in   the  motion  being  in  the  global  Y  direction.  Likewise,  if  the  vessel's   heading  was  130°  and  the  direction  of  motion  was   50°   relative   to   the   Vessel   heading   then   the   motion   would   be   an   azimuth  of   130°   +   50°   =   180°   relative   to   global,   that   is  in  the  global  -­‐X  direction.   Note:  

The  direction  of  motion  is  not  restricted  to  be  in  the  same  direction  as  the  vessel  heading.  

Rate  of  Turn  

In  addition  to  varying  the  velocity  of  the  mean  position,  you  can  specify  a   Rate  of  Turn  for  each  stage.  This  is  the   angle  change  per  second  to  be  applied  to  the  vessel's  heading  throughout  the  stage.   If  the  Direction  of  Motion  is  specified   relative  to  the  Vessel  heading  and  the  Rate  of  Turn  is  non-­‐zero,  then  the  vessel   velocity   direction   varies   during   the   stage.   In   addition,   if  the   direction   is   0°   relative   to   Vessel   heading   then   the   vessel   velocity,  whilst  varying,  is  always  in  the  Vessel  heading  direction.   If   the   direction   is  specified   relative  to   the   Global   X   direction  then   the   vessel   velocity  direction   is  not   affected   by   the   Rate  of  Turn.  

Harmonic  Motion   The  Harmonic  Motion  page  (on  the  vessel  data  form)  only  applies  if  the  vessel's  superimposed  motion  is  set  to  RAOs   +  Harmonic.  It  allows  you  to  specify  a  number  of  harmonic  motions  of  the  vessel.   The   harmonic   motions   are   in   addition   to   any  wave-­‐generated   motion   specified   by   the  RAO   data,   so   if  you   only   want   the  wave-­‐generated  motion  then  you  should  set  the  number  of  harmonic  motions  to  zero.   Each  harmonic  motion  is  a  single-­‐period  sinusoidal  motion  of  the  vessel,  specified  by  giving:   x

the  Period  of  the  harmonic  motion;  this  applies  to  all  6  degrees  of  freedom,  

x

the   Amplitude   and   Phase   of  the   motion   for   each   of  the   6   degrees   of  freedom   of   the   vessel.   If   you   are   modelling   slow   drift,   then   note   that   slow   drift  normally   only   applies   to   surge,   sway   and   yaw,   in   which   case   the   amplitudes   for  heave,  roll  and  pitch  should  be  set  to  zero.   Note:  

The  harmonic  motion  amplitudes  (unlike  the   RAO  responses  of  the  vessel)  are  not  specified  relative   to   a   wave   amplitude   Ȃ   they   are   specified   directly   in   length   units   (for   surge,   sway   and   heave)   or   degrees  (for  roll,  pitch  and  yaw).  

Similarly,   the   phases   are   not   specified   relative   to   the   phase   of   a   wave  Ȃ   they   are   the   phase   lags   from   the   global   time   origin  T=0  until  the  maximum  harmonic  motion  occurs.  More  precisely,  the  phase  that  should  be  specified  for   the   harmonic  motion  is  given  by   360  ×  ((Tmax  /  P)  mod  1)   where  P  is  the  period  of  the  harmonic  motion  and  Tmax  is  the   global  time  at  which  you  want  the  maximum  of  the   motion  to  occur.  

234  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

Warning:  

Harmonic  motions  can  be  used  to  model  pre-­‐calculated  vessel  slow  drift.  If  you  do  this,  then  if  you   move   the   vessel's   Initial   Position   in   the   wave   direction,   or   if   you   change   the   data   for   the   waves   (other  than  changing  the  simulation  time  origin),  then  you  will  normally  also  then  have  to  adjust   the   phases   of   the   slow   drift.   This   is   because   such   changes   affect   the   global   time   at   which   a   particular   part   of   the   wave   train  will   reach   the   vessel   and   hence  will   also   affect   the   global   time   at   which  maximum  slow  drift  motion  is  achieved.  

Time  History   The   Time   History   pages   (on   the   vessel   data   form)   only   apply   if   the   vessel's   primary   or   superimposed   motion,   or   both,  are  set  to  Time  History.  It  allows  you  to  specify  the  motion  by  giving  a  time  history  file.  To  do  this:   x

On  the  Calculation  page  set  the  primary  motion  or  superimposed  motion  data  item  (or  both)  to  Time  History.  

x

Create  a  tab-­‐delimited  text  file  containing  the  time  history  motion  you  want,  as  described  below.  

x

On  the  appropriate  Time  History  page,  setup  the  time  history  data  as  described  in  Data  in  Time  History  Files.  

Contents  of  Time  History  File   The  time  history  file  must  contain  a  time  column  and  columns  for  all  6  degrees  of  freedom  of  the  vessel.  For  primary   time  history   motion  these  are  Primary   X,  Primary   Y,  Primary  Z,  Primary  Rotation  1,   Primary  Rotation   2  and  Primary   Rotation  3,  measured  relative  to  the  global  axes.   For  superimposed  time  history  motion  the  degrees  of  freedom  that  must  be  specified  are  Surge,  Sway,  Heave,  Roll,   Pitch  and  Yaw.  They  are  measured  relative   to  the  primary  position  of  the  vessel,  as  specified   by  the  vessel's  primary   motion.   For  details  of  how  rotations  are  applied,  see  order  of  application  of  rotations.   The   time   values   in   a   vessel   time   history  file   need   not  be  equally  spaced.   The   units   used   for   all   the   columns   must  be   the  same  as  those  used  in  the  OrcaFlex  model,  so  the  time  values  must  be  in  seconds  and  angles  in  degrees.   For  further  details  of  the  file  format  see  Time  History  Files.   Notes:  

If  there  is  any  wave-­‐generated  motion  present  in  a  vessel's  time  history  motion  then  the  OrcaFlex   wave  data  needs  to  match  the  wave  that  generated  that  motion.  If  you  have  suitable  data  for  the   wave  elevation  then  you  can   use  that  to   specify  the  wave  by  time  history.  This  can  be   done  either  in   a  separate  time  history  file  for  the  wave  or  else  in  an  extra  column  in  the  vessel's  time  history  file.  

 

The   position   and   velocity   specified   by   a   time   history   file   for   the   start   of   the   simulation   (i.e.   for   SimulationTime   =   -­‐BuildUpDuration)   will   not,   in   general,   match   the   static   state   from   which   OrcaFlex  starts  the  simulation.  To  handle  this  OrcaFlex  uses  ramping  during  the  build-­‐up  stage  to   smooth  the  transition  from   the   static   state   to   the   position   and   motion   specified   in   the   time   history   file.  

Applied  Loads   You  can  optionally  include  applied  loads  on  a  vessel.   You   can   apply   to   the   vessel   external   Global   Loads   that   do   not   rotate   if   the   vessel   rotates.   These   are   specified   by   giving   the   components   of   Applied   Force   and   Applied   Moment   relative   to   global   axes.   These   components   can   be   constant,  vary  with  simulation  time  or  be  given  by  an  external  function.  If  the  vessel  rotates  then  the  loads  do  not   rotate  with  it.   In   addition,   you   can   specify   external   Local  Loads   that   do   rotate   with   the   vessel.   These   are   specified   by   giving   the   components   of  Applied   Force   and   Applied   Moment   relative   to   vessel   axes.   Again   these   components   c an   be   constant,   vary  with  simulation  time  or  be  given  by  an  external  function.  If  the  vessel  rotates  then  the  loads  do  rotate  with  it.   These  are  suitable  for  modelling  thrusters,  for  example.   In  both  cases  the  Point  of  Application  of  the  load  is  specified  by  giving  its  x,y,z  coordinates  relative  to  vessel  axes.   Note:  

Applied   loads   will   only   affect   vessel   static   position   if   the   corresponding   degree   of   freedom   is   included   in   the   static   analysis,   and   will   only   affect   the   motion   if  the  Primary   Motion   is   set   to   one   of   the  calculated  options  which  includes  the  degree  of  freedom.  

235  

w

 

System  Modelling:  Data  and  Results,  Vessels    

Multiple  Statics   The  offsets  for   multiple  statics  calculations  are  specified  here.  Offsets  are  from  the  vessel's  initial  position  and  are   specified  by  giving  a  range  of  azimuth  and  offset  values.  For  example:  

  The   Azimuths   table   determines   which  directions   are   to   be   analysed.  The  Offsets   table   specifies   how  far   in   the   given   direction  the  vessel  is  to  be  placed.  With  the  above  data,  the  offsets  analysed  by  the  multiple  statics  calculation  are   as  illustrated  by  the  dots  in  the  diagram  below:  

Y 90 deg

X

45 deg

135 deg

180 deg

0 deg 0m

20m 40m 60m 80m 100m

Vessel Initial Position   Figure:  

Example  Offsets  

A   diagram   showing   the   selected   offsets   is  drawn   on   the   Vessel   Offsets   data   form,   to   help   visualise   which   offsets   will   be  analysed.  

Drawing   Vessels  are  drawn  as  wire  frames   defined   in  the   data  as  a  set  of   Vertices  and  Edges.  The   Vertices  are  defined  by   giving  their  coordinates  relative  to  the  vessel  axes  Vxyz.  The  Edges  are  lines  drawn  between  two  vertices.   For   shaded   graphics   views,   by   default,   the   vessel   is   drawn   using   a   solid,   filled-­‐in   shape   based   on   the   vertices   and   edges.  As  an  alternative  you  can  use  the  vertices  and  edges  to  define  a  frame  like  structure.  If  the  edge  diameter  is   '~'  then  that  edge  will  be  used  to  build  a  filled  in  shape,  otherwise  that  edge  is  drawn  as  a  cylinder  with  the  specified   diameter.  Note  that  you  can  use  a  mixture  of  edge  diameters  (some  defined,  some  set  to  '~')  to  combine  both  filled   in  and  framework  shapes.   You   can   define   wire   frame   drawing   data   in   two   places   Ȃ   for   the   vessel   and   also   for   its   vessel   type.   The   vessel   is   drawn  by  first  drawing  a  wire  frame  based  on  the  vertices,  edges  and  pen  specified  for  its  vessel  type  (see  the  vessel  

236  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

types   data  form).   Then  a   further  vessel-­‐specific  wire  frame  may  be  drawn,  using  any  vertices,   edges  and  pen  that   you  specify  on  the  vessel's  data  form.   This  allows   you   to   specify   a   wire   frame   drawing   of   the  basic   vessel   type,   and   then   optionally  add   to   it   (possibly  in   a   different  colour)  a  wire  frame  drawing  of  some  equipment  that  is  specific  to  that  vessel.  If  the   vessel  length  differs   from   the  vessel  type  length,  then   the   vessel   type   wire   frame   is  scaled   accordingly.  Note   that   either,   or   both,   of   these   wire  frames  can  be  empty  (i.e.  no  edges)  if  desired.   The   drawing   data   do   not   affect   the   mathematical   model   in   any   way   Ȃ   they   are   purely   for   drawing   3D   views.   The   vertices  and  edges  follow  the  motions  of  the  vessel,  and  thus  may  be  used  to  improve  understanding  of  the  motion   of  the  model.  They  can  also  be  used  to  represent  a  spar  or  other  equipment  attached  to  the  vessel,  so  that  you  can   then   look   for   clashing   with   other   parts   of   the   system.   For   example   during   a   simulation   replay   you   can   adjust   the   viewpoint  to  look  exactly  along  the  edge  of  interest,  and  check  visually  if  other  parts  of  the  model  pass  through  it.  

Shaded  Drawing   By  default,  for  shaded  3D  Views,  vessels  are  drawn  using  the  wire  frame  data.   The   wire   frame   drawing   data   comprises   vertices   and   edges,   but   OrcaFlex   needs   a   solid   surface   for   the   shaded   graphics   representation.   OrcaFlex   uses   the   following   procedure   to   generate   this   surface   from   the   wire   frame   vertices  and  edges.   First  any  edges  with  specified  diameters  (i.e.  diameters  not  equal  to  '~')  are  drawn  as   cylinders.  This  allows  you  to   use   such   edges   to   visualise   parts   of   the   structure   that   are   not   solid,   e.g.   crane   boom   latticework.   These   edges   are   now  handled  and  are  excluded  from  the  remainder  of  the  procedure.   The  remaining  edges  are  used  to  partition  the  vertices  into  sets  of  connected  vertices.  Two  vertices  are  deemed  to   be  connected  if  there  exists  a  path  of  edges  between  the  two  vertices.   Finally,  for  each  set  of  connected  vertices,  the  smallest  convex  hull  enclosing  the  set  is  drawn.   This   algorithm   does   not   always   generate   the   shaded   drawings   that   you   might   expect.   Consider   the   following   two   wire  frame  vessels.  When  drawn  in  wire  frame  mode  they  look  the  same,  but  in  shaded  mode  they  differ.  

237  

w

 

System  Modelling:  Data  and  Results,  Vessels    

  Figure:  

Wire  Frame  and  Shaded  Drawing  

For  the  green  vessel  the  superstructure  and  the  hull  share  vertices  and  so  all  vertices  are  connected.  This  results  in  a   single  convex  hull  for  all  vertices  being  drawn.  In  the  red  vessel,  the  superstructure  and  hull  do  not  share  vertices   and   so   there   are   two   distinct   sets   of   connected   vertices.   This   results   in   two   separate   convex   hulls   and   a   better   representation.   Alternatively  the   object   can  be  represented  by  an  imported   3D   model  by   specifying  the  Shaded  Drawing   File.  This   must  be  a   Direct   X  format  file,  usually  with   the  .x   file   extension.  If  you  use  a  relative   path  then  the  path   will  be  taken   as  relative  to  the  folder  containing  the  OrcaFlex  file.   The   Browse   button   allows   you   to   specify   the   Shaded   Drawing   File   quickly   and   also   provides   quick   access   to   the   Orcina  sample  drawings  via  the  Navigate  to  Orcina  sample  drawings  button.   The  Use   Culling  option   is   normally   selected   since   it   can   provide   a   useful   performance   benefit.   However,   in   order   to   work   it  requires  that  the  triangles  defined  in  the  .x   file  have  their  outward  facing  directions  defined  correctly.  In   the  

238  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

unusual   situation   where   the   outward   facing   directions   are   not   defined   correctly   then   the   .x   file   will   not   display   correctly.  If  this  happens  then  typically  sections  of  the  model  will  be  missing  w hen  drawn  by  OrcaFlex.  Disabling   the   Use  Culling  option  resolves  this  problem.   Draw  Size  is  provided  to  allow  you  to  scale  the  drawing.  All  directions  are  scaled  equally  to  arrange  that  the  longest   side  in  the  drawing  is  drawn  to  the  specified  Draw  Size.  This  longest  side  is  calculated  by  first  fitting  the  smallest   possible  cuboid  around  the  vertices  of  the  shaded  drawing  (these  are  defined  in  the  .x  file).  This  cuboid  is  aligned   with  the  shaded  drawing's  local  axes.  Then  the  length  of  the  longest  side  of  t his  cuboid  is  found.   Specify  a  value  of  '~'  to  display  the  drawing  using  the  absolute  coordinates  as  specified  in  the  .x  file.   Note:  

If  you  use  a   value  of  '~'  for   Draw  Size  then  OrcaFlex   uses  the  coordinates  in  the  .x   file  directly.  If   these   coordinates   use   a   different   length   units   system   from   your   OrcaFlex   model   then   you   should   specify   the   units   used   in   the   .x   file   by   including   an   auxiliary   file   called   AdditionalInformation.txt.   Examples  of  this  can  be  found  in  the  sample  shaded  drawings  provided  by  Orcina.  

Shaded   Drawing   Origin   is   provided   because   the   shaded   drawing   and   the   vessel   may   have   different   origins.   The   Shaded   Drawing   Origin   defines   the   origin   of   the   shaded   drawing   with   respect   to   the   vessel's   local   axis   system.   Similarly   Shaded   Drawing   Orientation   allows   you   to   reorient   the   shaded   drawing   to   match   the   vessel's   axis   system.  

External  Functions   Parameters  

This  data  item  specifies  the   External  Function  Parameters,  a  free   form  multi-­‐line  text   field   which  is  passed  to   any   external  function  used  by  the  Environment.  

Properties  Report   The  Vessel  properties  report  is  available  from  the   popup-­‐menu  on  the  data  form.  It  reports  the  following:   Length  

The  length  of  the  vessel.   Wave  direction  relative  to  vessel  

The  relative  wave  direction.  If  there  is  more  than  one  wave  train  then  the  direction  of  the  first  wave  train  is  used.   Statics  force  accuracy,  Statics  moment  accuracy  

These  are  only  reported  if  the  vessel  is  included  in  statics.   The   static   analysis   searches   for   an   equilibrium   position   for   the   vessel   Ȃ   that   is   a   position   for   which   the   resultant   force   and   moment   on   the   vessel   is   zero.   We   refer   to   the   resultant   force   and   moment   as   the   out   of   balance   load.   Because   computers   have   limited   numerical   precision   the   static   analysis   cannot   always   find   a   configuration   where   the  out  of  balance  load  is  exactly  zero.  OrcaFlex  accepts  a  position  as  a  static  equilibrium  position  if  the  largest  out  of   balance  load  component  is  less  than  the  statics  accuracy.   The   Statics  force  accuracy  equals   Tolerance   *  vessel  typical  force  and  the  Statics   moment  accuracy  equals  Tolerance   *   vessel   typical   moment.   The   vessel   typical   force   and   moment   are   based   on   the   forces   and   moments   applied   by   connected  objects.   Reducing   the   Tolerance   value   will   give   a   more   accurate   static   equilibrium   position,   but   will   take   more   iterations.   OrcaFlex   may   not   be   able   to   achieve   the   Tolerance   specified   if   it   is   too   small,   since   the   computer   has   limited   numerical  precision.   Note:  

6.7.2

The  statics  accuracies  change  during  the   static  analysis   because  the  forces  and   moments  applied   by  connected  objects  vary  with  the  position  of  the  vessel.  The  statics  accuracies  reported  in   Reset   state   may   be   quite   different   from   those   used   for   the   final   equilibrium   position   and   should   be   treated  as  rough  approximations  to  the  true  statics  accuracies.  

Vessel  Types  

Each  vessel  has  a  vessel  type  that  determines  a  lot  of  its   data  and  which  is   defined   on  the  vessel  types  form.   You   can   define   a   number   of   different   vessel   types   and  each  type   is   given   a  name,  which  is  then   used   on   the   vessel   data   form   to  specify  the  type  of  that  particular  vessel.  

239  

w

 

System  Modelling:  Data  and  Results,  Vessels    

Two   different   vessels   can   have   the   same   type.   To   illustrate   this,   consider   a   model   of   a   pipe   being   towed   by   two   identical   tugs.   This   is   modelled   by   creating   a   vessel   type   called   'Tug'   and   then   creating   two   vessels,   each   of   type   'Tug'.  The  RAOs,  for  example,  are  data  of  the  'Tug'  vessel  type,  since  they  apply  to  both  tugs.  On  the  other  hand  the   two  tugs  differ  in  their  positions  and  any  prescribed  motion,  so  these  are  properties  of  the  individual  vessel  objects.   You   don't   have   to   use   all,   or   even   any,   of   the   vessel   types   you   define.   For   example   you   can   set   up   a   data   file   that   defines  a  number  of  vessel  types  but  has  no  vessels.  Such  a  file  can  then  act  as  a  library  of  vessel  types  that  can  be   imported  into  other  OrcaFlex  data  files.   Vessel  Type  Data   For   each  Vessel  Type   you   can   enter   data   for   several   different   draughts,   each  draught   h aving   a   user-­‐specified  Name.   Each   vessel   in   the   model   must   specify   (on   its   vessel   data   form)   which   draught   to   use.   It   is   not   possible   to   use   different  draughts  at  different  times  during  the  same  simulation.   Some  of  the  vessel  type  data  apply  to  all  draughts,   but  a  lot  of  the  data  is  draught-­‐dependent  and  so  separate   data  is   defined  for  each  defined  draught.  The  principal  categories  of  vessel  type  data  are:   x

Geometry  and  drawing  data  which  apply  to  all  draughts.  

x

Conventions  define  the  meaning  of  any  RAO  and  wave  drift  QTF  data.  The  conventions  apply  to  all  draughts.  

x

RAO  data.  Separate  RAOs  are  specified  for  each  different  draught.  There  is  a   Check  RAOs  facility  that  provides   RAO  graphs  that  help  detect  errors.  

x

Wave  Drift  data.  Separate  values  are  specified  for  each  different  draught.  

x

Hydrodynamic  and  Wind  Damping  data.  Separate  values  are  specified  for  each  different  draught.  

x

Inertia  and  Damping  data.  Separate  values  are  specified  for  each  different  draught.  

Default  Vessel  Type  Data  

When  you  create  a  new  vessel  type,  it  is  given  initial  default  data  that  corresponds  to  a  tanker.  You  should  replace   this  with  accurate  data  for  the  vessel  you  are  modelling.   If  your  vessel  is  similar  to  the  default  data  tanker  then  you  might  find  the  default  data  useful  if  you  have  no  better   data   available.   However   note   that   the   default   vessel   tanker   has   a   significant   heave   resonance   in   beam   seas   at   7s   period.   The  default  data  was  obtained  as  follows.   x

The   default   structure   data,   displacement   RAOs,   wave   load   RAOs,   wave   drift   QTFs,   stiffness,   added   mass   and   damping  data  all   come  from  an  NMIWave  diffraction  analysis  of  a  103m  long  tanker  in  400m  water  depth.  The   tanker   used   in   this   analysis   had   the   following   properties:   Breadth   15.95m,   Draught   6.66m,   Transverse   GM   1.84m,  Longitudinal  GM  =  114m,  Block  Coefficient   0.804.  The  diffraction  analysis  used  8%   extra  damping  in   roll   about  CG.  

x

The  default  hydrodynamic  and  wind  drag  coefficients  are  based  on  graphs  given  in  the  standard   OCIMF  book.   Note   that   the   OCIMF   book   gives   different   coefficients   for   different   vessel   types   and   draughts.   The   default   vessel   type   data   are   approximate   averaged   deepwater   coefficients,   and   they   have   been   rounded   to   only   1   or   2   significant  figures.  

x

The  hydrodynamic  and  wind  drag  areas  and  area  moments  are  set  for  the  above  103m  tanker.  The  wind  drag   areas  are  based  on  an  assumed  average  upperworks  height  of  12m  for  surge  and  9m  for  sway.  

x

The  origin  for  these  default  data  is  on  the  centreline  at  midships  and  at  the  mean  water  level.  

OrcaFlex   automatically   Froude   scales   vessel   type   data   to   the   vessel   length   you   specify.   So   this   default   data   might   sometimes  still  be  useful  if  your  vessel  is  a  different  length  to  the  tanker  described  above,  but  is  otherwise  similar.  

Structure   Vessel  Type  Length  

The   length   to   which   the   vessel   type   RAO   and   drawing   data   apply.   This   may   be   left   unspecified   ('~').   If   a   value   is   specified,  then  it  may  be  used  to  scale  the  vessel  type  data  to  the  length  of  the  vessel.  See   Vessel  Length  for  details.   The   remaining  data  on  the  Structure  page  (on  the  vessel  type  form)  are  only  used  if  the  vessel   Primary  Motion  is   set   to  Calculated  (3  DOF)  or  Calculated  (6  DOF).  Note  that  these  data  are  draught-­‐specific.  

240  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

Mass  and  Moments  of  Inertia  

The   vessel   type's   mass   and   its   moments   of  inertia   about   axes   through   the   CG   in   the   vessel   x,   y   and   z   directions.   This   should  include  the  structural  and  contents  mass  and  inertia,  but  not  the  added  mass.   Centre  of  gravity  (CG)  

The  coordinates  of  the  vessel  type's  centre  of  mass,  relative  to  vessel  axes.  

Conventions   The  conventions  page  (on  the  vessel  types  data  form)  contains  settings  that  define  the  meaning  of  the  RAO,  QTF,  and   (in  the  case  of  the  direction  conventions)    stiffness,  added  mass  and  damping  data.  This  enables  you  to  enter  these   data  directly  from  many  other   programs  without  having  to  convert  the  values  between  different  sets  of  conventions.   Instead  you  can  tell  OrcaFlex  the  conventions  that  apply  to  those  data  and  OrcaFlex  will  then  automatically  allow  for   those  conventions  when  it  uses  the  data.   Warning:  

In   general,   the   conventions   apply   to   all   RAO   and   QTF   data.   You   cannot   mix   data   with   differing   conventions  in  the  same  model  (but  see  rotational  RAOs  below).  

Although  RAOs  are  simple  enough  in  principle,  a  number  of  complications  make  them  notoriously  error-­‐prone  and   difficult  to  check  in  practice.  The  main  issues  are:   x

Different  coordinate  systems.  

x

Different  definitions  of  phase  angle  and  rotational  RAOs.  

x

Use   of  vessel   symmetry,  e.g.   to  obtain   motions   in   seas   from   the   port   side   given   data   for   seas   from   the   starboard   side.  

OrcaFlex  provides  easy  ways  of  handling  these  problem  areas.   The   use   of   differing   coordinate   systems   and   conventions   by   different   suppliers   of   data   is   the   main   source   of   confusion.  It  is  vital  that  you  know  the  conventions  that  apply  to  the  RAO  tables  that  you  are  using.  Unfortunately,   not   all   RAO   tables   fully   document   the   conventions   used:   see   RAO   data   checklist   for   help   finding   out   what   conventions  apply  to  your  data  and  see  Checking  RAOs  to  check  that  the  conventions  are  set  correctly.   Translational  RAO  Conventions  

Translational  displacement  RAOs  are  always  non-­‐dimensional  (e.g.  metres/metre  or  feet/foot).   Translational  load  RAOs  are  always  given  as  force  per  unit  wave  amplitude  (e.g.  kN/m).   Rotational  RAO  Conventions  

Roll,  pitch  and  yaw  displacement  RAOs  may  be  specified  using  one  of  the  following  three  possible  conventions:   x

As   rotation   angles   per   unit   wave   amplitude.   The   values   are   dimensional,   e.g.   in   degrees/metre,   radians/foot   etc.  

x

As   rotation   angles   per   unit   maximum   wave   slope.   Maximum   wave   slope   is   the   true   maximum   slope   of   the   ™ƒ˜‡•—”ˆƒ…‡ǡ™Š‹…Š‹•Ɏ Ȁ”ƒ†‹ƒ•ȋαͳͺͲ Ȁ†‡‰”‡‡•Ȍˆ‘”™ƒ˜‡Š‡‹‰Š– ƒ†™ƒ˜‡Ž‡‰–ŠǤŠ‡˜ƒŽ—‡•ƒ”‡ non-­‐dimensional,   which   can   be   thought   of   as   either   degrees/degree   or   radians/radian   which,   of   course,   are   equivalent.  

x

As  rotation  angles  per  unit   wave  steepness.  Wave  steepness  is  a  commonly  used  angular  measure  of  a  wave,   defined  by  steepness  =  H/L  radians  (=  (18ͲȀɎȌ Ȁ†‡‰”‡‡•ȌǤŠ‡˜ƒŽ—‡•ƒ”‡ƒ‰ƒ‹‘-­‐dimensional,  i.e.  either   degrees/degree  or  (equivalently)  radians/radian,  but  of  course  the  RAOs  are  numerically  larger  (by  a  factor  of   ɎȌ–Šƒ–Š‘•‡—•‹‰–Š‡ƒš‹—™ƒ˜‡•Ž‘’‡…‘˜‡–‹‘Ǥ  

In   each   case,   the   angles   of   rotation   may   be   given   in   either   degrees   or   radians;   the   convention   on   the   data   form   should   be   specified   accordingly.   For   the   two   non-­‐dimensional   cases,   the   units   (degrees   or   radians)   of   the   max   wave   slope   or   wave   steepness   must   be   the   same   as   those   for   the   angles   of   rotation.   Note   that,   so   long   as   the   units   are   consistent   in   this   way,   the   actual   numerical   values   for   these   non-­‐dimensional   RAOs   are   the   same   regardless   of   whether  the  unit  selected  is  degrees  or  radians.   Rotational   load   RAOs   must  be   given   as   moment   per   unit   wave   amplitude   (e.g.   kNm/m),   and  rotational   QTFs   (i.e.   the   yaw   QTF)   are   non-­‐dimensional.   Therefore,   the   degrees/radians   switch   and   the   unit   amplitude   /   steepness   /   maximum  slope  switch  do  not  apply  to  load  RAOs.  This  is  the  exception  to  the  above  rule  that  the  conventions  apply   to  all  RAOs  and  QTFs.  

241  

w

 

System  Modelling:  Data  and  Results,  Vessels    

Warning:  

If   rotational   displacement   RAOs   are   given   relative   to   wave   slope   or   steepness,   then   OrcaFlex   (internally)   converts   them  to   be   relative   to   wave   amplitude   using   the  deep   water  wavelength,   not   the  wavelength  for  the  water  depth  specified  in  the  model.  

Waves  are  referred  to  by  

The   RAO   and   QTF   data   can   be   specified   by   period   in   seconds,   by   angular   frequency   in   radians/second   or   by   frequency  in  Hertz.   RAO  Phases  

The  RAO  Phase  convention  is  specified  by  3  data  items:   1.

Phases  are  either  leads  or  lags.  

2.

Phases  are  specified  either  in  degrees  or  radians.  

3.

The   phase   defines   the   time   at   which   the   maximum   positive   value   of   the   motion   occurs.   This   is   relative   to   the   time   at   which   the   wave   crest,   trough,   zero   up-­‐crossing   or   zero   down-­‐crossing   passes   the   phase   origin.   Wave   Drift  QTFs  are  not  affected  by  this  convention,  since  they  are  entered  without  any  phase  information.  

Directions  

You   must   specify   the   directions   that   correspond   to   positive   motion   or   load   in   the   RAO   and   QTF   data,   and   in   the   stiffness,   damping   and   added   mass   matrices.   The   most   common   convention   is   as   given   by   the   default   OrcaFlex   vessel  type:  a  right-­‐handed  system  with  Z  upwards  and  clockwise  rotations  being  positive.   Symmetry  

You   can   specify   symmetry   of   the   vessel   type.   OrcaFlex   will   then   use   the   user-­‐specified   RAO/QTF   tables   for   wave   directions  on  one  side  of  the  symmetry   plane   to   derive  tables   for  the  reflected  directions  on  the  other  side   of  the   plane.   The  Symmetry  can  be  set  to:   x

None:   The  vessel  type  has   no  symmetry.  The  directions   specified  must  cover  all  the   wave   directions  used  in  the   simulation.  

x

XZ   plane   (or   YZ   plane):   This   specifies   that   the   XZ   (or   YZ)   plane   through   the  RAO   origin   is   a   plane   of  symmetry.   For  each  direction  given   OrcaFlex  uses  symmetry  to   derive  tables  for  the  reflected   direction  on  the  other  side  of   the  plane.  

x

XZ   &   YZ   planes:  This  specifies   that  both  the  XZ   and   YZ   planes   through   the   RAO   origin   are   planes   of   symmetry.   For   each   direction   given   OrcaFlex   uses   symmetry   to   derive   tables   for   the   reflected   directions   in   the   other   3   quadrants.  

x

Circular:  This  specifies  that  the  vessel  has  circular  symmetry  about  the  RAO  origin.  RAO/QTF  tables  can  only  be   given  for  one  wave  direction,  and  OrcaFlex  uses  symmetry  to  derive  tables  for  all  other  directions.   Warning:  

If  you   specify   some   planes   of   symmetry   then   the   RAO   origin  must   be   on   all   the   planes   of   symmetry.   Or  if  you  specify  circular  symmetry  then  the  RAO  origin  must  be  at  the  centre  of  symmetry.  

RAOs   OrcaFlex   uses   two   different   types   of   RAO   (response   amplitude   operator):   Displacement   RAOs   and   Wave   Load   RAOs.   Displacement  RAOs  are  specified  on  the  RAOs  page  on  the  vessel  type  data  form.  They  define  the  1st  order  motion   of   the   vessel   in   response   to   waves   of   given   period   and   amplitude.   They   are   only   used   if   the   vessel   superimposed   motion  is  set  to  RAOs+Harmonic.  In  the  dynamic  analysis  the  vessel  moves  harmonically,  in  all  6  degrees  of  freedom,   about  its  primary  position.  These  harmonic  motions  are  specified  by  giving  the  RAO  amplitudes  and  phases,  for  all   six   degrees   of   freedom,   usually   for   a   range   of   wave   periods   and   directions.   For   further   information   see   RAOs   and   Phases.   Wave   load   RAOs   are   specified   on   the  Load   RAOs   page   on   the   vessel   type   data   form.   They  define   the   1st   order   wave   force   and   moment   on   the   vessel   due   to   waves   of   given   period   and   amplitude.   They   are   only   used   if   the   1st   order   wave  loads  are  included  for  the  vessel  and  they  only  affect  the  motion  if  the  vessel  primary  motion  is  set  to  one  of   the  calculated  options.   The  2  types  of  RAOs  are  specified  in  very  similar  ways,  using  the  following  data.  

242  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

RAO  Origin  

The  RAO  origin  is  the  point  on  the  vessel  whose  motion  is  defined  by  the  RAOs.  The  RAO  origin  is  specified  by  giving   its  coordinates   with  respect  to   the  OrcaFlex   vessel   axes  (not  the   directions   specified  on   the  Conventions   page).   It   is   commonly,  but  does  not  need  to  be,  at  the  centre  of  gravity.  Different   draughts  can  use  different  RAO  origins.   RAO  Phase  Origin  

The   RAO  phase   origin   is   the   point   on   the   vessel   that   the   RAO   phase   values   are   relative   to.   It   is   specified   by   giving   its   x  and  y-­‐coordinates  with   respect   to  the   OrcaFlex  vessel  axes  (not  the  directions  specified  on  the   Conventions  page).   The  phase  values  given  in  the  RAOs  must   be   relative  to  the  time   that  the   wave  crest  or  trough  (depending   on   the   RAO  phase  conventions  specified)  passes  the  specified  RAO  phase  origin.   Often  the  phase  origin  is  the  same  as  the  RAO  origin,  i.e.  the  phases  are  relative  to  the  time  the  crest  or  trough  passes   the   point   whose   motion   the   RAOs   define.   In   this   case   the   phase   origin   can   be   set   to   '~',   meaning   'same   as   RAO   origin'.   But   note   that   some   programs   (one   example   being   Moses)   generate   RAOs   where   the   phase   origin   is   not   necessarily  the  same  as  the  RAO  origin.   RAO  Data  

RAO   data   can   be   specified   for   a   number   of   different   wave   directions   relative   to   the   vessel,   using   the   OrcaFlex   direction   convention.   This   relative   wave   direction   is   the   direction   in   which   the   wave   is   progressing,   measured   positive   from   the   vessel   x-­‐direction   towards   the   vessel   y-­‐direction,   and   on   the   RAOs   page   of   the   vessel   type   data   form  each  RAO  table  is  labelled  with  its  associated  direction.   To  change  the  value  of  the  wave  direction  for  one  of  the  RAO  tables,  select  the  tab  for  that  table  and  edit  the  Selected   Direction.  To  insert  a  new  wave  direction  after  an  existing  direction,  select  the  existing  direction's  page  and  click  the   Insert  Direction  button.  Similarly,  the  Delete  Direction  button  deletes  the  currently  selected  direction.   For  each  direction,  the  RAO  table  covers  a  range  of  wave  periods  or  frequencies,  as  specified  in  the  conventions  data.   The  periods/frequencies  need  not  be  entered  in  order  Ȃ  they  will  be  sorted  before  use.   In   the   case   of   a   circular   symmetric   vessel,   RAOs   are   specified   for   only   one   wave   direction   Ȃ   OrcaFlex   will   derive   RAOs  for  all  other  directions.   RAO  Interpolation/Extrapolation   On  direction  

You   must   provide   RAO   tables   that   include   or   span   the   wave   direction(s)   involved   in   the   simulation.   If   RAOs   are   required   for   a   wave   direction   for   which   an   RAO   table   has   not   been   supplied,   then   OrcaFlex   will   use   linear   interpolation  to  obtain  an  RAO  table  for  that  direction.   Warning:  

Interpolation   is   likely  to   be  poor   if  the   interval   involved   is   large.   We   therefore   recommend   that   the   RAO  directions  defined  cover  all  the  wave  directions  that  will  be  used  and  in  steps  of  30°  or  less.  

On  period  

For  regular   wave  analysis,  RAO  data  is  only  needed  for  the  appropriate  wave  period,  or  for   wave   periods  either  side   of  that  period.  For  random  sea  simulations,  RAO  data  should  be  specified  for  a  wide  enough  range  of  wave  periods  to   cover   the   spectrum.   The   View  Wave   Components   button   (on   the   Waves   page   of  the   environment   data   form)   reports   the  wave  frequencies  that  OrcaFlex  will  use  to  represent  the  spectrum.   Note:  

If  the  vessel  length  differs  from  the  vessel  type  length  then  the  RAO  periods  specified  on  the  vessel   type   form   are   Froude   scaled,   and   it   is   these   Froude   scaled   periods   that   must   cover   the   actual  wave   period(s).  

Linear   interpolation   is   used   if   RAOs   are   required   for   a   period   that   is   between   the   periods   given   in   the   table.   We   strongly  recommend  that  your  RAO  tables  provide  data  for  periods  that  include  or  span  all  the  wave  periods   that   will  be  involved  in  the  simulation.   For   periods   shorter   than   the   smallest   value   in   the   table,   OrcaFlex   will   assume   a   zero   amplitude   response   at   zero   period   and   apply   linear   interpolation   between   that   and   the   shortest   period   value.   If   you   specify   an   RAO   for   zero   period  which  has  non-­‐zero  response,   OrcaFlex  will  over-­‐ride  this  with  a  zero   response,  since  a  non-­‐zero  response  at   zero   period  is  physically  implausible.  In  both  of  these  cases,  OrcaFlex  will  issue  a  warning.  To  avoid  this  warning,   simply  specify  a  zero-­‐amplitude  response  at  zero  period  in  your  RAO  tables.  

243  

w

 

System  Modelling:  Data  and  Results,  Vessels    

For   periods   longer   than   the   largest   value   in   the   table,   OrcaFlex   will   use   linear   extrapolation,   based   on   the   two   longest   periods   in   the   RAO  table.   This   may   give   unexpected   results,   especially   if  the   extrapolation   required   is   over   a   large   period   range,   so   again   OrcaFlex   will   issue   a   warning.   To   avoid   this   warning,   you   may   specify   the   limiting   infinite  period  RAO,  which  for  a  free-­‐floating  vessel  can  be  derived  from  the  knowledge  that  the  vessel  must  follow   the  surface  in  a  sufficiently  long  wave.  See  RAO  Quality  Checks  for  details.   Complex-­‐valued  Interpolation  and  Extrapolation  

Note  that  RAO  interpolation  and  extrapolation  is  done  using  the  complex  value  representation  of  the  RAOs,  in  which   the  RAO  with  amplitude  A  and  phase  lag  P  is  represented  by  the  complex  number:   C(A,P)  =  A[cos(P)  +  i  sin(P)].   For   example,   given   RAOs   (A1,P1)   for   direction   D1   and   (A2,P2)   for   direction   D2,   the   interpolated   RAO   for   the   intermediate  direction  D=(D1+D2)/2  is  (A,P),  where:   C(A,P)  =  [C(A1,P1)  +  C(A2,P2)]/2   This  gives  better  results  than  interpolating  the  amplitude  and  phase  separately.   OrcaFlex   first   interpolates   on   direction,   to   obtain   an   RAO   table   appropriate   to   the   wave   direction.   It   then   interpolates   (or   extrapolates)   that   table   to  obtain   the   RAO   for   the   appropriate   wave   period.   Both   of  these   stages   are   done  using  complex  valued  interpolation  or  extrapolation.   Obtaining  the  data   All   of  the   above   data   can   generally   be   obtained   from   the   results   of   a   diffraction   program.   OrcaFlex   can   import   these   data  from  the  output  files  of   some  specific  programs  (AQWA  and  WAMIT)  and  from  generic  text  files  with   OrcaFlex-­‐ specific  markers  added.  There  are  two  different  ways  to  do  this  import.   The  easiest  and  most  reliable  way   is  to   import  all  the  hydrodynamic  data  using  the  Import  Hydrodynamic   Data   button  on  the  Vessel  Types  data  form.  Alternatively,  you  may  use  the  Import  RAOs  button  to  import  only  the  RAO   data.  

Wave  Drift  Loads   The  Wave  Drift  page  on  the  vessel  type  form  contains  the   Quadratic  Transfer  Functions  (QTFs)  that  OrcaFlex  uses  to   calculate  a  wave  drift  load  (sometimes  called  the  slow  drift  load).   Note:  

The  wave  drift  load  is  only  calculated   for  a  vessel  if  the   Wave   Drift  Load  (2nd  Order)  option  is   checked   in   the   list   of   vessel   Included   Effects,   and   will   only   affect   the   vessel   motion   if   the   vessel   primary  motion  is  set  to  one  of  the  Calculated  options.  See  Modelling  Vessel  Slow  Drift  for  details.  

See  Wave  Drift  Load  Theory  for  details  of  how  OrcaFlex  calculates  the  wave  drift  l oads.   QTF  Origin  

The  QTF   origin   is  the   point  on   the   vessel   to   which  the  QTFs   apply.  The  wave  drift   loads  are   calculated   based   on  the   wave  conditions  at  this  point  and  they  are  applied  at  this  point.  The  QTF  origin  is  specified  relative  to  the  OrcaFlex   vessel   axes  (not  the   directions   specified   on   the  Conventions   page);   different  draughts  can   use   different   origins.   The   z-­‐coordinate  of  this  origin  is  not  specified,  since  the  loads  are  calculated  and  applied   only  for  the  'horizontal'  (surge,   sway  and  yaw)  degrees  of  freedom.   Wave  Drift  QTFs  

QTFs,   like   wave   load   RAOs,   are   specified   in   dimensional   form   and   with   respect   to   waves   of   unit   amplitude.   Translational   QTFs   have   units   of   force   per   unit   length   squared,   and   rotational   QTFs   moment   per   unit   length   squared.   No  phases  are  required  because  only  the  diagonal  terms  of  the  full  QTF  matrix  are  entered  in  OrcaFlex,  and  these   diagonal  terms  necessarily  have  zero  phase.  OrcaFlex  uses  Newman's  approximation  to  obtain  the  off-­‐diagonal  QTFs   from  the  diagonal  QTFs  specified.  See  Wave  Drift  Load  Theory  for  details.   The  way  QTF  data  are  entered  in  OrcaFlex  is  much  the  same  as  that  for  RAO  data.  In  particular:   x

For  each  draught,  QTF  tables  are  specified  for  each  of  a  number  of  wave  directions,  specified  using  the   Orcina   direction   convention.   To   insert   a   new   table   use   the   Insert   Direction   button   and   to   delete   a   table   select   that   table's  page  and  then  click  the  Delete  Direction  button.  To  change  the  direction  associated  with  a  table,  select   that  table's  page  and  then  edit  the  Selected  Direction  value.  

244  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

x

The  direction  conventions  apply  to  the  QTFs.  

x

If   the   vessel   type   has   some   symmetry   (see   the   conventions   page)   then   OrcaFlex   automatically   generates   QTF   tables   for   all   the   reflected   directions   implied   by   that   symmetry.   You   must   provide   QTF   tables   for   enough   directions   for   OrcaFlex   to   have   data   (either   user-­‐specified   or   generated  based   on   symmetry)   for   directions  that   cover  the  wave  directions  the  vessel  will  experience.  

x

Each   QTF   table   consists   of   data   for   a   range   of   wave   periods   or   frequencies   (depending   on   the   convention   specified).  You  should  provide  data  for   periods  that  (after  allowing   for  Froude  scaling  if  the  vessel  length  differs   from  the  vessel  type  length)  cover  the  wave  periods  the  vessel  will  experience.   Warning:  

The   settings   on   the   conventions   page   apply   to   all   draughts   and   they   apply   to   both   the   vessel   type's   RAOs   and   to   its   wave   drift   QTFs.   If   your   RAO   and   QTF   data   use   different   conventions   you   will   therefore  need  to  take  account  of  this,  for  example  by  specifying  the  different  convention  systems   for  each  in  the  files  from  which  they  are  imported.  

Obtaining  the  data   All   of  the   above   data   can   generally   be   obtained   from   the   results   of   a   diffraction   program.   OrcaFlex   can   import   these   data   directly   from   the   output   files   of   some   specific   programs   (AQWA   and   WAMIT)   and   from   many   other   formats   with  OrcaFlex-­‐specific  markers  added.  There  are  two  different  ways  to  do  this  import.   The  easiest  and  most  reliable  way   is  to   import  all  the  hydrodynamic  data  using  the  Import  Hydrodynamic   Data   button  on  the  Vessel  Types  data  form.  Alternatively,  you  may  use  the   Import  QTFs  button  to  import  only  the  QTF   data.  

Stiffness,  Added  Mass  and  Damping   Reference  Origin  

The  origin  to  which  the  stiffness,  added  mass  and  damping  matrices  all  refer,  specified  as  coordinates  with  respect   to  the  OrcaFlex  vessel  axes.   Equilibrium  Position  

The  Z  above   mean   water  level  and  the  heel  and  trim  (relative  to  global  axes)  of  the   reference  origin   when   the  vessel   is  in  hydrostatic  equilibrium  for  this  draught.  Note  that  Z  above  mean  water  level  (rather  than  Z  relative  to  global   axes)   is   specified,   so   that   the   vessel   type   data   is   independent   of   mean   water   level   or   choice   of   position   of   global   origin.   Added  Mass,  Damping  and  Hydrostatic  Stiffness   All   these   matrices   must   be   specified   with   respect   to   axes   through   the   given   Reference   Origin   in   the   conventions   directions,  i.e.  with  respect  to  the  directions  specified  on  the  conventions  page  of  the  vessel  types  form.   For  details  of  the  units,  and  the  theory  used,  see  Vessel  Theory:  Stiffness,  Added  Mass  and  Damping.   Hydrostatic  Stiffness  

The  hydrostatic  stiffness  matrix  is  only  specified  for  heave,  roll  and  pitch  directions.  It  is  applied  in  Statics  only  if  the   vessel's  Static  Analysis  includes  6  DOF,  and  in  dynamics  only  if  the  vessel's   Primary  Motion  is  set  to  "Calculated  (6   DOF)".   Added  Mass  and  Damping  

The   added   mass   and   damping   matrices   are   specified   in   all   6   degrees   of   freedom.   They   are   always   applied   to   the   vessel,  but  will  only  influence  the  motion  of  the  vessel  if  the  Primary  Motion  is  set  to  one  of  the  calculated  modes.   Added  Mass  and  Damping  Method  

If   you   choose   Constant   for   the   Added   Mass   and   Damping   method,   then   single-­‐valued   added   mass   and   damping   matrices  will  be  used.   If  you  choose  Frequency  Dependent,  then  you  may  specify  a  number  of  added  mass  and  damping  matrices,  each   pair   corresponding   to   a   particular   given   frequency   or   period.   Whether   you   specify   period   or   frequency   values   is   determined  by  the  Waves  are  referred  to  by  setting  on  the  Vessel  Type  Conventions  page.   If  you  use  the  Constant  (i.e.   frequency  independent)  method,  then  you  should  specify  values  that  are  appropriate   to   the   frequency  of   vessel   motion   you   expect.   To   calculate   slow   drift  motion   of   the   vessel   it   is  normally  appropriate   to  

245  

w

 

System  Modelling:  Data  and  Results,  Vessels    

enter   low   frequency   values.   Otherwise   values   corresponding   to   the   dominant   wave   frequency   are   perhaps   more   appropriate.  Clearly,  if  the  vessel  experiences  a  wide  range  of  frequencies,  the  frequency-­‐dependent  method  is  more   appropriate  and  would  be  expected  to  give  better  results.   If  you  use  the  Frequency  Dependent  method  then  you  need  to  specify   both  the  added  mass  and  damping  matrices,   and  for  a  range  of  frequencies.  Also.  the  added  mass  and  damping  data  should  be  consistent  in  the  sense   that  they   obey  the  Kramers-­‐Kronig  relations  Ȃ  see  Consistent  Added  Mass  and  Damping  for  details.   Cutoff  Time  

When   you   use   frequency-­‐dependent   added   mass   and   damping,   OrcaFlex   applies   the   frequency-­‐dependent   data   in   the   time   domain   by   calculating   and   applying   the   vessel's   Impulse   Response   Functions   (IRF).   See   Vessel   Theory:   Impulse  Response  and  Convolution  for  details.   Realistic   IRFs   decay   to   zero   with   increasing   time   lag.   So   to   improve   the   calculation   speed   OrcaFlex   truncates   the   Impulse   Response   Function  at   the   time   lag   specified   by   the   Cutoff   Time.   The   IRF   is   assumed   to   be   zero   for   time   lags   greater  than  the  Cutoff  Time.   Larger   Cutoff   Time   values   might   give   more   accurate   results   but   require   more   calculation.   In   order   to   choose   the   Cutoff  Time,  you  may  find  it  useful  to  use  the  Report  Vessel  Response  window  to  view  the  graphs  of  the  components   of  the  IRF.  From  the  graphs  you  could  decide  the  time  lag  at  which  the  function  has  decayed  sufficiently  close  to  zero   as  to  have  little  or  no  effect  on  the  calculation.   Note  1:  

Frequency-­‐dependent   added  mass   and  damping   can   be   quite   time-­‐consuming   to   compute.   For  this   reason,  it  is  not  calculated  for  vessels  which   do  not  have  calculated  primary   motion:   in  this  case,   the  added  mass  and  damping  load  is  simply  set  uniformly  to  zero.  

Note  2:  

The   damping   matrix   given   by   a   diffraction   program   models   wave   radiation   damping.   However   there   is   another,   often   more   important,   source   of   damping,   namely   wave   drift   damping.   See   Damping   Effects   on   Vessel   Slow   Drift.   Wave   drift   damping   can   be   modelled   in   OrcaFlex   by   adjusting  the  diagonal  entries  in  the  damping  matrix.  

Obtaining  the  data   All   of  the   above   data   can   generally   be   obtained   from   the   results   of   a   diffraction   program.   OrcaFlex   can   import   these   data  from  the  output  files  of   some  specific  programs  (AQWA  and  WAMIT)  and  from  generic  text  files  with   OrcaFlex-­‐ specific  markers  added.  There  are  two  different  ways  to  do  this  import.   The  easiest  and  most  reliable  way   is  to   import  all  the  hydrodynamic  data  using  the  Import  Hydrodynamic   Data   button   on   the   Vessel   Types   data   form.   Alternatively,   you   may   use   the   Import   Matrices   button   to   import   the   frequency-­‐dependent  added  mass  and  damping  matrices.  

Hydrodynamic  and  Wind  Damping   Hydrodynamic   and   wind   damping   loads   on   a  vessel   are   loads   due   to   the   relative   velocity   of  the   fluid   past   the   vessel.   They  can  be  modelled  using   the  data  on  the  Hydrodynamic  Damping  and  Wind  Damping  pages  on  the  vessel  type   data   form.   If   the   length   of   the   vessel   differs   from   that   of   the   vessel   type   then   the   vessel   type   data   will   be   scaled   accordingly.   These  loads  are  an  important  source  of  damping  when   modelling  vessel  slow  drift.  For  a  discussion  of  the  various   damping  sources  see  Damping  Effects  on  Vessel  Slow  Drift.   The   velocity   used   to   calculate   the   drag   loads   is   the   relative   velocity   of   the   fluid   past   the   vessel.   This   includes   any   current  or   wind  velocity  and  the  vessel  velocity  due  to   any   primary  motion.  The  drag  forces  and  moments  due  to   translational   motion   are   modelled   using   the   standard   OCIMF   method.   The   drag   forces   and   moments   due   to   any   vessel   rate   of   yaw   are   modelled   using   yaw   rate   drag   load   factors.   For   details   of   how   the   loads   are   calculated,   see   Vessel  Theory:  Drag  Loads.   Warning:  

The   current   and   wind   loads   are   based   on   theory   for   surface   vessels   and   are   not   suitable   for   submerged  vessels.  

Load  Origin  

The  coordinates  (relative  to  vessel  axes)  of  the  point  on  the  vessel  at  which  the  hydrodynamic  or  wind  drag  loads   are  calculated  and  at  which  they  will  be  applied.  This  need  not  be  at  the  vessel  origin.   It  is  normally  best  to  place  the   load  origin  at  the  centre  of  the  vessel.  

246  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

The   velocity   used   in   the   hydrodynamic   drag   load   calculation   is   the   current   velocity   at   the   load   origin,   minus   any   velocity  of  the  load  origin  due  to   primary  motion  of  the  vessel.  Note  that  if  the  load  origin  is  above  the   water  surface   then  the  current  velocity  used  is  that  at  the  water  surface.   The   velocity   used   in   the   wind   load   calculation   is   the   wind   velocity   (as   specified   on   the   Wind   page   on   the   environment  data  form),  minus  any  velocity  of  the  wind  load  origin  due  to   primary  motion  of  the  vessel.  Note  that   the  wind  velocity  specified  should  be  that  at  10m  above  mean  water  level,  since  that  is  the  height  used  by  the  OCIMF   vessel  wind  load  model  (see  Wind  Speed  for  more  details).   Load  Symmetry  

Specifies  what  symmetry  the  vessel  type  has  below  (for  hydrodynamic  damping)  or  above  (for  wind  damping)  the   water   line,   about   the   load   origin.   For   XZ   and   YZ   symmetry,   OrcaFlex   will   use   the   symmetry   to   derive   load   coefficients  for   extra  directions  generated  by  reflection   in  the  specified  vessel  axes  planes.  For  circular  symmetry,   you   must   specify   coefficients   for   one   direction   only   and   OrcaFlex   will   use   symmetry   to   derive   coefficients   for   all   other  directions.   Note:  

The  symmetry  for  hydrodynamic  drag,  wind  drag  and  RAOs  (see   RAO  Symmetry)  need  not  be  the   same,  though  of  course  the  symmetry  for  hydrodynamic  drag  would  normally  be  the  same  as  that   for  RAOs.  

Areas  and  Area  Moment  

The  surge  and  sway  areas  and  yaw  area  moment   that  will  be  used  to  calculate  the  current  or  wind  loads.  For  details   see  Vessel  Theory:  Drag  Loads.   Coefficients  

Load  coefficients  are  specified  for  the  vessel  surge,  sway  and  yaw  directions.  They   depend  on   the   direction  of  the   current  or  wind,  relative  to  the  vessel  (direction  0  meaning  from  astern,  90  meaning  from  starboard,  etc.).   OrcaFlex  uses  any  symmetry  specified  to  derive  coefficients  for  other  directions  and  then  uses  linear  interpolation   to  derive  coefficients  for  intermediate  directions.   Note:  

When   the   symmetry   is   XZ   and   YZ   the   yaw   moments   must   be   zero,   so   OrcaFlex   forces   zero   yaw   coefficients  in  this  case.  

The   View  Coefficients   button   allows   you   to  view  the   coefficients   that   will   be   used  Ȃ   the   blobs   on   the   graph   show   the   coefficients   you   have   specified   plus   any   that   OrcaFlex   has   derived   using   reflection,   and   the   curve   shows   the   interpolated   coefficients   that   will   be   used   for   intermediate   directions.   You   should   specify   sufficient   directions   to   define  the  shape  of  the  curve  and  to  cover  the  range  of  directions  that  the  vessel  will  experience.   Yaw  Rate  Drag  Factors  (hydrodynamic  damping  only)  

The  yaw  rate  drag  factors  determine  the  yaw  drag  moment,  and  any  surge  and  sway  drag  forces,  that  result  if  the   vessel  has  a  non-­‐zero  rate  of  yaw.   For  a  slender  ship,  and  if  the  load  origin  has  been  placed  at  the  centre  of  the  vessel,  then  the  surge  and  sway  drag   factors  can  usually  be  taken  to  be  zero,  and  then  yaw  drag  factor  can  be  estimated  based  on  the  vessel  length  and   draught.  See  Drag  Loads  due  to  Yaw  Rate  for  details.   Roll  Damping  (hydrodynamic  damping  only)  

These  coefficients  allow  you  to  include  a   roll  damping  moment  that   is  proportional  to  the  roll  component  of  angular   velocity.  Both  linear  and  quadratic  terms  can  be  modelled.  See  Roll  Damping  for  details.  

Drawing   Drawing  Data  

Each   vessel  of  this  type  is  drawn  as  a   wire   frame,   based  on  vertices  and   representing  the   vessel  type,  plus  a   wire   frame  representing  vessel-­‐specific  features.  See  Drawing.   Edge  diameter  (used  only  for  shaded  drawing)  

For   shaded   graphics   views,   by   default,   the   vessel   is   drawn   using   a   solid,   filled-­‐in   shape   based   on   the   vertices   and   edges.  

247  

w

 

System  Modelling:  Data  and  Results,  Vessels    

As   an   alternative   you   can   use   the   vertices   and   edges   to   define   a   frame   like   structure.   If   the   edge   diameter   is   '~'   then   that   edge   will   be   used   to   build   a   filled   in   shape,   otherwise   that   edge   is   drawn   as   a   cylinder   with   the   specified   diameter.  Note  that  you  can  use  a  mixture  of  edge  diameters  (some  defined,  some  set  to  '~')  to  combine  both  filled   in  and  framework  shapes.  

  Figure:  

Wire  frames  with  different  edge  diameter.  A   value  of  '~'  is  used  for  the   wire  frame  on  the  left   and  a  value  of  1m  is  used  for  the  wire  frame  on  the  right.  

Shaded  Drawing   By  default,  for  shaded  3D  Views,  vessels  are  drawn  using  the  wire  frame  data.   The   wire   frame   drawing   data   comprises   vertices   and   edges,   but   OrcaFlex   needs   a   solid   surface   for   the   shaded   graphics   representation.   OrcaFlex   uses   the   following   procedure   to   generate   this   surface   from   the   wire   frame   vertices  and  edges.   First  any  edges  with  specified  diameters  (i.e.  diameters  not  equal  to  '~')  are  drawn  as   cylinders.  This  allows   you  to   use   such   edges   to   visualise   parts   of   the   structure   that   are   not   solid,   e.g.   crane   boom   latticework.   These   edges   are   now  handled  and  are  excluded  from  the  remainder  of  the  procedure.   The  remaining  edges  are  used  to  partition  the  vertices  into  sets  of  connected  vertices.  Two  vertices  are  deemed  to   be  connected  if  there  exists  a  path  of  edges  between  the  two  vertices.   Finally,  for  each  set  of  connected  vertices,  the  smallest  convex  hull  enclosing  the  set  is  drawn.   This   algorithm   does   not   always   generate   the   shaded   drawings   that   you   might   expect.   Consider   the   following   two   wire  frame  vessels.  When  drawn  in  wire  frame  mode  they  look  the  same,  but  in  shaded  mode  they  differ.  

248  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

  Figure:  

Wire  Frame  and  Shaded  Drawing  

For  the  green  vessel  the  superstructure  and  the  hull  share  vertices  and  so  all  vertices  are  connected.  This  results  in  a   single  convex  hull  for  all  vertices  being  drawn.  In  the  red  vessel,  the  superstructure  and  hull  do  not  share  vertices   and   so   there   are   two   distinct   sets   of   connected   vertices.   This   results   in   two   separate   convex   hulls   and   a   better   representation.   Alternatively  the   object   can  be  represented  by  an  imported   3D   model  by   specifying  the  Shaded  Drawing   File.  This   must  be  a   Direct   X  format  file,  usually  with   the  .x   file   extension.  If  you  use  a  relative   path  then  the  path   will  be  taken   as  relative  to  the  folder  containing  the  OrcaFlex  file.   The   Browse   button   allows   you   to   specify   the   Shaded   Drawing   File   quickly   and   also   provides   quick   access   to   the   Orcina  sample  drawings  via  the  Navigate  to  Orcina  sample  drawings  button.   The  Use   Culling  option   is   normally   selected   since   it   can   provide   a   useful   performance   benefit.   However,   in   order   to   work   it  requires  that  the  triangles  defined  in  the  .x   file  have  their  outward  facing  directions  defined  correctly.  In   the  

249  

w

 

System  Modelling:  Data  and  Results,  Vessels    

unusual   situation   where   the   outward   facing   directions   are   not   defined   correctly   then   the   .x   file   will   not   display   correctly.  If  this  happens  then  typically  sections  of  the  model  will  be  missing  when  drawn  by  OrcaFlex.  Disabling   the   Use  Culling  option  resolves  this  problem.   Draw  Size  is  provided  to  allow  you  to  scale  the  drawing.  All  directions  are  scaled  equally  to  arrange  that  the  longest   side  in  the  drawing  is  drawn  to  the  specified  Draw  Size.  This  longest  side  is  calculated  by  first  fitting  the  smallest   possible  cuboid  around  the  vertices  of  the  shaded  drawing  (these  are  defined  in  the  .x  file).  This  cuboid  is  aligned   with  the  shaded  drawing's  local  axes.  Then  the  length  of  the  longest  side  of  this  cuboid  is  found.   Specify  a  value  of  '~'  to  display  the  drawing  using  the  absolute  coordinates  as  specified  in  the  .x  file.   Note:  

If  you  use  a   value  of  '~'  for   Draw  Size  then  OrcaFlex   uses  the  coordinates  in  the  .x   file  directly.  If   these   coordinates   use   a   different   length   units   system   from   your   OrcaFlex   model   then   you   should   specify   the   units   used   in   the   .x   file   by   including   an   auxiliary   file   called   AdditionalInformation.txt.   Examples  of  this  can  be  found  in  the  sample  shaded  drawings  provided  by  Orcina.  

Shaded   Drawing   Origin   is   provided   because   the   shaded   drawing   and   the   vessel   type   may   have   different   origins.   The   Shaded   Drawing   Origin   defines   the   origin   of   the   shaded   drawing   with   respect   to   the   vessel   type's   local   axis   system.   Similarly   Shaded   Drawing   Orientation   allows   you   to   reorient   the   shaded   drawing   to   match   the   vessel   type's  axis  system.   Vessel  length  scaling  

If   the   vessel   length   and   the   vessel   type   length   are   different   then   the   shaded   drawing   is   scaled   accordingly.   This   scaling  is  applied  in  addition  to  the  scaling  of  the  shaded  drawing  specified  by  the  Draw  Size.  

Importing  Hydrodynamic  Data   The   Import   Hydrodynamic   Data  button   on   the   vessel   types   form   enables  the   import   of   data,   in   the   form   of   results   from   common   ship   response   calculation   packages,   into   OrcaFlex.   Presently,   OrcaFlex   can   import   data   which   has   been   output   by   AQWA   and   WAMIT.   OrcaFlex   can   also   import   data   from   Moses,   Hydrostar/ARIANE   and   Wadam   output  files,  after  the  addition  of  some  markup  text.  The  form  of  this  markup  also  allows  for  import  from  any  generic   text  file.   Importing  data  from  AQWA  and  WAMIT   OrcaFlex   attempts   to   import   as   much  data   as   possible   from   the   file   to   the   appropriate   data   items   of  the   vessel   type.   These   data   are:   displacement   RAOs,   load   RAOs,   QTFs,   added   mass   and   damping,   and   hydrostatic   stiffness.   The   imported   data   is   converted   to   match   the   vessel   type   conventions   and   scaled   to   match   the   units   of   the   OrcaFlex   model.   Mass,   inertia   and   hydrostatic   equilibrium   position   are   not   always   output   by   these   packages,   but   will   be   imported   by   OrcaFlex   if   they  are   available.  The  remaining   vessel   type   data   are  not  typically  calculated   by   such  programs,   so   will   need  to  be  obtained  from  another  source.  Any  OrcaFlex  data  for  which  there  are  not  any  corresponding  data  in  the   imported  file  will  not  be  changed.   If   the   import   fails   for   any  reason,   OrcaFlex   will   report   an   error   and  reinstate   the   vessel   type   data   to   that   which  was   present   before   the   import  began.   If  the  data   in   the   file   are   ambiguous,  OrcaFlex  will   prompt   you  for   clarification.   At   the   end   of   the   import   process,   OrcaFlex   will   warn   you   of   anything   noteworthy   Ȃ   for   example,   if   no   QTFs   were   present  in  the  file,  or  if  a  damping  matrix  was  non-­‐symmetric.   Importing  data  from  AQWA   AQWA  results  are  output  in  a  text  file  with  a  .lis  extension.  The  data  are  in  dimensional  form.  OrcaFlex  reads  data   into  the  presently  selected  draught.   Units  

OrcaFlex  first  attempts  to  determine  the  units  in  use  in  the  file,  from  the  values  of  g  (acceleration  due  to  gravity)  and   ɏȋ™ƒ–‡”†‡•‹–›Ȍǡƒ†–Š‡‘™Ž‡†‰‡–Šƒ––Š‡—‹–‘ˆ–‹‡‹•–Š‡•‡…‘†Ǥ ˆ–Š‹•‹••—……‡••ˆ—Žǡ–Š‡”…ƒ Ž‡š™‹ŽŽ•…ƒŽ‡ the   data   (if   necessary)   from   the   units   in   use   in   the   file   to   the   units   of   the   OrcaFlex   model.   If   the   units   cannot   be   determined,  or  if  they  are  a  mixture  of  SI  and  US,  OrcaFlex  will  be  unable  to  convert  the  data  in  this  way:  in  this  case,   the  raw  data  will  be  imported  from  the  file  unscaled  and  a  warning  will  be  issued.   Reference  origins  

All  of  the  AQWA  data  have   their  reference  origin  at  the  vessel  centre  of  gravity.   So  OrcaFlex  sets  the   RAO   origins,   QTF  origin,  and  reference  origin  for  stiffness,  added  mass  &  damping  to  the  value  of  the  centre  of  gravity  given  on  

250  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

the  vessel  types  Structure  page.  You  will  need  to  ensure  that  this  value  is  appropriate  before  importing  the  AQWA   data.  RAO  phase  origins,  likewise,  are  at  the  vessel  centre  of  gravity,  so  are  set  appropriately  by  OrcaFlex.   Added  mass  and  damping  

AQWA  outputs  the  whole  6x6  matrices  for   added  mass  and  damping,  but  OrcaFlex  uses  only  the  upper   right  triangle   of   each.   You   will   receive   a   warning   if   any   of   these   matrices,   in   the   imported   file,   are   significantly   non-­‐symmetric,   since  OrcaFlex  is  unable  to  handle  this.   Note:  

Frequency-­‐independent   added   mass   and   damping   are   not   imported.   If   you   have   such   data,   you   may   wish   to   pre-­‐process   your   AQWA   file   to   add   them   to   each   frequency-­‐dependent   item   before   importing  into  OrcaFlex.  

RAOs  

Displacement   RAOs   are   imported   in   a   straightforward   way.   Load   RAOs   are   imported   including   the   Froude-­‐Krylov   contribution.  Since  AQWA  and  OrcaFlex  use  the  same  conventions  for  vessel  headings,  RAO   directions  are  read  in   directly  and  require  no  adjustment.   QTFs  

Wave  drift  QTFs  in  all  six  degrees  of  freedom  are  imported.  As  with  RAOs,  no  adjustment  is  necessary  for  directions.   Stiffness  and  equilibrium  position  

The  heave,  roll  and  pitch  components  of  the   hydrostatic  stiffness  matrix  are  imported.  Again,  only  the  upper  right   triangle  is  used,  and  a  warning  is  issued  if  the  AQWA  data  are  non-­‐symmetric.  The  Z-­‐component  of  the  equilibrium   position  is  imported.  Since  AQWA  does  not  give  the  heel  and  trim  values  at  the  equilibrium  position,  these  values  are   set  to  zero:  if  necessary,  you  should  edit  these  values  yourself.   Note:  

AQWA   may   output   the   stiffness   matrix   twice,   under   the   hydrostatic   and   hydrodynamic   headings.   If   so,  and  if  the  two  matrices  are  not  the  same,  a  warning  will  be  given  and  the  'hydrostatic'  form  will   be  imported.  

Mass  and  inertia  

Mass   and   moments   of   inertia   are   imported,   the   latter   from   the   diagonal   elements   of   the   inertia   matrix   output   by   AQWA.   Importing  data  from  WAMIT   WAMIT   results   are   output   in   a   text   file   with   an   .out   extension.   The   data   are   almost   all   non-­‐dimensional,   the   main   exceptions   to   this   being   g,   acceleration   due   to   gravity,   and   L,   WAMIT's   length   scale   which   is   used   to   non-­‐ dimensionalise  the  data.  OrcaFlex  reads  data  into  the  presently  selected  draught.   Please  note  that  OrcaFlex  is  unable  to  import  WAMIT  data  representing  multi-­‐body  problems.   Units  

The   length  unit   in   use   in   the   file   will   be   determined,   if   possible,   from   the   v alue   of   g  and   the   knowledge   that   the  unit   of  time  is  the  second.  If  this  is  successful,  then  OrcaFlex  will  convert  the  data  from  non-­‐dimensional  to  the  units  of   the   OrcaFlex   model,   using   WAMIT's   length   scale   (converted   to   the   length   unit   of   the   OrcaFlex   model)   and   the   current  OrcaFlex  values  of  g  (acceleration  due  to  gravityȌƒ†ɏȋnominal  water  density),  using  the  scaling  factors   defined  in  the   WAMIT  user   manual.  If  the  length  unit  cannot   be  determined  then  the  raw  data  will  be  imported  from   the  file  unscaled  and  a  warning  will  be  issued.   Reference  origins  

All  of  the  WAMIT  data  have  their  reference  origin  at  WAMIT's  vessel  origin.  OrcaFlex  determines  the  position  of  this   origin,   with   respect   to   OrcaFlex's   vessel   origin,   from   the   difference   in   the   centre   of   gravity   as   given   in   the   two   coordinate  systems,  and  sets  the  RAO  origins,  QTF  origin,  and  added  mass  &  damping  origin  to  have  this  value.  (If   you   choose   the   OrcaFlex   vessel   origin   to  be   the   same   as   the   WAMIT   vessel   origin,   then   the   coordinates   of  the   centre   of  gravity  will  coincide  and  the  reference  origins  will  all  be  zero).   In   WAMIT,   the   phase   origin   is   at   the   global   origin,   so   the   OrcaFlex   RAO   phase   origins   are   represented   by   the   coordinates  of  the  WAMIT  global  origin  relative  to  the  OrcaFlex  vessel  origin.  We  determine  this  by  (i)  expressing   the  WAMIT  global  origin  relative  to  the  WAMIT  vessel  origin  (i.e.  in  "body-­‐fixed  axes"),  using  the  values  of  XBODY,   YBODY,   ZBODY   and  PHIBODY   in   the   WAMIT   output   file,  and   (ii)  translating   this   value   to  be   relative  to   the  OrcaFlex   vessel  origin,  using  the  difference  between  the  centres  of  gravity  in  the  two  systems.  

251  

w

 

System  Modelling:  Data  and  Results,  Vessels    

You   will   therefore   need   to   ensure   that   the   centre   of   gravity   is   set   appropriately   in   OrcaFlex   before   importing   the   WAMIT  data.   Note:  

In  some  cases,  WAMIT  does  not  output  the  horizontal  position  of  the  centre  of  gravity.  If  so,  then   OrcaFlex  will  warn  you  of  this,  and  you  will  need  to  set  these  reference  origins  yourself.  

Added  mass  and  damping  

WAMIT   outputs  added   mass   and   damping  as   a   list  of   (i,j)  components.   OrcaFlex   imports  these   values,   but  uses   only   the  upper  right  triangle  of  each  matrix.  You  will  receive  a  warning  if  any  of  these  matrices,  in  the  imported  file,  are   significantly  non-­‐symmetric,  since  OrcaFlex  is  unable  to  handle  this.  Components  for  which  an  (i,j)  value  is  not  given   are   assumed   to   be   zero.   The   non-­‐dimensional   WAMIT   values   for   added   mass   (a)   and   damping   (b)   are   scaled   to   dimensional  (A,B  respectively)  values  according  to  the  formulae   Aij  αɏkaij   Bij  αɏkɘ„ij   where   k  =  3  for  i,j=1,2,3;   k  =  4  for  i=1,2,3,  j=4,5,6  or  i=4,5,6,  j=1,2,3;   k  =  5  for  i,j=4,5,6  and   ɘ‹•–Š‡ˆ”‡“—‡…›‹”ƒ†Ȁ•Ǥ   RAOs  

Displacement   RAOs   are   imported   in   a   straightforward   way.   They   are   re-­‐dimensionalised   by   multiplying   by   the   factor  Lk,  k  =  0  for  i  =  1,2,3;  k  =  1  for  i  =  4,5,6.  WAMIT  may  calculate   load  RAOs  in  one  of  two  ways,  and  either  or  both   methods   (labelled   "Haskind"   and   "Diffraction")   may   be   present   in   the   file.   If   both   are   given,   OrcaFlex   will   prompt   you   to   choose   one   method   and   will   import   the   data   for   that   method   only.   The   re-­‐dimensionalising   factor   for   load   •‹•ɏ‰k,  k  =  2  for  i=1,2,3;  k  =  3  for  i=4,5,6.  The  RAO  directions  in  OrcaFlex  are  relative  to  the  vessel,  and  are   determined  by  subtracting  the  value  of  PHIBODY  in  the  WAMIT  input  file  (the  vessel  heading  at  rest)  from  each  of   the  values  of  wave  heading,  which  are  given  in  WAMIT  with  respect  to  global  axes.   QTFs  

Wave   drift   QTFs   are   re-­‐†‹‡•‹‘ƒŽ‹•‡† ™‹–Š –Š‡ ˆƒ…–‘” ɏ ‰ k,   k   =   1   for   i=1,2,3;   k   =   2   for   i=4,5,6.   Directions   are   adjusted  in  the  same  way  as  for  RAOs.   WAMIT   may   calculate   QTFs   in   a   number   of   different   ways:   Momentum   Conservation,   Pressure   Integration,   and   Control  Surface.  OrcaFlex  is  presently  unable  to  import  data  resulting  from  the  Momentum  Conservation  calculation,   so  this  method  is  disregarded.  As  with  load  RAOs,  if  data  from  more  than  one  of  the  remaining  methods  is  present,   then  you  will  be  prompted  to  choose  just  one  of  them.   Notes:  

WAMIT   outputs   QTFs   for   pairs   of   wave   headings.   OrcaFlex   allows   only   unidirectional   QTFs,   so   imports  only  the  data  for  which  the  two  headings  coincide.  

Stiffness  and  equilibrium  position  

The  heave,  roll  and  pitch  components  of  the   hydrostatic  stiffness  matrix  are  imported.  The   equilibrium  position  is   not  specified  by  WAMIT,  so  you  should  edit  this  value  yourself.   Mass  and  inertia  

Mass  and  moments  of  inertia  are  not  imported  from  WAMIT  files.  You  should  enter  these  data  yourself.   Importing  data  from  generic  text  files   The   Import   Hydrodynamic   Data   button   can   also   be   to   import   data   from   a   generic   text   file   containing   special   markup  to  identify  and  describe  the  data.  The  following  data  are  imported:   x

Displacement  RAOs.  

x

Load  RAOs.  

x

Wave  Drift  QTFs.  

x

Frequency  dependent  added  mass  and  damping  matrices.  

252  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

This  mode  of  import  is  equivalent  to  i mporting   each  of  the  above  categories  of  data  individually  by  using  the  import   buttons  as  follows:   x

Import  RAOs  on  the  Displacement  RAOs  page.  

x

Import  RAOs  on  the  Load  RAOs  page.  

x

Import  QTFs  on  the  Wave  Drift  page.  

x

Import  Matrices  on  the  Stiffness,  Added  Mass,  Damping  page.  

If   the   import   fails   for   any  reason,   OrcaFlex   will   report   an   error   and  reinstate   the   vessel   type   data   to   that   which  was   present  before  the  import   began.  At  the  end  of  the   import  process,  OrcaFlex  will  warn  you  of  anything   noteworthy  Ȃ   for  example,  if  no  QTFs  were  present  in  the  file,  or  if  a  damping  matrix  was  non-­‐symmetric.  

Importing  RAOs   RAOs   may   be   imported   using   the   Import   RAOs   button   on   the   vessel   types   form.   Data   may   be   imported   from   AQWA   or   WAMIT   output   files   and,   with   the   addition   of   some   markup   text,   from   Moses,   Hydrostar/ARIANE   and   Wadam   output  files.  The  form  of  this  markup  also  allows  for  import  from  other  text  files.   Import  RAOs  from  specific  program  output   OrcaFlex   can   import   RAOs   directly   from   AQWA   and   WAMIT   output   files,   as   described   under   Importing   Hydrodynamic  Data.  All  existing  RAO  data  for  the  selected  draught  will  be  overwritten.   Import  RAOs  from  marked-­‐up  text  files   You  can  use  generic  text  files  to  import  RAO  data  from,  for  example,  a  ship  response  calculation  program  or  model   test  results.   When  you  import  RAOs  from  a  text  file,  any  RAO  data  previously  present  in  OrcaFlex  for  the  draughts  given  in  the   text   file   will   be   deleted.   Other   data   for   these   draughts,   e.g.   wind   and   hydrodynamic   damping   data,   will   not   be   affected.   So   for   each   draught   you   import,   all   the   RAOs   for   that   draught   must   be   in   a   single   file.   You   can   therefore   either  put  all  the  RAOs  in  a  single  file,  or  else  have  separate  files  for  separate  draughts.   RAO   data   in   a   text   file   can   be   imported   providing   that   the   data   appears   in   tabular   form   and   markers   are   first   inserted  into  the  file  to  identify  the  data  to  OrcaFlex.   A  text  RAO  file  must  contain  the  RAO  data  in  the  following  form.  It  is  usually  easy  to  create  a  suitable  file  by  adding  a   few  lines  to  your  original  response  data  file  Ȃ  see  the  examples  below  for  specific  program  output.   x

The  RAO  data  must  appear  in  the  file  in  one  or  more  tables,  each  table  being  for  one  draught  and  direction.  To   enable   OrcaFlex   to   find   the   tables,   each   table   must   be   preceded   by   a   line   containing   the   string   OrcaFlex   Displacement   RAO   Start.   The   table   must   be   immediately   followed   by   a   line   containing   the   string   OrcaFlex   Displacement   RAO   End.   There   must   not  be   any   blank   lines   between   these   two   marker   lines.   For   load   RAOs   these   strings  should  be  OrcaFlex  Load  RAO  Start  and  OrcaFlex  Load  RAO  End.   Note:  

OrcaFlex   does   not   distinguish   between   text   files   containing   displacement   RAOs   and   those   containing  wave  load  RAOs.  You  must  take  care  to  import  them  into  the  correct  location.  

x

Immediately   following   the   line   containing   the   OrcaFlex   RAO   Start   string   there   must   be   two   lines   (in   either   order)  specifying  the  draught  and  direction  that  applies  to  that  table.  The  line  specifying  the   draught   must   be  of   the  form  Draught  DraughtName,  where  DraughtName  is  the  name  of  the  draught.  If  the  name  contains  spaces,   then  DraughtName  must  be  enclosed  in  quotes.  The  line  specifying  the  direction  must  be  of  the  form  Direction  n,   where   n   is   a   number   specifying   the   direction   the   wave   is   progressing,   in   degrees,   measured   positive   from   forward  towards  the  port  side.  So  direction  0  means  waves  coming  from  astern  and  direction  90  means  waves   coming  from  the  starboard  side.  

x

Following  these  two  lines,  the  first  line  of  the  table  must  be  a  set  of  headers  defining  the  subsequent  columns.   This   headers   line   consists   of   a   number   of   character   strings,   separated   by   spaces.   The   strings   indicate   the   contents  of  the  columns  Ȃ  see  Header  Strings  for  Text  RAO  Tables.  Columns  with  unrecognised  header  strings   are  ignored.  

x

If  you   want   OrcaFlex  to  ignore  a  column,   for  example   because   it  contains  irrelevant  or  superfluous  data,  then   insert  an  unrecognised  header  string,  (e.g.  "N/A"  or  "~").  In  particular,  if  the  table  contains   both  wave  period   and  frequency  you  must  indicate  that  one  of  these  is  to  be  ignored,  since  OrcaFlex  will  not  accept  two  columns   specifying  the  same  information.  

253  

w

 

System  Modelling:  Data  and  Results,  Vessels    

x

The  remaining  lines  in  the  table  must  contain  numbers,  one  for  each  header  in  the   headers  line,  separated  by   tabs   and/or   spaces.  Please  note  that   it   is   the   order   of   the   columns  that   matters,   not   their   actual   position   across   the   page.   Hence,   although   it   is   natural   to   align   the   headers   above   the   columns   of   numbers,   this   is   not   in   fact   necessary.  

x

The  conventions  and  units  used  in  the  text  file  can  be  specified  Ȃ  see  below.  Doing  so  allows  the  imported  data   to  be  converted  to  match  the  vessel  type  conventions  and  scaled  to  match  the  units  of  the  OrcaFlex  model.  

x

The  RAO  and  phase  origins  are  not  read  in  and  should  be  set  on  the  Vessel  Type  form.  

Specifying  conventions  and  units  

You   can   add  markup   to   the  file   to   specify   the   conventions   and   units  of   the   data   contained   in   the   file.   For  example,   a   typical  markup  block  for  conventions  is  as  follows:   ***  OrcaFlex  Conventions  Start  ***   RAOResponseUnits  =  degrees   RAOWaveUnit  =  amplitude   RAOPhaseConvention  =  lags   RAOPhaseUnitsConvention  =  degrees   RAOPhaseRelativeToConvention  =  crest   SurgePositive  =  forward   SwayPositive  =  port   HeavePositive  =  up   RollPositiveStarboard  =  down   PitchPositiveBow  =  down   YawPositiveBow  =  port   ***  OrcaFlex  Conventions  End  ***  

The   block   begins   with   a   line   containing   the   string   OrcaFlex   Conventions   Start   and   ends   with   a   line   containing   OrcaFlex   Conventions   End.   The   lines   in   between   specify   the   conventions   using   the   standard   OrcaFlex   batch   script   names  and  values.   The  above  list  contains  all  possible  conventions  settings  that  can  be  included  in  the  block.  Note  that  the   waves  are   referred  to  by  convention  is  not  allowed  because  the   WP,  WFH  or  WFR  header  string  identifies  the  period/frequency   convention.  In  addition  the  symmetry  convention  should  not  be  specified  in  this  block  Ȃ  instead  you  should  simply   set  it  to  the  appropiate  value  in  the  OrcaFlex  model.   If   some   conventions   are   omitted   then   OrcaFlex   uses   the   corresponding   value   from   the   OrcaFlex   model   conventions.   If  the  block  is  omitted  altogether  then  a  warning  is  issued  and  no  conventions  conversion  is  performed.   The  units  of  the  data  contained  in  the  file  are  specified  in  a  similar  way:   ***  OrcaFlex  Units  Start  ***   LengthUnits  =  m   ForceUnits  =  kN   ***  OrcaFlex  Units  End  ***  

The   above   list   contains   all   possible   units   settings   that   can   be   included   in   the   block.   For   displacement   RAOs   the   length  unit  is  the  only  one  which  is  used  and  consequently  the  force  units  can  b e  omitted.  For  load  RAOs  both  length   and  force  units  should  be  specified.   If   some   units   settings   are   omitted   then   OrcaFlex   uses   the   corresponding   value   from   the   OrcaFlex   model.   If   the   block   is  omitted  altogether  then  a  warning  is  issued  and  no  units  scaling  is  performed.   Moses  output  

See  this  example  for  Moses  displacement  RAOs,  and  this  one  for  Moses  wave  load  RAOs.  Remember  that  RAO  and   phase   origins   are   not   imported   from   text   files   Ȃ   you   should   take   particular   care   with   Moses   output   to   set   these   correctly,  since  RAO  origin  may  differ  from  phase  origin  in  this  case.   Note:  

Moses  reports   encounter  period  or  frequency,  to  account  for  the  effect  of  the  speed  of  the  vessel  on   the   apparent   wave   period   or   frequency.   OrcaFlex   requires   the   data   at   actual   wave   period   or   frequency,  so  if  possible  your  vessel  in  your  Moses  model  should  not  have  any  forward  speed.  

Hydrostar/ARIANE  output  

Hydrostar  has  an  option  to   output   results  in  a  format  suitable  for  import  directly  into  ARIANE,  and  this  format  is   also   suitable   for   import,   with   the   mark-­‐up   described   above,   into   OrcaFlex.   This   format   is   usually   indicated   by   the   phrase  'pour  ARIANE'  at  the  head  of  the  file.   This  example  demonstrates  the  necessary  mark-­‐up  and  the  setting  of  

254  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

conventions   and   units  for  displacement   RAOs.   Hydrostar   requires   that   the   'horizontal'   components   of   the   RAOs  are   listed  first,  then  the  'vertical'  components:  you  must  take  particular  care  here  to  order  the  column  headers  correctly.   Note:  

As  with  Moses,  Hydrostar  takes  account  of  the  effect  of  the  speed  of  the  vessel  in  determining  the   added  mass  and  damping,  and  reports  results  at  encounter  period  or  frequency.  Your  vessel  in  your   Hydrostar  model  should  not  have  any  forward  speed.  

Other  output  

OrcaFlex   can   import   data   produced  by  other   programs   (or,   more   generally,  text   data   obtained   by  any  other   means)   so  long  as  it  is  in  the  format  illustrated  by  the  examples  above,  ie  a  series  of  tables  of  data,  one  for  each  direction,   arranged   with   rows   representing   wave   period   or   frequency   and   columns   for   surge/sway/heave/roll/pitch/yaw   dimensional  RAO  values.  

Header  Strings  for  Text  RAO  Tables   When  importing  RAOs  from  a  text  file,  the  following  strings  can  be  used  in  the  header  line.   Header  string   Column  contains   WP  

Wave  period  in  seconds  

WFH  

Wave  frequency  in  Hertz  

WFR  

Angular  wave  frequency  in  radians/second  

XA  

Surge  amplitude  

XP  

Surge  phase  

YA  

Sway  amplitude  

YP  

Sway  phase  

ZA  

Heave  amplitude  

ZP  

Heave  phase  

RXA  

Roll  amplitude  

RXP  

Roll  phase  

RYA  

Pitch  amplitude  

RYP  

Pitch  phase  

RZA  

Yaw  amplitude  

RZP  

Yaw  phase  

In  these  header  strings   X,   Y  and  Z  represent  the  vessel  axes,   A  denotes  amplitude,   P  denotes  phase  and   R  rotation   about  the  given  axis.  

RAO  Data  Checklist   To   derive  vessel  point   motions,   you   need   to   obtain   data   giving   both  RAOs  and   phases   for   the   vessel   for   the   relevant   wave   period.   You  also   need  to   know  what  conventions  apply  to   your   data;   these   may  be   documented   with  the   data,   but  sometimes  you  may  have  to  deduce  what  they  are.  You  should  have  answers  to  all  the  following  questions:   To  what  point  on  the  vessel  do  the  data  apply?  

This  is  the  RAO  origin  and  is  often  the  vessel  centre  of  gravity,  but  you  need  to  be  sure.  If  it  is  not  specified  check   with  your  data  supplier.   To  what  point  on  the  vessel  are  the  phases  relative?  

This  is  the  RAO  phase  origin  and  is  usually,  but  not  always,  the  same  as  the  RAO  origin.   Are  the  responses  in  dimensional  or  RAO  form?  

RAO   form   (i.e.   for   unit  wave   amplitude)   is  the   most  common;   data   giving   dimensional   form   would   have  to   also   give   the  associated  wave  amplitudes/heights.  OrcaFlex  will  only  accept  RAO  form.   In  what  form  are  the  rotational  roll,  pitch  and  yaw  RAOs?  

Units  such  as  degrees/metre  or  radians/metre  (displacement  RAOs),  or  kN.m/m  (wave  load  RAOs),  almost  always   mean  the  rotational  motions  are  relative  to  waves  of  unit  amplitude.  

255  

w

 

System  Modelling:  Data  and  Results,  Vessels    

Very   rarely,   rotational   RAO   amplitudes   are   given   per   unit   wave   height   (i.e.   double   amplitude)   Ȃ   check   your   data   source.  In  this  case  you  will  have  to  multiply  the  RAOs  by  2  manually,  before  entry  to  OrcaFlex.   Units   such  as   degrees/degree,  radians/radian,   or   no   units   (displacement   RAOs),   or   kN.m/radian   (wave   load   RAOs),   imply  rotational  RAOs  relative  to  waves  of  unit  steepness  or  maximum  slope.   For  long   wave  periods  in  deep  water,  the  rotational  displacement  RAOs  in  the  wave   plane  (e.g.  pitch  in  head  or  stern   seas)  should  tend  to  1  for  RAOs  relative  to  unit  maximum  slope,  or  to  pi  for  RAOs  relative  to  unit  steepness.   Are  the  phases  in  degrees  or  radians?  

Unless  you  only  have  a  small  amount  of  data,  this  should  be  obvious  from  the  range  of  phase  values.   What  directions  are  positive  for  surge,  sway,  heave,  roll,  pitch  and  y aw?  

Often  they  are  surge  positive  forward,  sway   positive  to   port,  heave  positive  up,   but   some  authors  use  heave  positive   downwards.   Roll,   pitch   and   yaw   are   usually   positive   when   clockwise   about   the   positive   surge,   sway   and   heave   directions.   Most  data  sources  use  right-­‐handed  axes,  but  not  all.  OrcaFlex  allows  complete  generality  in  its  data  input,  but  you   must  find  out  how  your  data  are  defined.   To  what  phase  time  origin  are  the  phases  relative?  

OrcaFlex   allows   you   to   specify   that   the   phases  to  be   relative   to   the   time   the   wave  crest,   trough,   zero   up-­‐crossing   or   zero  down-­‐crossing  passes  the  phase  origin.   The  passage  of  the  crest   past  the   RAO  origin  is  the  most  common  phase   time  origin,  but  you  need  to  check  and  tell  OrcaFlex  Ȃ  see  note  on  phase  leads/lags  below.   Are  the  phases  leads  or  lags?  

Phase   conventions   are   sometimes   documented   by   giving   the   formula   used   to   represent   the   harmonic   motion.   Commonly  used  ones  are:   x

Ǥ…‘•ȋɘ–-­‐  P)  or  A.cos(P  -­‐  ɘ–Ȍ‹’Ž›–Šƒ–’Šƒ•‡‹•ƒŽƒ‰Ǥ  

x

…‘•ȋɘ–Ϊ̵Ȍ‹’Ž‹‡•–Šƒ–’Šƒ•‡̵‹•ƒŽ‡ƒ†Ǥ  

Using  sin  rather  than  cos  in  the  above  formulae  has  no  effect  on  whether  the  phases  are  leads  or  lags.  

Checking  RAOs   The   Check   RAOs   button   on   the   vessel   types   form   allows   a   visual   check   on   the   RAO   data   for   either   displacement   RAOs  or  wave  load  RAOs.  For  a  given  draught  and  wave  direction,  it  displays  graphs  (one  for  each  vessel  degree  of   freedom)  showing  how  the  RAO  and  phase  vary  with  wave  period.   There  are  3  types  of  graph  available:   x

Complex  Values  

x

Amplitude  

x

Phase  

Amplitude  and  Phase  Graphs  

These   graphs   provide   a   straightforward   graphical   representation   of  the   RAO   data   as   input   on   the   Vessel   Types   data   form.   The   amplitude   or   phase   is   plotted   on   the   Y   axis   of   the   graph.   For   the   X   axis   you   have   the   choice   of   plotting   period,  frequency  in  rad/s  or  frequency  in  Hz.   Complex  Value  Graphs  

The   graphs   initially   show   the   RAOs   for   the   currently   selected   draught   and   direction.   You   can   switch   to   other   draughts  and  directions,  either  by  using  the  navigation  buttons  at  the  bottom  of  the  form  to  step  through  the  data,   else   or   by   selecting   from   the   drop-­‐down   lists.   You   can   change   the   scale   of  the   graphs   (double   click   on   the   graph   and   change  the  ranges  of  the  axes).  This  is  useful  if  the  curve  does  not  initially  fit  on  the  graph.  

256  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

R

I 0 0 Figure:  

 

T

 

The  graphs  depicts  the  RAO  data  specified  by  the  user  for  the  specified  RAO  origin.  The  graph  has  two  parts:   x

A  curve  showing  the  RAO  data  specified  by  the  user  as  a  series  of  points  joined  in   order  of  increasing  period.   The  curve  starts  from  the  'short'  wave  response,  which  should  have  zero  or  very  small  amplitude,  so  the  curve   should   start   from   near   the   origin.   Moving   along   the   curve   away   from   the   origin   corresponds   to   the   wave   period   increasing  from  zero.  For  surge,  sway  and  heave,  the  other  end  of  the  curve  is  the  'long'  wave  RAO  data  specified   for  period  'Infinity'.  For  roll,  pitch  and  yaw,  the  RAO  data  for  period  'Infinity'  cannot  (for  technical  reasons)  be   included  in  the  curve,  so  instead  the  other  end  of  the  curve  is  the  RAO  data  for  the  largest  finite  period  specified.  

x

A  solid  circle  representing  the   expected  long   wave   response  limit  for  a  freely  floating  vessel.  See   RAO  Quality   Checks  for  details  of  the  expected  long  wave  RAOs.   Warning:  

The  expected  long  wave  response  limits  calculated  by  OrcaFlex  only  apply  to  free-­‐floating  vessels.   Also,   the   yaw   response   limit   only   applies   to   slender   vessels   (i.e.   vessels   that   are   long   in   the   x-­‐ direction  and  narrow  in  the  y-­‐direction).  

The  purpose  of  the  graph  is  help  you  check  your  RAO  data   Ȃ  the  curve  should  normally  be  reasonably  smooth  and   tend  towards  the  expected  limit  shown  by  the  solid  circle.  See   How  to  Check  RAO  Data  for  details.   The  graph  repr‡•‡–••ƒ•’‘‹–•‹’‘Žƒ”…‘‘”†‹ƒ–‡•ȋǡɔȌǡ™Š‡”‡ǣ   x

R   is   the   non-­‐dimensional   amplitude.   For   surge,   sway   and   heave   R   is   the   vessel   motion   amplitude   divided   by   the   wave  amplitude.  And  for  roll,  pitch  and  yaw,  R  is  the  rotational  response  normalised  with  respect  to  maximum   wave  slope  Ȃ  i.e.  it  is  vessel  rotation  amplitude  divided  by  the  maximum  wave  slope.  

x

ɔ ‹• –Š‡ ’Šƒ•‡ lag,   from   the   time   the   wave   crest   passes   the   user-­‐specified   phase   origin   until   the   maximum   positive  motion  occurs.   Note:  

Positive  here  means  as  in  the  OrcaFlex  conventions  (not  necessarily  the  same  as  the  vessel  type  RAO   conventions).  So  positive  surge  is  forward,  positive  sway  is  to  port,  positive  heave  is  up,  positive  roll   is  starboard  down,  positive  pitch  is  bow  down  and  positive  yaw  is  bow  to  port.  

This  polar  coordinates  way   of  representing  RAOs  is  better  than   drawing  separate   graphs  of  amplitude  and  phase,   since   it   presents   all   the   information   on   a   single   graph   and   also   the   resulting   curves   are   smooth,   whereas   phase   graphs  frequently  show  phase  jumps.  

How  to  Check  RAOs   For   each   draught   and   wave   direction,   you   should   check   that   the   curves   on   the   Complex   Value   RAO   graphs   are   reasonably   smooth   and   approach   the   circle,   which   is   the   expected   long-­‐wave   limit   for   a   free-­‐floating   vessel.   Note   that:   x

The  curve  may  not  approach  the  expected  long  wave  limit  if  the  RAO  data  does  not  include  values  for  any  long   waves.   Wave   periods   over   20   seconds   for   ships,   or   30   seconds   for   semisubmersibles,   are   considered   to   be   sufficiently  long  for  this  purpose.  

x

The   curve   might   also   not   approach   the   circle   if   the   vessel   is   not   free-­‐floating.   For   example   the   heave   displacement  RAO  amplitude  of  a  tension  leg  platform  will  not  approach  the  usual  long  wave  limit  of  1.  

257  

w

 

System  Modelling:  Data  and  Results,  Vessels    

x

The   circle   on   the   yaw   graph   only   applies   to   slender   vessels   (i.e.   long   in   the   x-­‐direction   and   narrow   in   the   y-­‐ direction).  

x

Smooth  graphs  can  only  be  expected  if  the  data  includes  RAOs  for  reasonably  closely  spaced  periods.  

As  examples,  consider  the  following  three  example  graphs:  

258  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

1.5

1

0.5

0

-0.5

-1

-1.5 -1.5

-1

-0.5

0

0.5

1

1.5

-1

-0.5

0

0.5

1

1.5

1.5

1

0.5

0

-0.5

-1

-1.5 -1.5

259  

w

 

System  Modelling:  Data  and  Results,  Vessels    

1.5

1

0.5

0

-0.5

-1

-1.5 -1.5

-1

-0.5

0

0.5

1

1.5

  The  first  graph  shows  a  typical,  well-­‐behaved  set  of  displacement  RAO  data   Ȃ  the  curve  is  smooth  and  the  long-­‐wave   limit  agrees  with  the  expected  value  marked  by  the  circle.   For  a  freely  floating  vessel,  the  second  graph  is  clearly  in  error,  since  the  curve  does  not  lead  to  the  expected  long   wave   limit.   The   RAO   data   for   long   waves   (represented   by   the   end   of   the   curve)   has   the   correct   amplitude,   but   its   phase  differs  by  180°  from  the  expected  long-­‐wave  value  (represented  by  the  circle).  There  are  two  likely  causes  Ȃ  it   may  be  that  the  phase  lead/lag  convention  data  has  been  set  wrongly  (this  would  give  a  phase  angle  sign  error)  or   else   that   the   convention   data   for   the   direction   of   positive   motion   has   been   set   wrongly   (this   would   give   a   phase   error  of  180°).   The  curve  on  the  third  graph  approaches  the  expected  long  wave  limit,  but  then  suddenly  goes  to  zero.  This  suggests   that  the  RAO  data  for  period  'Infinity'  has  not  been  set  correctly  and  is  zero.   Common  Problems  

It   is   not  unusual   to  be   given   RAO  data   for   a   vessel   but   not   be   given   all   the   conventions   that   apply   to   the   data.   Below   are   some   common   problems   and   their   symptoms.   But   beware   that   several   common   problems   have   very   similar   symptoms,   so   it   is   not   possible   to   be   sure   what   the   problem   is   unless   you   are   sure   about   most   of   the   data's   conventions  and  only  unsure  about  one.  It  is  therefore  important  to  get  as  much  information  as  possible  from  the   original  RAO  data  supplier.   x

The   quoted   wave  direction   might  be  measured  clockwise  (viewed  from  above)  from   the  x-­‐direction,   rather  than   anticlockwise   (which   is   the   OrcaFlex   convention).   The   effect   would   be   a   180°   shift   in   the   sway,   roll   and   yaw   phases.  

x

The   quoted   wave   direction   may   be   the   direction   the   wave   is   coming   from,   rather   than   the   direction   it   is   progressing  towards  (which  is  the  OrcaFlex  convention).  The  effect  would  be  to  negate  all  the  phase  values.  

x

The  phases  may  be  leads  instead  of  lags  (OrcaFlex  will  accept  either  Ȃ  see  RAO  Phase  Conventions).  The  effect  of   an  error  here  would  be  to  negate  all  the  phase  values.  

Importing  QTFs   Wave   drift   QTF   data   may   be   imported   using   the   Import   QTFs   button   on   the   vessel   types   form.   Data   may   be   imported  directly  from  AQWA  or  WAMIT  output  files,  from  NMIWAVE  output  files,  and,  with  the  addition  of  some   markup  text,  from  Moses  and  Hydrostar/ARIANE  output  files.  The  form  of  this  markup  also   allows  for  import  from   other  text  files.   Import  QTFs  from  specific  program  output   OrcaFlex   can   import   QTFs   directly   from   AQWA   and   WAMIT   output   files,   as   described   under   Importing   Hydrodynamic  Data.  All  existing  QTF  data  for  the  selected  draught  will  be  overwritten.  

260  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

NMIWAVE  files  

NMIWAVE   text  output   files   may   be   imported   with   the   addition   of  a   single   line   containing   the   string  NMIWAVE  Wave   Drift  at  the  start  of  the  file,  as  in  this  example   NMIWAVE   uses   the   ITTC   conventions,   which   are   surge   +ve   forward,   sway   +ve   to   starboard,   heave   +ve   down,   roll   +ve   starboard   down,   pitch  +ve   bow   up,   yaw   +ve   bow   to  starboard.   As   with  AQWA   and   WAMIT,  OrcaFlex   recognizes   the   type   of   file   and   'knows'   what   these   conventions   are,   so   you   do   not   need   to   specify   them   explicitly   in   the   file   before  import.  The  imported  data  is  converted  from  these  conventions  to  match  the  vessel  type   conventions.   The   wave   heading   convention   used   by   NMIWAVE   is   that   wave   heading   is   measured   +ve   clockwise   when   viewed   from   above,   and   zero   wave   heading   means   a   stern   wave.   This   is   the   opposite   sense   to   OrcaFlex,   and   is   handled   automatically  on  import  by  changing  the  sign  of  the  wave  headings  when  an  NMIWAVE  file  is  imported.   Also,  note  that  NMIWAVE  QTFs  are  non-­‐dimensional:  OrcaFlex  will  re-­‐dimensionalise  them  on  import,  in  the  same   way  as  for  WAMIT.   Import  QTFs  from  marked-­‐up  text  files   QTF   data   can   be   imported   from   generic   text   files   using   a   similar   format   to   that   for   RAOs.   Only   the   direction   conventions  are  required  in  this  case:   ***  OrcaFlex  Conventions  Start  ***   SurgePositive  =  forward   SwayPositive  =  port   HeavePositive  =  up   RollPositiveStarboard  =  down   PitchPositiveBow  =  down   YawPositiveBow  =  port   ***  OrcaFlex  Conventions  End  ***  

Since   QTFs   contain   no   phase   information,   the   conventions   relating   to   phase   have   no   relevance;   and   all   QTFs   are   assumed   to   be   given   relative   to   waves   of   unit   amplitude,   so   the   RAOWaveUnit   convention   is   implicitly   set   to   be   "amplitude".   The  units  of  the  data  contained  in  the  file  are  specified  in  the  same  way:   ***  OrcaFlex  Units  Start  ***   LengthUnits  =  m   ForceUnits  =  kN   ***  OrcaFlex  Units  End  ***  

The  QTF  format  differs  from  that  for  RAOs  as  follows:   x

The   identifier   line   is   OrcaFlex   QTF   Start   rather   than   OrcaFlex   Displacement   RAO   Start   or   OrcaFlex   Load   RAO   Start.  

x

The  column  headers  are  surge,  sway,  heave,  roll,  pitch  and  yaw  instead  of  X,  Y,  Z,  RX,  RY  and  RZ  respectively.  

x

Phase  data  are  not  specified.  

Otherwise,  the  format  follows  that  for   importing  RAOs.   The  QTF  origin  is  not   read  in  and  should  be  set  on  the  Vessel   Type  form.   Moses  output  

QTFs  in  all  six  degrees  of  freedom  are  included  in  this  example  of  Moses  output.   Note:  

Moses  reports   encounter  period  or  frequency,  to  account  for  the  effect  of  the  speed  of  the  vessel  on   the   apparent   wave   period   or   frequency.   OrcaFlex   requires   the   data   at   actual   wave   period   or   frequency,  so  if  possible  your  vessel  in  your  Moses  model  should  not  have  any  forward  speed.  

Hydrostar/ARIANE  output  

As  is  the  case  with  RAOs,  Hydrostar's  output  format  'pour  ARIANE'  is  suitable,  with  the  mark-­‐up  described  above,  for   import  into  OrcaFlex.   This  example  demonstrates  the  necessary  mark-­‐up  and  the  setting  of  conventions  and  units   for  horizontal-­‐plane  QTFs  (surge,  sway  and  yaw).  OrcaFlex  will  set  the  data  for  the  degrees  of  freedom  which  are  not   imported  (so  heave,  roll  and  pitch  in  this  case)  to  zero.  Data  in  this  form  are  common  for  vessels  with  large  water   plane  areas,  such  as  barges,  tankers,  etc.  

261  

w

 

System  Modelling:  Data  and  Results,  Vessels    

Note:  

As  with  Moses,  Hydrostar  takes  account  of  the  effect  of  the  speed  of  the  vessel  in  determining  the   added  mass  and  damping,  and  reports  results  at  encounter  period  or  frequency.  Your  vessel  in  your   Hydrostar  model  should  not  have  any  forward  speed.  

Other  output  

OrcaFlex   can   import   data   produced  by  other   programs   (or,   more   generally,  text   data   obtained   by  any  other   means)   so  long  as  it  is  in  the  format  illustrated  by  the  examples  above,  ie  a  series  of  tables  of  data,  one  for  each  direction,   arranged   with   rows   representing   wave   period   or   frequency   and   columns   for   surge/sway/heave/roll/pitch/yaw   dimensional  QTF  amplitudes.  As  in  the   Hydrostar/ARIANE  example  above,  any  missing  columns   will  have  their   data   set  to  zero.  

Importing  Added  Mass  and  Damping  Data   Added   mass  and  damping   matrices  may   be  imported   using  the   Import  Matrices   button  on  the  vessel  types  form.   Data   may   be   imported   directly   from  AQWA   or   WAMIT   output   files,   and   with   the   addition   of  some   markup   text   from   Moses,  Hydrostar/ARIANE  and  Wadam  output  files.  The  form  of  this  markup  also  allows  for  import  of  reasonably-­‐ formatted  6x6  added  mass  and  damping  matrices  from  any  generic  text  file.   Direct  import  from  specific  program  output   OrcaFlex   can   import   data   directly   from   AQWA   and   WAMIT   output   files,   as   described   under   Importing   Hydrodynamic  Data.  Note,  however,  that  in  this  case  the   reference  origin  is  not  imported.  All  existing  added   mass   and  damping  data  for  the  selected  draught  will  be  overwritten.   Import  from  marked-­‐up  text  files   With   the   addition   of   some   tags   indicating   the   start   and  end  of   the   data,   and   others  defining   which  vessel   properties   should   receive   the   imported   data,   data   may   be   imported   from   the   plain   ASCII   text   files   which   are   written   by   a   number   of   analysis  programs   in   addition   to   AQWA   and   WAMIT.   The   form   of   these   tags   is  given   below   for   each  such   program.   Common   to   each   is   a   line   of   the   form   "Draught   [DraughtName]"   specifying   the   vessel   type   draught   into   which  the  data  are  to  be  imported.  If  DraughtName  already  exists,  the  existing  added  mass  and  damping  data  will  be   overwritten;  if  not,  then  it  will  be  created  and  all  the  other  data  will  be  set  to  those  of  the  OrcaFlex  default  vessel   type.   The   conventions   used   in   the   text  file   can  be   specified   in  exactly  the   same  way  as   for  RAO  and  QTF  import.   Doing   so   allows   the   imported   data   to   be   converted   to   match   the   vessel   type   conventions.   Note   that,   as   for   QTFs,   only   the   direction   conventions   apply   to   the   added   mass   and   damping   data   and   so   only   these   conventions   need   to   be   specified.  So  a  typical  conventions  block  might  look  like  this:   ***  OrcaFlex  Conventions  Start  ***   SurgePositive  =  forward   SwayPositive  =  port   HeavePositive  =  up   RollPositiveStarboard  =  down   PitchPositiveBow  =  down   YawPositiveBow  =  port   ***  OrcaFlex  Conventions  End  ***  

You   can   also   specify   scaling   factors   to   allow   for   the   different   ways   in   which   various   programs   scale   their   output.   OrcaFlex  requires  unscaled,   fully-­‐dimensional  values  for   added   mass  and  damping   matrices,  so  some  data  may  need   to   be   un-­‐scaled   or   re-­‐dimensionalised.   You   may   also   use   these   scaling   factors   to   allow   for   differences   in   units   systems  Ȃ  unlike  RAOs,  OrcaFlex  does  not  read  units  definitions  and  convert  them  automatically  for  added  mass  and   damping  matrices.  These  scaling  factors  are  described  in  the  sections  below.   Moses  output  

Added  mass  and  damping  may  be  output  in  two  forms  by  Moses:   (i)  Diagonal  elements  only   The  standard  Moses  output  file  may  contain  tables  of  the  diagonal  elements  (surge-­‐surge,  sway-­‐sway,  yaw-­‐yaw  etc   terms   only)   of  the   added   mass   and   damping   matrices,   listed   against   period  or   frequency.   OrcaFlex   will   import   these   values,   setting   the   off-­‐diagonal   elements   to   zero.   These   data   are   identified   to   OrcaFlex   by   the   addition   of   a   line   containing   the   phrase   "OrcaFlex   Added   Mass   Table   Start"   (analogously   for   Damping),   followed   by   a   draught   specifier  and  a  line  containing  one   of  the  strings  "WP",   "WFR"  or  "WFH"  to  indicate  that  the  data  are  given  in   the   Moses   file   by   period,   frequency   in   radians/sec,   or   frequency   in   Hz   respectively.   These   lines   must   immediately  

262  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

precede   the   table   of   data,   after   any   column   headings,   and   each   table   must   be   terminated   by   a   corresponding   "OrcaFlex  Added  Mass  Table  End"  (or  Damping)  line.  The  use  of  this  markup  is  illustrated  in  this  truncated  example   (ii)  Full  matrices   Moses   will   also   optionally   output   the   full   6x6   matrices   to   a   PPO   file:   in   general,   if   you   have   this   file   you   should   import   the   data   from   here   in   preference   to   the   diagonal   values   only.   In   this   case,   the   required   tags   are   "OrcaFlex   Added   Mass   Matrix   Start"   (again,   analogously   for   Damping),   followed   by   a   draught   specifier   and   one   of   "WP   x",   "WFR   x"   or   "WFH   x",   where   x   is   the   value   of   the   period   or   frequency,   at   the   beginning   of   each   matrix,   and   the   corresponding  End  tag  immediately  after  each  matrix.  This  sample  demonstrates  the  use  of  these  tags.   In  both  cases,  Moses  presents  added  mass  and  damping  values  which  are  scaled  by  mass.  Since  OrcaFlex  requires   non-­‐normalized   data,   the   import   process   must   account   for   this   scaling   factor.   To   do   so,   note   that   towards   the   beginning  of  each  of  the  two  examples  above  is  a  section  of  the  form   OrcaFlex  Scaling  Factor  Start  1381.8  OrcaFlex  Scaling  Factor  End     The   number   1381.8   here   is   the   mass   of  the   vessel,   as   specified   in   the   standard   output   file.   OrcaFlex   will   scale   all   the   imported  data  by  this  value.   Note:  

Moses  reports   encounter  period  or  frequency,  to  account  for  the  effect  of  the  speed  of  the  vessel  on   the   apparent   wave   period   or   frequency.   OrcaFlex   requires   the   data   at   actual   wave   period   or   frequency,  so  if  possible  your  vessel  in  your  Moses  model  should  not  have  any  forward  speed.  

Hydrostar/ARIANE  output  

Full  6x6  added  mass  and  damping  matrices  are  output,  without   any  normalising  or  scaling  factors.  Other  than  the   lack  of  scaling  factors,  the  tags  required  by   OrcaFlex  are  the  same  as  those  for  Moses  6x6  matrices:  you   will  need   Start   and   End   tags   surrounding   each   matrix  and   a   " Draught"   line,   and   a   "WP   x",   "WFR   x"   or   "WFH   x"   line,   where   x   is   the  value   of  the   period  or  frequency,  at  the   beginning  of  each  matrix.  Note   that  OrcaFlex  will  allow,  as  necessary,  for   the   row  and  column  headings  1,  2,   ...,  6  if  they  are  present.  A  short   example  of  marked-­‐up  Hydrostar   output  is  given.   You   should   be   aware   that   Hydrostar   results   may   be   given   in   either   Hydrostar's   own   axis   conventions   or   those   of   ARIANE,   and   that   the   two   differ.   The   latter   is   usually   indicated   by   the   phrase   'pour   ARIANE'   in   the   file:   the   conventions  in  the  example  here  are  those  for  ARIANE.   Note:  

As  with  Moses,  Hydrostar  takes  account  of  the  effect  of  the  speed  of  the  vessel  in  determining  the   added  mass  and  damping,  and  reports  results  at  encounter  period  or  frequency.  Your  vessel  in  your   Hydrostar  model  should  not  have  any  forward  speed.  

WADAM  output  

WADAM   also   outputs   full   6x6   added   mass   and   damping   matrices,   but  these   a re  non-­‐dimensional.   The   tags   required   for   each   matrix   are   as   for   Hydrostar/ARIANE   as   above;   in   addition,   the   non-­‐dimensionalising   factors   must   be   specified  in  the  file.  Since  the  data  are  fully  non-­‐dimensional,  the  scaling  is   rather  more  complex  than  Moses'  scaling   by   mass   alone:   the   factors   differ   between   the   added   mass   and   damping   matrices,   and   each   matrix   requires   a   different  factor  for  each  constituent  3x3  sub-­‐matrix  (since  their  units  differ).  Full  details  of  the  calculation  of  these   factors   are   given  in   the  WADAM   output   file   itself   (search  for   the   string   'non-­‐dimensional');   see  this  edited   example,   which  shows  the  relevant  text  and  the  corresponding  markup  text  required  by  OrcaFlex.   Note:  

The  conventions  are  not  specified  in  this  example,  as  we  (Orcina)  do  not  have  any  information  on   them.  If  you  are  able  to  let  us  know  what  they  are,  and  help  us  improve  this  example,  then  please  do   get  in  touch  with  us.  

Output  from  other  programs  

OrcaFlex  should  be  able  to  import  added  mass  and  damping  data  from  other  programs,  not  listed  above,  so  long  as   they  are  presented  in  text  files  as  6x6  matrices.  As  in  the  above   examples,  you  will  need   to:  add  the  appropriate  text   strings   in   the   file   to   delimit   the   data,   nominate   the   draught  into   which  the   data   are   to   be   imported,   and   indicate  the   wave  period  or  frequency  for  each  matrix.  If  the  matrices  are  non-­‐dimensional,  either  partially  or  fully,  you  will  need   to  enter  the  scaling  factors  into  the  file.  Note  that  these  factors  may  be  entered  multiple  times  and  will  be  updated   each  time  Ȃ  this  may  be  useful  if,  say,  they  depend  on  frequency.  

6.7.3

Modelling  Vessel  Slow  Drift  

When   a   vessel   is   exposed   to   waves   it   experiences   wave   loads   that   can   be   split   into   first   order   and   second   order   terms.   The   first   order   terms   generate   motion   at   wave   frequency   and   this   is   modelled   in   OrcaFlex   using   RAOs   to   specify   either   the   displacement   or   the   load.   The   second   order   terms   are   much  smaller   but  they  include   loads  with  a  

263  

w

 

System  Modelling:  Data  and  Results,  Vessels    

much   lower   frequency.   These   low   frequency   terms   are   called   the   wave   drift   loads   and   they   can   cause   significant   slow  drift  motions  of  the  vessel  if  their  frequencies  are  close  to  a  natural  frequency  of  the  vessel.   One   common   situation   where   the   wave   drift   loads   can   matter   is   with   a   moored   vessel.   The   vessel's   natural   frequencies  in  surge,  sway  and  yaw  are  typically  quite  low  and  so  the  low  frequency  wave  drift  loads  can  generate   quite  significant  slow  drift  excursions.   If   you   have   already   calculated   the   vessel   slow   drift   motion   then   that   motion   can   be   applied   in   OrcaFlex   using   harmonic  motion  or  a  time  history  file.  But  OrcaFlex  can  calculate  and  apply  the  slow  drift  motion  for  you.  To  do  this   you  need  to  do  the  following:   x

Specify  QTF  data  on  the   wave  drift  page  of  the  vessel  type  form  (the   wave  drift  loads  are  calculated  based  on   this  data).  Also,  ensure  that   Wave   Drift  Load  (2nd  Order)   is  in  the  vessel's  Included  Effects.  This  tells  OrcaFlex   to  apply  the  mean  wave  drift  load  to  the  vessel  during  the  static  analysis,  and  then  in  the  dynamic  analysis  to   apply  the  time  varying  wave  drift  load.  

x

Optionally,  specify  appropriate  data  for  hydrodynamic  and  wind  damping  and  any  applied  load,  etc.,  and  include   these   loads   in   the   vessel's   Included   Effects.   OrcaFlex   will   automatically   include   loads   from   added   mass   and   damping  and  loads  from  any  lines  or  other  objects  that  are  connected  to  the  vessel.  

x

On  the  Structure  page  on  the  vessel  type  form,  specify  the  vessel  centre  of  gravity,  mass,  moments  of  inertia  data   for  the  appropriate  draught.  And  on  the   Stiffness,  Added  Mass  and  Damping  page,  specify  the  added  mass  and   damping   matrices   and   the   reference   origin   to   which   they   apply.   The   stiffness   and   hydrostatic   equilibrium   position  data  are  not  required  for  3DOF  analyses,  since  they  only  apply  in  the  heave,  roll  and  pitch  directions.  

x

On   the   vessel   form,   include   3   degrees   of   freedom   in   the   static   analysis.   The   OrcaFlex   static   analysis   will   then   calculate   the   equilibrium   position   allowing   for   the   mean   wave   drift   load.   And   set   the   primary   motion   to   Calculated   (3DOF).   The   OrcaFlex   simulation   will   then   calculate   the   vessel   surge,   sway   and   yaw   motion   that   results.  

x

Set   the   vessel's   superimposed   motion   according   to   whether   and   how   you   want   to   model   first   order   wave   frequency  motion.  

In  the   dynamic   simulation   OrcaFlex   will   then   calculate   all   the   loads   on   the   vessel   and   the   resulting   slow   surge,   sway   and  yaw  motion.   Damping  Effects  on  Vessel  Slow  Drift     Drag   and   damping   loads   have   an   important   effect   on   vessel   slow   drift   motions.   The   following   discussion   documents   the  various  damping  effects  and  how  they  are  modelled  in  OrcaFlex.  See   CMPT  (1998)  section  3.12.   x

Hydrodynamic  drag  and  skin  friction  on  the  vessel  hull.  This  is  modelled  in  OrcaFlex  using  a  combination  of   the  OCIMF  approach,  a  yaw  drag  moment  proportional  to  (yaw  rate) 2  and  roll  damping  terms  (both  linear  and   quadratic).   See   the   Hydrodynamic   Damping   data   on   the   vessel   type   data   form.   For   details   of   the   theory   see   Vessel   Theory:  Hydrodynamic  and   Wind   Damping.  Note   that  OrcaFlex  does   not  yet  have  the  dependency  of  yaw   drag  on  sway  velocity  proposed  by  Wichers,  1979.  

x

Wind   drag  on   the   vessel   hull.   This   is   the   aerodynamic   drag   due   to   wind   and   any   vessel   velocity.   It   is   modelled   in   OrcaFlex   based   on   the   OCIMF   approach.   See   the  Wind   Damping   data   on   the   vessel   type   data   form.   For   details   of  the  theory  see  Vessel  Theory:  Hydrodynamic  and  Wind  Damping.  

x

Hydrodynamic   drag   on   the   risers/moorings.   This   is   modelled   in   OrcaFlex   by   the   drag   force   part   of   the   Morison  force  on  the  lines  that  model  the  risers/moorings.  

x

Wave  radiation  damping.  This  is  not  usually  very  significant  at  low  frequencies,  because  the  asymptotic  limit   of  the  wave  frequency  damping  is  zero.  It  can  be  modelled  in  OrcaFlex  using  the  constant  damping  matrix  on  the   vessel  type  form.  

x

Wave  drift  damping.  This  arises  because  the  wave  drift  loads  vary  with  vessel  velocity.  It  can  be  modelled  in   OrcaFlex   by   including   it  in   the   constant  damping   matrix   on   the   vessel   type   data   form.   See  CMPT   (1998)   page   3-­‐ 78  and  Faltinsen  (1990)  page  161.  

x

Material   damping   in   the   risers/moorings.   This   is   the   structural   damping   in   the   material   of   the   risers   and   mooring  lines.  This  can  be  modelled  in  OrcaFlex  by  the  line   target  damping  value.  However  Triantafyllou  et  al   (1994)  concluded  that  its  effect  is  negligible.  

x

Seabed   soil   friction   on   the   risers/moorings.   This   arises   from   the   frictional   force   acting   on   the   part   of   a   mooring/riser   that   is   lifting   off   and   touching   down   on   the   seabed.   It   is   modelled   in   OrcaFlex   by   the   friction  

264  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

between   the   seabed   and   the   line   used   to   model   the   mooring/riser.   However   Triantafyllou   et   al   (1994)   concluded  that  its  effect  is  negligible.  

6.7.4

Vessel  Response  Reports  

The   vessel   response   window   can   be  opened   using  the  Results   Menu  or   using   the  Report   Vessel   Response  button   on   the  vessel  data  form.  It  offers  two  pages:  Displacement  Response,  and  Impulse  Response,  Added  Mass  and  Damping.   Generation   of   vessel   response   reports   can   be   automated   through   the   batch   script,   or   from   the   OrcaFlex   programming  interface.   Note:  

The   reported   RAOs   are   first   order   RAOs   calculated   using   small-­‐angle   theory   to   transform   the   RAOs   from  the  vessel  RAO  origin  to  the  points  specified  on  the  Vessel  Response  form.  Theses  results  are   accurate   for   small-­‐angle   rotational   motion,   but   if   the   vessel   undergoes   large-­‐angle   rotational   motion  it  will  exhibit  a  non-­‐linear  higher-­‐order  response  that  can  only  be  approximated  by  RAOs.   More   accurate   results   can   be   obtained   in   this   case   by   running   a   simulation,   which   uses   large   angle   theory,  and  this  can  often  be  done  very  quickly  by  removing  from  the  model  all  other  objects  that   do  not  affect  the  motion  of  the  vessel.  

Displacement  Response   The  Displacement  Response  page  gives  access  to  two   spreadsheet  reports,  one  giving  displacement  RAOs  and  one   giving   spectral   values,   of   the   vessel's   response   to   waves,   based   on   its   vessel   type's   displacement   RAOs.   It   can   be   accessed  from  the  Results  menu  or  from  the  Vessel  data  form.   Directions  and  Output  Points   You  specify  one  or  more  vessel-­‐relative  wave  directions,  and  one  or  more  points  on  the  vessel,  for  which  the  RAOs   and  spectral  response  will  be  reported.   If  you  specify  a  direction  of  '~'  then   the  direction  of  the   wave   relative  to  the  vessel   is  used.  If  there  is   more   than   one   wave  train  then  the  direction  of  the  first  wave  train  is  used.   The  output  points  are  specified  by  giving  their  coordinates  relative  to  vessel  axes.   Report  RAOs   The   RAO   report   gives   a   separate   worksheet   for   each   specified   direction,   containing   the   RAOs   for   each   of   the   specified  output  points,  derived  from  the  displacement  RAOs  of  its  vessel  type.   The  reported  RAOs  allow  for  the  following  effects:   x

They  allow  for  the  position  of  the  output  point  relative  to  the  RAO  origin  to  which  the  vessel  type  RAOs  apply.  

x

They  give   the   RAOs  for   the   specified   wave   direction,   relative   to   the   vessel.   If   this   is   not   one   of   the   directions   for   which  RAOs  are  specified,  then  the  RAOs  are  derived  by  interpolation  on  direction.  

x

They  allow   for  Froude   scaling  if   there   is  a   difference  between   the  vessel   type   length  and   the  vessel   length.  This   affects  the  periods  at  which  the  RAOs  are  reported.  

RAOs  are  given  for  the  position,  velocity  and  acceleration  of  the  output   points,  for  all  6  degrees  of  freedom  and   for  Z   Above  Wave.   The   RAO   report   covers   the   wave   periods   specified   on   the   vessel   type   data   form,   plus   (if   appropriate)   the   regular   wave  period  specified  on  the  environment  data  form.   Phase  Origin  

You  can  choose  the  Phase  Origin  to  use  for  reporting  phases,  the  options  being:   x

RAO   Phase   Origin.   The   phases   will   then   be   reported   relative   to   the   time   that   the   wave   (crest   or   trough,   as   specified  in  the  vessel  type  RAO  conventions)  passes  the  RAO  phase  origin  specified  on  the  vessel  type  form.  

x

Each  Point.  The  phases  will  then  be  reported  relative  to  the  time  the  wave  (crest  or  trough,  as  specified  in  the   vessel  type  RAO  conventions)  passes  that  particular  output  point.  

The  translational  RAOs  depend  on  the  position  of  the  output  point.  The  rotational  RAO  amplitudes  do  not  depend  on   the   position   of   the   output   point,   since   the   roll,   pitch   and   yaw   of   a   vessel   are   the   same   everywhere   on   it.   The   rotational  RAO  phases  depend  on  the  position  of  the  output  point  only  if  you  specify  the   Phase  Origin  to  be  Each   Point.  

265  

w

 

System  Modelling:  Data  and  Results,  Vessels    

Conventions  

You  can  choose  for  the  RAOs  to  be  reported  either  using  the  same   RAO  conventions  as  specified  for  this  vessel's  type   or  using  Orcina  standard  conventions.   Z  Above  Wave  

Z   Above   Wave   reports   the   heave   of   the   vessel   relative   to   the   water   surface.   This   can   be   useful   when   evaluating   whether  an  object  at  some  point  on  the  vessel  will  contact  the  water  surface  and  if  so  with  what  relative  velocity.   Warning:  

Z  Above  Wave  does  not  take  account  of  wave  surface  disturbance  due  to  the  presence  of  the  vessel.  

Report  Spectral  Response   The   spectral   response   report   contains   a   separate   worksheet   for   each   specified   direction   and   each   random   wave   train  specified  on  the  Environment  data  form.  The  worksheet  contains  tables  of  spectral  values  for  position,  velocity   and   acceleration,   for   all   6   degrees   of   freedom   and   for   Z   Above   Wave.   Separate   tables   are   given   for   each   of   the   specified  output  points.   Notes:  

The  spectral  response  report  is  not  available  if  no  random  wave  trains  are  specified.  

 

It   is   also   not   available   if   (for   any   of   the   specified   directions   and   for   any   degree   of   freedom)   the   vessel   has   a   non-­‐zero   RAO   amplitude   for   the   zero   wave   period   limit.   This   is   because   some   of   the   spectral   integrals  do   not   converge   in   such   cases.   Such   RAO   data   is   not   realistic,   since  no   vessel   can   respond  to  infinitely  high  frequency  waves.  

The  following  spectral  values  are  reported.   x

The  significant  amplitude.  

x

The  most  probable  maximum  amplitude  for  a  period  of  the  specified  Storm  Duration.  

x

The  average  period  of  the  response.  This  is  the  mean  period  between  zero  up-­‐crossings.   Note:  

The   significant   and   maximum   values   are   reported   as   single   amplitudes,   i.e.   the   motion   is   +/-­‐   the   value  reported.  

Warnings:   Wave   directional   spreading   spectra   are   not   taken   into   account.   The   calculation   assumes   that   all   the  spectral  energy  is  in  the  specified  direction  for  each  wavetrain.    

The  standard  formulae  used  to  calculate  the  spectral  values  (see  below)  are  based  on  the  further   assumptions   that   the   spectrum   is   narrow-­‐banded   and   Gaussian,   and   that   the   extremes   are   Rayleigh  distributed.  They  are  also  based  purely  on  small  amplitude  linear  theory.  

 

The  storm  duration  should  be  short-­‐term,  ie  of  the  order  of  hours  rather  than  days  or  years,  since   the  sea-­‐state,  as  represented  by  (Hs,Tz),  is  assumed  to  remain  constant  over  this  duration.  

These  spectral  values  are  calculated  as  follows:   The  spectrum  of  the  wave  train  is  combined  with  the  vessel  response  specified  by  its  RAOs,  to  obtain  the  response   spectrum   for   each   degree   of   freedom.   The   zeroth   and   second   spectral   moments,   m0   and   m2,   of   the   response   spectrum  are  then  calculated.  The  spectral  values  are  then  given  by:   Significant  Amplitude  =  As  =  2m0½   Average  Period  =  Tave  =  (m0/m2)½   Max  Amplitude  =  As(½ln(D*60*60/Tave))½   where   As   is   the   significant   amplitude,   D   is   the   specified   Duration   (in   hours)   and   Tave   is   the   average   period   (in   seconds).  See,  for  example,  Faltinsen,  pages  24-­‐27,  or  Ochi,  pages  151-­‐152.   Notes:  

For   Z   Above   Wave,   no   spectral   values   can   be   reported   for   acceleration   and   only   the   significant   amplitude   can   be  reported   for   velocity.   This   is   because   the   RAO   of   Z   Above   Wave   does   not   decay   to   zero   as   wave   period   goes   to   zero,   so   some   of   the   integrals   for   the   spectral   moments   do   not   converge.   For   velocity   of   Z   Above   Wave   you   could   perhaps   estimate   the   expected   maximum   amplitude   by   assuming   that   the   ratio   of   maximum   amplitude   to   significant   amplitude   is   roughly   the  same  for  velocity  as  for  position.  

266  

w

 

System  Modelling:  Data  and  Results,  Vessels  

 

 

All  these  reports  are  based  on  the  vessel  type   displacement  RAOs.  Any   wave  load  RAO  data  are   ignored.  

Impulse  Response,  Added  Mass  and  Damping   The  Impulse  Response,  Added  Mass  and  Damping   page   has  only  one   button:   Show   Graphs.  This  button  is  enabled   only   if   the   Vessel   Type   and   Draught   used   by   the   vessel   have   Added   Mass   and   Damping   Method   set   to   Frequency-­‐ Dependent.  Clicking  the  button  will  open  a  form  showing:   x

A  graph  of  the  Impulse  Response  Function  (IRF)  for  the  vessel.  

x

Graphs  of  the  original  frequency-­‐dependent  added  mass  or  damping  data.  

The  scale  of  the  time  lag  axis  for  the  IRF  is  determined  by  the  cutoff  time  chosen.  This  graph  may  be  useful  in  judging   what  value  should  be  selected  for  this  cutoff  time:  a  shorter  cutoff  time  will  allow  for  faster  calculation,  but  too  short   may  mean  significant  IRF  values  are  discarded  and  accuracy  lost  as  a  result.   The   graph   of   Added   Mass   against   frequency   has   superimposed   on   it   a   horizontal   intercept   representing   the   calculated   value   of   the  "infinite-­‐frequency"   added   mass.   So   long  as   the   data   are  given   to   sufficiently  high   frequency,   this  is  an  indicator  of  the  consistency  of  the  added  mass  and  damping  data.   Note  that  the  values  of  added  mass  and  damping  shown  may  have  been  Froude  scaled  according  to  Vessel  length.  

6.7.5

Vessel  Results  

For  details  on  how  to  select  results  variables  see   Selecting  Variables.   Motion  Results   The   vessel   motion   is   split   into   two   components,   the   primary   and   superimposed   motions.   Both   the   total   overall   motion  and  these  two  components  are  available  as  time  history  results.   X,  Y,  Z,  Rotation  1,  Rotation  2,  Rotation  3  and  Sea  Surface  Clearance  

The  position,  orientation  and  sea  surface  clearance  of  the  vessel,  relative  to  global  axes,  due  to  the  combination  of   the  primary  and  superimposed  motion.   X,   Y   and   Z   are   the   global   coordinates   of  a   user   specified   point   P   on   the   vessel.   The   point   P   is   specified   in   vessel   local   coordinates.  If  P=(0,0,0)  then  the  global  X,  Y  and  Z  coordinates  of  the  vessel  origin  are  reported.   Sea   Surface   Clearance   is   the   vertical   clearance   of   point   P   from   the   instantaneous   sea   surface.   A   positive   value   indicates  the  point  P  is  above  the  sea  surface.   Rotation   1,   2   and   3   define   the   final   orientation   relative   to   global   axes.   The   3   rotations   (called   Euler   angles)   are   3   successive  rotations  that  take  the  global  axes  directions  to  the  final  axes  directions.  Rotation  2  is  in  the  range  -­‐90°  to   +90°.  Range   jump   suppression  is  applied   to   the   Rotation   1   and   Rotation   3   angles   (so   values   outside   the   range  -­‐360°   to  +360°  might  be  reported).   Velocity,  GX-­‐Velocity,  GY-­‐Velocity,  GZ-­‐Velocity  

The  magnitude  and  components  of  the  velocity  of  the  vessel,  relative  to  global  axes,  due  to  the  combination  of  the   primary  and   superimposed  motion.   These   results   are   reported   at   a   user   specified   point   P   on   the   vessel.   The   point   P   is  specified  in  vessel  local  coordinates.   Angular  Velocity,  x-­‐Angular  Velocity,  y-­‐Angular  Velocity,  z-­‐Angular  Velocity  

The   magnitude   and   components   of   the   angular   velocity   of  the   vessel,   relative   to   vessel   axes,   due   to   the   combination   of  the  primary  and  superimposed  motion.   Acceleration,  GX-­‐Acceleration,  GY-­‐Acceleration,  GZ-­‐Acceleration  

The  magnitude  and  components  of  the  acceleration  of  the  vessel,  relative  to  global  axes,  due  to  the  combination  of   the   primary   and   superimposed   motion.   These   results   are   reported   at   a   user   specified   point   P   on   the   vessel.   The   point  P  is  specified  in  vessel  local  coordinates.   Acceleration  (incl.  g),  x-­‐Acceleration  (incl.  g),  y-­‐Acceleration  (incl.  g),  z-­‐Acceleration  (incl.  g)  

The  magnitude  and  components  (with  respect  to  vessel  axes)  of  the  vector  a   -­‐  g  where  a  is  the  acceleration  of  the   vessel  and   g  is  the  acceleration  due  to  gravity,  a  vector  pointing  vertically  downwards.  These  results  are  reported  at   a  user  specified  point  P  on  the  vessel.  The  point  P  is  specified  in  vessel   local  coordinates.  

267  

w

 

System  Modelling:  Data  and  Results,  Vessels    

These  results  can  be  used  to  compare  against  accelerometer  readings.   Angular  Acceleration,  x-­‐Angular  Acceleration,  y-­‐Angular  Acceleration,  z-­‐Angular  Acceleration  

The   magnitude   and   components   of   the   angular   acceleration   of   the   vessel,   relative   to   vessel   axes,   due   to   the   combination  of  the  primary  and  superimposed  motion.   Primary  X,  Primary  Y,  Primary  Z,  Primary  Rotation  1,  Primary  Rotation  2  and  Primary  Rotation  3  

The   primary   position   of   the   vessel,   as   produced   by   any   primary   motion,   relative   to   global   axes.   So   Primary   X,   Primary  Y  and  Primary  Z  are  the  global  X,Y,Z  coordinates  of  the  primary  position  of  the  vessel  origin,  and  Primary   Rotation  1,  Primary  Rotation  2  and  Primary  Rotation  3  are  the  primary  orientation  of  the  vessel,  again  relative  to   global  axes.   Primary   Rotation   2   is   in   the   range   -­‐90°   to   +90°.   Range   jump   suppression   is   applied   to   Primary   Rotation   1   and   Primary  Rotation  3  (so  values  outside  the  range  -­‐360°  to  +360°  might  be  reported).   Surge,  Sway,  Heave,  Roll,  Pitch  and  Yaw  

The   offset   of   the   vessel,   due   to   any   superimposed   motion,   relative   to   the   primary   position   of   the   vessel.   They   are   usually,   therefore,   the   wave-­‐generated   part   of   the   motion,   so   Surge,   Sway   and   Heave   are   the   offsets   from   the   primary  position  to  the   final  position  and  are   measured  in  the   primary  vessel  axes   directions.  And  Roll,  Pitch  and   Yaw  are  the  wave-­‐generated  rotations  and  are  relative  to  the  primary  vessel  axes  directions.   Pitch  is  in  the  range  -­‐90°  to  +90°.  Range  jump  suppression  is  applied  to  the  Roll  and  Yaw  angles  (so  values  outside   the  range  -­‐360°  to  +360°  might  be  reported).   Sea  Surface  Z  

The  global  Z  coordinate  of  the  sea  surface  directly  above  the  instantaneous  vessel   primary  position.   Force  and  Moment  Results   For   each   load   on   the   vessel,   so   long   as   it  is   not  explicitly   excluded   from   the   list   of  that   vessel's  Included   Effects,   time   history   results   are   available   for   the   magnitude   of   the   force   and   moment,   and   for   each   component   of   the   force   or   moment.   All   these   results   are   reported   in   vessel   axes   directions,   with   moments   taken   about   the   vessel   origin.   If   a   load  is  excluded  then  it  will  not  be  calculated  and  will  not  appear  in  the  list  of  available  results.   Total  Load  

The  magnitude  and  components  (in  vessel  axes  directions)  of  the  sum  of  the  constituent  loads  on  the  vessel  which   are  included  in  the  calculation.   Connections  Load  

The  sum  of  the  loads  from  all  attached  lines,  links,  winches,  shapes,  etc.  Available  only  if  at  least  one  such  object  is   attached   to   the   vessel.   For   convenience,   the   components   of   Connection   Force   and   Moment   are   also   available   in   Global   axes   directions   (GX,GY,GZ).   Details   of   the   loads   exerted   by   each   individual   object   are   available   as   results   tables,  and  can  also  be  found  under  the  results  for  each  object  itself.   Applied  Load  

The  sum  of  all  the  local  and  global   applied  loads.  Available  only  if  Applied  Loads  is  in  the  list  of  the  vessel's  included   effects.   Hydrodynamic  Damping  Load  

The   hydrodynamic   damping   force   and   moment   on   the   vessel.   Available   only   if  Hydrodynamic  Damping   is   checked   in  included  effects.   Wind  Damping  Load  

The  wind  damping  force   and   moment  on   the   vessel.   Available   only  if  Wind  Damping  is   checked   in   included  effects   and  Include  Wind  Loads  on  Vessels  is  checked  on  the  Wind  data  page  of  the  Environment  data  form.   Wave  Drift  Load  

The  wave  drift  force  and  moment  exerted  on  the  vessel.  These  results  are  available  only  if   Wave  Drift  Load  (2nd   order)  is  included.  

268  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Added  Mass  &  Damping  Load  

The  force  and  moment  on  the  vessel  due  to  added  mass  and  damping.   Hydrostatic  Stiffness  Load  

The  Hydrostatic  stiffness  force  and  moment  on  the  vessel.   Wave  Load  RAO  Force  and  Moment  

The  force  and  moment  on  the  vessel  due  to  the  wave  load  RAOs.  Available  if  Wave  Load  (1st  Order)  is  checked  in   included  effects.   Note:  

First  order  wave  RAO  load  and  second  order  wave  drift  load  results,  though  reported  with  respect   to  vessel  axes,  are  applied  with  respect  to   primary  heading  axes.  In  most  cases  there  will  be  little   difference   between   the   two   frames   of   reference,   but   large   values   of   Primary   Rotations   1&2   may   have  a  significant  effect  on  these  results.  

Multiple  Static  Results   For  multiple  statics  calculations  the   results  variables  available  are  as  follows.   The  loads  reported  are  the  total  loads,   including  those  from  current,  wind,  applied  loads  and  attached  lines  and  other  objects.   Restoring  Force  

The   magnitude   of   the   horizontal   component   of   the   total   force   applied   to   the   vessel.   Note   that   this   force   is   not   necessarily  in  the  offset  direction.   Vertical  Force  

The  vertically  downwards  component  of  the  total  force  applied  to  the  vessel.   GZ-­‐Moment  

The  total  moment,  about  the  vertical,  applied  to  the  vessel.   Worst  Tension  

The  largest  tension  in  any  segment  of  any  Line  connected  to  the  Vessel.  

6.8

LINES  

Lines  are  flexible  linear  elements  used  to  model  cables,  hoses,  chains  or  other  similar  items.  Lines  are  represented  in   OrcaFlex  using  a  lumped  mass  model.  That  is,  the  line  is  modelled  as  a  series  of  lumps  of  mass  joined  together  by   massless   springs,   rather   like   beads  on   a   necklace.   The   lumps   of   mass   are   called  nodes  and   the   springs   joining   them   are   called   segments.   Each   segment   represents   a   short   piece   of   the   line,   whose   properties   (mass,   buoyancy,   drag   etc.)   have   been   lumped,   for   modelling   purposes,   at   the   nodes   at   its   ends.   See   the   figure   below,   which   shows   an   example  line  spanning  from  a  Vessel  to  a  Buoy.  

269  

w

 

System  Modelling:  Data  and  Results,  Lines    

End positions and no-moment directions are defined relative to the objects to which the ends are connected and move with those objects.

z

y

End A z V

B

x

y x End B

section 1 (3 segments)

section 3 (9 segments)

Clump

section 2 (7 segments)

  Figure:  

Line  Model  

The  properties  of  a  Line  are   specified   by  dividing   it  up  into  a  number  of  consecutive  sections  that  are  chosen   by   the   user.  For   each  section  you  must  define  its  length,  the  Line  Type  of  which  it  is  made  and  the  number  of  segments  into   which   it   should   be   divided   for   modelling   purposes.   A   Line   Type   is   simply   a   set   of   properties   (for   example   the   diameter,  mass  per  unit  length  and  bend  stiffness)  given  a  name  so  that  they  can  be  called  by  that  name.  The  Line   Types  are  defined  separately,  on  the  Line  Types    data  form.  This  allows  the  same  set  of  line  properties  to  be  used  for   a   number   of   different   sections   of   the   line,   or   for   different   lines.   There   is  also   a  Line   Type  Wizard  tool   that   helps  you   set  up  Line  Types  representing  common  structures  like  chains,  ropes,  etc.   In   addition,   a   number   of   attachments   may   be   specified,   to   represent   items   that   are   connected   to   the   Line.   For   example,   attachments   may   be   used   to   model   clump   weights,   drag   chains   or   buoyancy   bags   attached   to   the   line.   Two   types  of  attachment  are  available  Ȃ  clumps  (buoyancy  or  heavy)  and  drag  chains.  

270  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Each  attachment  attached  to  the  Line  is  specified  by  giving  the   Attachment  Type  and  the  arc  length,  measured  from   End   A,   at   which   it   should   be   attached.   The   attachment   is   then   attached   to   the   nearest   node   to   that   arc   length.   Attachment  Types  are  similar  to  Line  Types  Ȃ  they  are  simply  named  sets  of  attachment  properties.  The  properties   themselves  are  then  given  separately,  on  the  Attachment  Types  data  form.  This  allows  the  same  set  of  attachment   properties  to  be  used  for  a  number  of  different  attachments.   The   two   ends   of   a   Line   are   referred   to   as   End   A   and   End   B   and   each   end   can   be   Free,   Fixed,   Anchored   or   else   connected  to  a  Vessel  or  Buoy.  The  two  ends  of  a  line  are  treated  in  essentially  the  same  way,  but  some  aspects  of   the  line  are  dependent  on  which  end  is   which.  In   particular  the   numbering  of  parts  of  a  Line  is  always  done  starting   at  End  A.  

6.8.1

Line  Data  

For   every   line   in   the   system   there   is   a   data   form   defining   its   structure   and   interconnection.   It   is   on   these   data   forms   that  the  system  is  built  up  by  connecting  lines  between  the  objects  that  have  been  defined.   Name  

Used  to  refer  to  the  Line.   Include  Torsion  

Torsional   effects   can   be   included   or   ignored,   for   each   line   in   the   model.   If   torsion   is   included   then   the   line   type   torsional  properties  must  be  specified.  See  Torsional  Stiffness.   To   see   the   line   orientation   visually   on   the  3D   views,   select   Draw  Node   Axes   on   the   View   menu.   OrcaFlex   then   draws   the  node  axes  Nxyz  at  each  node,  and  these  axes  allow  you  to  see  how  the  line  is  behaving  torsionally.   Notes:  

The  node  axes  are  drawn  using  the  node  pen,  specified  on   the  line  data  form.  

 

If  torsion  is  included  for  a  line,  you  must  specify  the  torsional  orientation  at  each  end  of  the  line.   This   is   done   by   setting   the   Gamma   angle   of   the   end   connections   on   the   line   data   form.   The   Gamma   angle   determines   the   torsional   position   of   the   line   end   Ȃ   for   details   see   Line   End   Orientation.   To   check  visually  that  you  have  the  orientation  you  expect,  select  Draw  Node  Axes  on  the  View  menu.  

 

If   torsion   is   included   for   a   line,   the   static   analysis   should   also   include   the   effects   of   torsion   Ȃ   otherwise   the   simulation   will   start   from   a   position   that   is   not   in   torsional   equilibrium   and   an   unstable   simulation   may  result.   We   recommend   that   the  Full   Statics   option   is   selected   because   this   is  the  only  statics  option  in  OrcaFlex  that  includes  the  effects  of  torsion.  

Top  End  

This  data  item  is  used  to  give  OrcaFlex  information  about  the   sense  of  the  Line.  Various  calculations  performed  by   the   program   need   to  know   which   end   of  the   line   (End   A  or   End   B)   is   at   the   top,   and   which   end   is   at   the   bottom.   You   specify  which  end  is  at  the  top,  and  the  program  assumes  that  the  other  end  is  at  the  bottom.   Suppose  you  have  a  line  with  the  top  end  connected  to  a  vessel,  and  the  bottom  end  anchored  to  the  seabed.  If  you   wish  to   measure   arc   length  from   the   vessel   then   you   should   connect   End   A   to   the   vessel,   make   End   B   anchored   and   set  Top  End  to  End  A.  On  the  other  hand,  if  you  wish  to  measure  arc  length  from  the  seabed  then  you  should  connect   End  B  to  the  vessel,  make  End  A  anchored  and  set  Top  End  to  End  B.   The  setting  of  the  Top  and  Bottom  Ends  is  used  by  the  program  as  follows:   x

The  Lay  Azimuth  data  defines  a  lay  direction  starting  from  the  Bottom  End  and  moving  towards  the  Top  End.  

x

The   Touchdown   results   point   is   determined   by   starting   at   the   Top   End   and   then   moving   towards   the   Bottom   End  until  the  first  node  in  contact  with  the  seabed  is  found.  

x

The   Contents   Pressure   Reference   Z   level   can   be   set   to   '~'   (indeed   this   is   the   default   value)   which   OrcaFlex   interprets  as  the  Z  level  of  the  Top  End  in  the  reset  state.  

x

The   Line   Setup   Wizard   uses   the   bottom   end   when   calculating   anchor   positions   and   also   for   the   layback   calculation.  

If   the   Line   is   not   in   contact   with   the   seabed   then   this   data   is   somewhat   arbitrary.   You   are   free   to   make   whatever   choice  suits  your  model,  but  remember  that  the  contents  pressure  will  be   referenced   from  the  Top  End.  If   the   entire   Line  is  in  contact  with  the  seabed  then  again  you  are  free  to  make  whatever  choice  of  Top  and  Bottom  Ends  suits   your  model.  

271  

w

 

System  Modelling:  Data  and  Results,  Lines    

Connections   The   line   end   connection   data   specifies   whether   the   line  ends  are   connected   to   other   objects,   the   position,   angle  and   stiffness  of  the  connection,  and  whether  the  end  is  released  during  the  simulation.   You   can   view   and   edit   an   individual   line's   connection   data   on   the   line's   data   form.   Or   you   can   view   and   edit   the   connection  data  for  all  the  lines  together  on  the  All  Objects  Data  Form.   Connect  to  Object  

The  line  spans  from  End  A  to  End  B  and  each  end  may  be  connected  to  another  object  in  the  model,  such  as  a  buoy  or   vessel,  or  else  Fixed,  Anchored  or  left  Free.   Object  Relative  Position  

Defines  the  position  of  the  centre  of  the  node  at  the  line  end.   x

If   the   end   is   connected   to   another   object   this   defines   the   coordinates   of   the   connection   point   relative   to   that   other  object's  local  axes.  

x

If  the  end  is  Fixed  this  defines  the  coordinates  of  that  point  relative  to  global  axes.  

x

If   the   end   is   Anchored   this   defines   the   X   and   Y   coordinates   of   the   anchor   relative   to   global   axes,   plus   the   Z-­‐ coordinate  relative  to  the  seabed  level  at  that  (X,Y)  position.  

x

If  the  end  is  Free  then  this  defines  the  coordinates  of  the  estimated  equilibrium  position  of  the  line   end,  relative   to  global  axes.  

Height  above  seabed  

This  data  item  is  only  available  for  Anchored  connections  and  specifies  the  vertical  height  above  the  seabed  of  the   pipe  underside.  This  value  is  coupled  to  the  Object  Relative  z  coordinate  Ȃ  changing  either  one  results  in  the  other   being  changed  to  match.   To  understand  how  this  data  item  should  be  used  consider,  for  simplicity,  a  line   end  anchored  to  a  flat  horizontal   seabed.  The  Object  Relative  z  coordinate  specifies  the  position  of  the  centreline.  If  it  is  set  to  0  then  the  end  node  will   penetrate  the  seabed  by  a  distance  of  ½D,  where  D  is  the  contact  diameter.   The   net   result   of   this   is   that   the   end   node   is   'buried'   in   the   seabed   and   receives   a   large   seabed   reaction   force.   Because  it  is  anchored  this  force  cannot  displace  the  end  node,  but  the  adjacent  node  is  free  to  move  and  it  will  try  to   take  up  a  position  sitting  on   top  of  the  seabed.   This  in  turn  will  lead  to  unrealistic  values  of  curvature,  bend  moment   etc.  at  the  end.   If,  however,  you  set  Height  above  seabed  to  0  then  the  end  node  centreline  will  have  a  z  coordinate  of  ½D,  relative  to   the  seabed.  The  node  sits  just  in  contact  with  the  seabed  and  the  above  problems  are  removed.   If  the  seabed  is  not  horizontal  then  the  mathematics  is  slightly  more  complicated  as  it  has  to  take  into  account  the   slope  of  the  seabed.  However,  the  recommendation  of  setting  Height  above  seabed  to  0  remains  valid.   End  Orientation  

When  a  line  is  connected  to  an  object,  it  is  connected  into  an  end  fitting  that  is  rigidly  attached  to  that  object  and  you   specify  the  orientation  of  this  connection  by  giving  its  Azimuth,  Declination  and  Gamma  angles.   These  angles  define  the  end  fitting  orientation  relative  to  the  object,  so  for  objects  that  rotate  (e.g.  vessels  and  6D   buoys)   the   fitting   rotates   with   the   object.   For   Fixed   or   Anchored   ends   the   end   orientation   is   defined   relative   to   global  axes.  For  Free  ends  the  end  orientation  is  not  used.   Azimuth,  Declination  and  Gamma  define  the  end  fitting  orientation  by  specifying  the  directions  of  the  axes  (Ex,  Ey,   Ez)  of  its  frame   of  reference,  where  E  is  the  end  fitting  origin  Ȃ  the   point  to  which  the   line  end  is  connected.  See  Line   End  Orientation.   The  direction  of  Ez  is  defined  by  specifying  its  Azimuth  and  Declination  angles.  Ez  is  the  end  fitting  axial  direction;   when   the   end   segment   is   aligned   with   Ez   then   no   bending   moment   is   applied   by   the   joint,   so   Ez   is   sometimes   called   the  no-­‐moment  direction.  Note  that  Ez  must  be  specified  using  the   End  A  to  End  B  convention,  i.e.  Ez  is  into  the  line   at  End  A,  but  out  of  the  line  at  End  B.   Ex  and  Ey  are  perpendicular  to  Ez  and  they  are  defined  by  specifying  the  Gamma  angle,  w hich  is  a  rotation  about   Ez.   The  Ex  and  Ey  directions  are  used  for  reporting   results  (e.g.  the  2  components  of  shear  force).  And  if  the  line   has  

272  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

torsion  included  and  the  joint  twisting  stiffness  is  non-­‐zero,  then  Ex  and  Ey  also  define  the  line  end  orientation  at   which  no  torsional  moment  is  applied  by  the  joint.   The   connection   at   a   line   end   is   modelled   as   a   ball-­‐joint   with   this   orientation   being   the   preferred   "no-­‐moment"   orientation,   i.e.   the   orientation   of   the   line   end   that   gives   rise   to   no   moment   from   any   rotational   stiffness   of   the   connection.   If  all  of  the  end  connection  stiffness  values  are  zero,  e.g.  to  model  a  ball  joint  that  is  completely  free  to  rotate,  then   the  end  orientation  angles  have  no  effect  on  the  line  behaviour.  The  angles  then  only  serve   to  define  the  local  x,  y   and  z-­‐directions  that  are  used  to  define  results  (e.g.  shear  and  bend  moment  components,  stress  components,  etc.)   that  depend  on  the  local  axes  directions.   Bending  and  Twisting  Stiffness  

The   connection   at   a   line   end   is   modelled   as   a   joint   with   the   specified   rotational   stiffness.   The   restoring   moments   applied   by   the   joint   depend   on   the   deflection   angle,   which   is   the  difference  between   the   end   fitting   orientation   and   the   orientation   of   the   line.   The   end   orientation   is   therefore   the   orientation   of   the   line   that   corresponds   to   zero   moment  being  applied  by  the  joint.   The  connection  stiffness  is  the  slope  of  the  curve  of  restoring  moment  against  deflection  angle.   The  bending  and  twisting  connection  stiffnesses  can  be  set  to:     x

Zero:  free  to  rotate  with  no  resistance.  

x

Non-­‐zero,  finite:  can  rotate  but  with  resistance.  

x

Infinity:  a  rigid  connection.  

x

Variable:  non-­‐linear  (for  bending  connection  stiffness  only).  

The  x  bending  and  y  bending  values  specify  the  connection  bending  behaviour  for  rotation  about  the  end  Ex  and  Ey   directions,  respectively.  For  an  isotropic  ball  joint  the  two  values  must  be  equal;  this  can  conveniently  be  specified   by  setting  the  y-­‐bending  value  to  '~',  meaning  'same  as  x-­‐value'.  A  non-­‐isotropic  ball  joint  can   be   modelled   by  giving   different  x  and  y  bending  values;  in  this  case  the  line  must  include  torsion.   The  x  bending  and  y  bending  behaviour  can  either  be  linear  or  non-­‐linear,  as  follows:   x

For   a   simple   linear   behaviour,   specify   the   bending   stiffness   to   be   the   constant   slope   of   the   curve   of   restoring   moment  against  deflection  angle.  

x

For   a   non-­‐linear   behaviour,   use   variable   data   to   specify   a   table   of   restoring   moment   against   deflection   angle.   OrcaFlex  uses  linear  interpolation  for  angles  between  those  specified  in  the  table,  and  linear  extrapolation  for   angles  beyond  those  specified  in  the  table.  The  restoring  bend  moment  must  be  zero  at  zero  angle.  

The  Twisting  Stiffness  value  is  only  relevant  if   torsion  is  included  for  the  line.  It   specifies  the  rotational  stiffness   about   the   end   Ez   direction.   For   the   twisting   stiffness   this   variation   is   always   modelled   as   linear   so   the   twisting   stiffness  you  specify  should  be  the  slope  of  the  linear  angle-­‐moment  curve.   A  flex  joint  can  be  modelled  by  setting  the  stiffness  values  to  be  non-­‐zero  and  finite.   Warning:  

Avoid   specifying   large   connection   stiffness   values   (except   the   special   value   Infinity)   since   they   require  very  short  simulation  time  steps.  

Release  at  Start  of  Stage  

If  desired  each  line  end  can  be  disconnected  at  the  start  of  a  given  stage  of  the  simulation.  If  no  release  is  wanted   then  set  this  item  to  "~",  meaning  "not  applicable".  

Structure   Each   line   can   be   made   up   of   up   a   number   of   sections   with   different   properties,   the   sections   being   defined   in   sequence  from  End  A  to  End  B.   Line  Type  

This  determines  the  properties  of  the  section.  

273  

w

 

System  Modelling:  Data  and  Results,  Lines    

Section  Length  

The   unstretched   length   of   the   section.   This   is  the   unstressed   length  (i.e.   zero   wall   tension)   at   atmospheric  pressure   inside  and  out.  Length  changes  due  to  external  and  internal  pressure,  and  allowing  for  the  Poisson  ratio  effect,  are   calculated  and  allowed  for  by  OrcaFlex.   If  the  line  type  is  profiled  then  the  section  length  is  determined  by  the  profile  data  and  so  cannot  be  edited  here.   Expansion  Factor  

The  expansion  factor  allows  you  to  model  time-­‐varying  changes  in  unstretched  length,  for  example  due  to  thermal   expansion  or  contraction.   A  value  of  '~'  means   that  no   expansion  factor  is  applied  Ȃ  this  is   equivalent  to  a  value   of  1.  Other  positive  values  can   be  used,  in  which  case  the  unstretched  length  remains  constant  throughout  the  simulation.   Alternatively  the  expansion  factor  can  be  a  variable  data  source  which  specifies  a  table  of  expansion  factor  against   simulation   time.   It   specifies   a   multiplicative   factor   which   is   applied   to   the   unstretched   length  when   calculating   axial   strain  which  in  turn  is  used  to  calculate  effective  tension  (see  Line  Theory:  Calculation  Stages).   Note:  

Expansion  factor  is  only  used  in  the  calculation  of  strain.  It  has  no  effect  on  mass,  buoyancy,  drag,   added  mass  etc.  

Target  Segment  Length,  Number  of  Segments  

These  data  items  determine  the  segmentation  of  the  section.   If  Target  Segment  Length  is  set  to  ~  then  the  number  of  segments  in  the  section  is  set  by  Number  of  Segments.   Otherwise,  the  segmentation  is  chosen   based   on   Target  Segment  Length.  The  Number  of  Segments  is   not  editable   and  reports  the  actual  number  of  segments  used  which  is  given  by  the  formula:   Number  of  Segments  =  Round(Section  Length  /  Target  Segment  Length)   where  Round  is  the  function  that  rounds  a  floating  point  value  to  the  nearest  integer.   Note:  

It   is   usually   preferable   to   determine   segmentation   by   specifying   Target   Segment   Length.   This   allows  you  to  alter  section  lengths  without  altering  segment  length.  

Clash  Check  

Clash  modelling  is  included  when  this  data  item  is  set  to   Yes.  If  it  is  set  to  No  then  the  section  will  be  ignored  for   clashing  purposes.   Notes:  

Line  clashing  is  not  modelled  during  statics.  

 

Clash  checking  is  quite  time-­‐consuming,  so  you  should  only  set  this  item  to   Yes  for  those  sections   for  which  you  need  clash  modelling  to  be  included.  See  Line  Clashing.  

Cumulative  Length,  Cumulative  Segments  

These  columns  report  the  cumulative  length  and  cumulative  number  of  segments  counted  from  the  first  section.  The   values  are  for  reporting  purposes  only  and  cannot  be  edited.   Profile  Graph  

The  profile  graph  plots  the  inner  and  outer  radii  of  the  line  as  they  vary  with  arc  length.  This  is  especially  useful  to   check  that  stress  joint  and  bend  stiffener  data  has  been  correctly  input.  

Pre-­‐bend   Pre-­‐bend  is  only  available  when   torsion  is  modelled.   Pre-­‐bend  is  provided  for  modelling  lines  which  are  not  straight   when  unstressed,  e.g.  spool  pieces.   The   pre-­‐bend   is   defined   for   each   section   by   specifying   the   pre-­‐bent   curvature   (in   radians   per   unit   length)   of   the   section.  The  pre-­‐bent  curvature  is  the  curvature  of  the  pipe  in  its  unstressed  state.  For  lines  which  are  straight  when   unstressed  then  pre-­‐bend  should  be  specified  to  be  zero  Ȃ  which  is  the  default  setting.   Pre-­‐bend   can   be   specified   in   both   the   line   local   x   and   y   directions.   However,   to   simplify   data   preparation   and   interpretation  of  results  we  recommend  that  you  arrange  the  line's  local  axes  such  that  the  pre-­‐bend  is  entirely  in   either  the  local  x  or  local  y  direction.  

274  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Note:  

When  pre-­‐bend   is  modelled   curvature  and   bend  radius  results  are  reported  relative  to  the  pre-­‐bent   curvature.  

Warning:  

Pre-­‐bend  breaks  the  assumptions  of  the  stress  results  and  fatigue  analysis.  

Attachments   A   number   of   attachments   may  be   added   to  each  line.   Each  attachment   can  either   be  of   a   specified   Attachment  Type   or  else  be  a  clone  of  a  specified  6D  buoy.   Attachment  Type  

Can  be  a  Clump  Type,  a  Drag  Chain  Type,  a  Flex  Joint  Type,  a  Stiffener  Type  or  an  existing  6D  Buoy.   If   you   specify   a   6D   buoy   as   the   attachment   type   then   the   attachment   is   a   clone   of   that   6D   buoy   and   changing   the   properties  of  the  6D   buoy  also  changes  the   properties  of  the  attachment.  The   6D   buoy  from  which  the  attachment  is   cloned  cannot  be  deleted,  without  first  deleting  all  the  attachments  that  are  clones  of  it.   6D   buoy   attachments   are   useful   when   you   want   a   number   of   identical   6D   buoys   attached   to   a   line.   To   attach   20   identical  buoys  to  a  line,  for  example,  first  create  the  first  buoy  separately  from  the  line  and  then  connect  it  to  the   line  by  setting   its  connection  data   item   on   the   buoy   data   form.   This   first   buoy   acts  as   the   master   from  which  all   the   other   attachment   buoys   are   cloned.   Then,   on   the   line   data   form,   specify   19   attachments   and   set   their   attachment   type  to  be  the  first  6D  buoy.   Note:  

6D  Buoy  attachments  can  only  be  used  when  the  Line  includes  torsion.  

Position  

The  x,  y  and  z  coordinates  specify  the  position  of  the  attachment  relative  to  the  line.   The  z  coordinate  specifies  the  arc  length  at   which   the  attachment  is  connected  and  this  arc  length   may   be  measured   relative  to  either  End  A  or  End  B  as  specified  by  the  user.   x

For   Clumps,   Drag   Chains,   Flex   Joints   and   Stiffeners   the   x   and   y   coordinates   must   be   zero   and   the   z   coordinate   is   the  arc  length.  These  attachments  are  connected  at  the  node  nearest  to  this  arc  length.  

x

For  Stiffeners  the  z  coordinate  specifies  the  arc  length  of  the  stiffener  connection  point.   Note:  

x

If   the   attachment   is   a   clump   then   it   is   also   offset   vertically   from   the   node   by   the   offset   distance   specified  in  the  clump  type  data.  Beware  that  the  sign  convention  for  this  offset  varies  depending   on   whether   the   clump   is   net   buoyant   (positive   offset   is   upwards)   or   heavy   (positive   offset   is   downwards).  

For  6D  Buoy  attachments  the  z  coordinate  specifies  the   arc  length  at  which  the  buoy  should  be  connected  to  the   line.  The  buoy  will  be  connected  to  the  nearest  node  to  that  arc  length.  The  buoy  will  be  connected  with  an  offset   (relative  to  that  node's  axes)  that  is  given  by  (x,  y,  0).  See   6D  Buoy  Initial  Position  for  more  details.  

Orientation  

For  6D  Buoy  attachments  only.  Rotation  1,  Rotation  2  and  Rotation  3  determine  the   Initial  Attitude  of  the  attached   buoy.   Name  

For  6D  Buoy  and  Drag  Chain  attachments  only.  This  is  the  name  of  the  attached  object  and  is  used  to  select  results   for  that  object.  

Contents   Contents  Method  

Three  methods  of  modelling  contents  are  available,  as  outlined  below:   x

If  uniform  is  selected  then  the  entire  line  is  assumed  to  be  filled  with  contents  of  a  uniform  density.  

x

The  free-­‐flooding  option  results  in  the  line  being  filled  with  sea  water,  up  to  the  instantaneous  water  surface.  

x

The   slug  flow   method   allows   for   spatial   and   temporal   variation   of  contents.   The   contents   flow   velocity   can   also   vary  with  time.  

275  

w

 

System  Modelling:  Data  and  Results,  Lines    

Include  axial  contents  inertia  

This  data  item  allows  the  component  of  inertia  due  to  contents  in  the  line's  axial  direction  to  be  omitted  from  the   analysis.   Typically   the   axial   contents   inertia   would   be   excluded   from   an   analysis   of   a   line   with   free-­‐flooding   contents,  for  example  a  drilling  riser  in  emergency  disconnect  mode.   Contents  Pressure  and  Reference  Z  level  

The  contents  pressure  specifies  the  internal  pressure  in  the  line  at  a  specified  fixed   reference  Z  level  (specified   relative   to   global   axes).   The   internal   pressure   at   this   Z   level   is   assumed   to   remain   constant   throughout   the   simulation.  The  internal  pressure  at  other  levels  is  calculated  allowing  for  the  static  pressure  head  due  to  differences   in  Z  level.  For  slug  flow,  the  static  pressure  head  is  calculated  using  the  mean  contents  density  in  the  line.   The  reference  Z  level  can  be  set  to  '~'  and  this  is  taken  to  mean  the  Z  level  of  the   Top  End  of  the  line  in  the   reset   state.   See  Line  Pressure  Effects  for  details  of  contents  pressure  modelling.   Uniform  contents  data   Contents  Density  

Each   section   of   the   line   is   assumed   to   be   full   of   contents   of   this   density   and   the   mass   of   the   section   is   increased   accordingly.   All   pressures   in   OrcaFlex   are   gauge   pressures,   not   absolute   pressures.   That   is,   they   are   relative   to   atmospheric   pressure  and  so  can  be  as  low  as  -­‐1  atmosphere  (-­‐101.3  kPa).   Flow  Rate  

The   rate   of   flow   of   mass   through   the   line.   If   it   is   non-­‐zero   then   it   is   used   to   calculate   the   centrifugal   and   Coriolis   forces   due   to   flow   of   fluid   in   the   line.   Positive   values   mean   flow   from   End   A   towards   End   B   and   vice   versa.   To   convert  between  mass  flow  rate,  volume  flow  rate  and  flow  velocity  use  the  following  simple  formulae:   ‘Ž—‡ˆŽ‘™”ƒ–‡αƒ••ˆŽ‘™”ƒ–‡Ȁɏ   Flow  velocity  =  Volume  flow  rate  /  A   ™Š‡”‡ɏ‹•–Š‡…‘–‡–•†‡•‹–›ƒ†‹•–Š‡‹–‡”ƒŽ…”‘•s-­‐sectional  area.   Slug  flow  data   The  slug   flow  data  allows  you  to  specify  variation  of  contents   density  along  the  arc  length  of  the  line.   This  pattern  of   contents   can   also   progress   along   the   line   over   time.   The   program   accounts   for   the   resulting   variations   of   mass,   weight  and  centrifugal  and  Coriolis  forces   Flow  velocity  

The  velocity  at  which  the  contents  pattern  flows  along  the  line.   This  value  can  be  constant,  or  can  vary  as  simulation   time  varies.  Positive  values  mean  flow  from  End  A  towards  End  B  and  vice  versa.   A   value   of  zero   can   be   used   to  specify   spatial   contents   variation   with   no   temporal   variation.   A   variable   flow   velocity   can  be  used  to  model,  for  example,  the  flow  of  contents  out  of  a  drilling  riser  in  emergency  disconnect  mode.   Note:  

There  is   no   need   to   ramp   flow   velocity   at   the   beginning   of   dynamics  Ȃ  in   fact   it   is   better   to   include   the   fluid   flow   in   the   static   calculation   because   this   removes   undesirable   transients   during   the   dynamic   analysis.   Therefore,   if   you   wish   to   model   a   constant   flow   rate,   you   should   set   the   flow   velocity  to  be  this  constant  value.  

Density  between  slugs  

The  contents  density  for  sections  of  the  line  that  fall  in  between  slugs.   Slug  flow  contents  pattern  

The  spatial  variation  of  contents  density,  i.e.  the  contents  pattern,  is  specified  in  a  table.  Each  row  of  the  table  allows   you  to  define  a  number  or  group  of  identical  slugs  specified  by  their   density  and  length  together  with  the  distance   between  slugs.   In   addition   you   specify   a   reference   point,   an   arc   length   which   can   be   relative   to   either   end   of   the   line,   and   the   simulation  time  at  which  the  first  slug  in  the  group  reaches  that  reference  point.  If  the  flow  velocity  is  zero  then  the   previous  definition  is  meaningless.  In  this  case  we  adopt  the  convention  that  the  group  of  slugs  covers  arc  lengths  

276  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

(measured   from   End   A)   greater   than   the   reference   point.   For   example,   if   the   flow   velocity   is   zero   and   you   have   a   single  slug  with  length  L  and  reference  point  at  End  A  then  the  slug  will  stretch  between  arc  lengths  0  and  L.   Simple  repeating  patterns  of  slugs  can  easily  be  modelled  using  a  single  row  in  the  table.  For  irregular  patterns  of   slug  you  can  model  each  slug  as  a  single  row  in  the  table.  Range  graphs  of   contents  density  can  be  used  to  confirm   that  your  data  corresponds  to  your  desired  pattern  of  slugs.  

Statics   The  line  static  calculation  is  performed  in  2  steps  as  described  in   Statics  of  Lines.   Included  in  Statics   This   switch   allows   you   to   exclude   certain   lines   from   the   statics   calculation.   This   is   mainly   useful   when   building   a   model  and  a  particular  line  is  not  converging.  In  this  situation  you  could  exclude  all  other  lines  from  statics  (this  is   easiest   from   the   All   Objects   Data   Form).   This   would   allow   you   to   experiment   with   different   statics   convergence   parameters  for  the  problematic  line  without  having  to  wait  for  all  the  other  lines  to  converge  each  time  you  tried  a   new  set  of  convergence  parameters.   Note:  

Results   are   not   available   for   such   lines   and   dynamics   is   disabled   if   you   have   any   lines   which   are   excluded  from  statics.  Lines  which  are  excluded  from  statics  have  no  influence  on  other  objects  in   the  model.  

Step  1  Statics  Method   This  can  be  either  Catenary,  Spline,  Quick,  Prescribed  or  User  Specified.   The   normal   setting   is  Catenary,   in   which  case   the   static   analysis  finds  the   equilibrium   catenary  position   of   the   line,   allowing   for   weight,   buoyancy,   drag,   but   not   allowing   for   bend   stiffness   or   interaction   with   shapes.   See   Catenary   Statics.   The  Catenary  solution  has  some  limitations  and  some  systems,  such  as  those  with  slack  or  neutrally  buoyant  lines,   can  be  troublesome.  For  such  lines  you  can  instead  specify  Spline,  in  which  case  the  line  is  instead  set  to  a  3D  spline   curve  based  on  spline  control  points  specified  by  the  user.  See   Spline  Data  and  Spline  Statics.   The  Quick  method  leaves  the  line  in  the  rough  catenary  shape  used  in  the  Reset  state.  See   Quick  Statics.   For  pull-­‐in  analysis  the  Prescribed  option  has  been  provided.  Here  the  user  specifies  the  starting  position  of  the  line   as  a  sequence  of  straight  line  or  curved  sections  on  the  seabed.  See   Prescribed  Starting  Shape.   The   User   Specified   option   allows   you   to  specify   the   position   for   each   node   on   the   line.   No   calculation   is   performed,   the  nodes  are  merely  placed  at  the  specified  positions.  See  User  Specified  Starting  Shape  and  User  Specified  Statics.   Step  2  Statics  Method  (Full  Statics)   This  can  be  either  None  or  Full  Statics.   If  None  is  selected  then  the  position  obtained  by  the  Step  1  Statics  Method  is  used.   The   Full   Statics   calculation   finds  a   full   equilibrium   position   for   the   model.   Unlike   the   Step   1   Catenary  method,   bend   stiffness  and  interaction  with  shapes  are  included.  Full  statics  needs  a  starting  shape  for  the  line,  and  it  uses  the  Step   1  Statics  Method  to  obtain  this;  it  then  finds  the  equilibrium  position  from  there.  You  should  therefore  set  the  Step  1   Statics  Method  to  give  a  reasonable  starting  shape.  See  Full  Statics.   For  more  details  of  the  Statics  Calculation  see  Statics  Analysis.   Warning:  

If   you   do   not   use   Full   Statics,   then   the   starting   position   will   not   (in   general)   be   an   equilibrium   position.  

Note:  

It  is  only  possible  to  include  buoys  in  the  static  analysis  (see   Buoy  Degrees  of  Freedom  Included  in   Static  Analysis)  if  either  the  Catenary  method  or  Full  Statics  is  used  for   all  lines  in  the  model.  

Include  Friction   Friction  can  be  included  in  the  static  analysis  only  if  the   Step  1  Statics  Method  is  Catenary  or  if  Full  Statics  is  used  for   the  Step  2  Statics  Method.   With   seabed   friction   present   there   is   not,  in   general,   a   unique   static   position   for   the   line,   since   the   position   it  adopts   depends  on  how  it  was  originally  laid  and  its  history  since  then.  In  order  to  define  a  unique  solution,  we  therefore  

277  

w

 

System  Modelling:  Data  and  Results,  Lines    

need  to  make  some  assumptions  about  how  the  line  was  originally  laid   and  friction  is  then  assumed  to  act  towards   this  position.   If  the   Step  1  Statics  Method  is  Prescribed,   then  this  'originally  laid'  position  is  assumed  to  be  the  position  defined  by   the  Prescribed  track.  Otherwise,  the  'originally  laid'  position  is  defined  by  specifying  the  Lay  Azimuth  and  As  Laid   Tension  values.   Lay  Azimuth   This   data   is   only   used   when   seabed   friction   is   included   in   the   static   analysis   and   the   Step   1   Statics   Method   is   not   Prescribed.  It  then  defines  the  position  in  which  the  line  is  assumed  to  have  been  originally  laid,  and  friction  is  then   assumed  to  act  towards  this  position.  When  Statics  Method  is  not  Prescribed,  it  is  assumed  that:   1.

The   line   was   originally   laid,   with   the   specified   As   Laid   Tension,   starting   with   the   Bottom   End   at   its   specified   position  (or  at  the  point  on  the  seabed  directly  below,  if  the  Bottom  End  is  not  on  the  seabed).  

2.

The  line  was  then  laid  in  the  Lay  Azimuth  direction,  leading  away  from  the  Bottom  End  position  and  with  the   specified  As  Laid  Tension.  

3.

The  line  was  laid  following  the  profile  of  the  seabed.  

4.

The  Top  End  was  then  moved  slowly  from  that  original  position  to  its  specified  position.  

To  help  set  this  data  item,  there  is  a  button  on  the  form  marked   Set.  This  button  sets  the  Lay  Azimuth  value  to  be  the   direction  from  the  Bottom  End  towards  the  Top  End,  based  on  their  current  positions.   Notes:  

Whilst   the  program  will   accept   any   Lay   Azimuth,  we   would   expect   the   statics   convergence  routine   to  have  increasing  difficulty  in  finding  a  solution  as  the  angle  between  the  Lay  Azimuth  direction   and  the  vertical  plane  through  the  line  ends  increases.  For  example,  if   we  have  a  line  top  at  X=0,   Y=0,  and  anchor  at  X=100,  Y=0,  we  would  expect  trouble  for  a  Lay  Direction  of  90°.  

 

The  Line  Setup  Wizard  also  uses  the  Lay  Azimuth  direction.  

As  Laid  Tension   This  data  specifies  the  effective  tension  with  which  the  line  was  originally  laid.  OrcaFlex  uses  this  to  determine  the   as-­‐laid   node   positions,   which   are   used   as   the   friction   target   positions   towards   which   friction   acts   in   the   static   analysis.  This  data  is  therefore  only  used  if  friction  is   included  in  statics.   If  the  Step  1  Statics  Method  is  set  to  Prescribed  starting  shape,  then  the  statics  friction  target  positions  are  laid  out   along  the  prescribed  shape  with  a  strain  determined  by  the  axial  stiffness  and  this  As  Laid  Effective  Tension  v alue.   If  the  Step  1  Statics  Method  is  not  Prescribed,  then  this  data  is  used  as  described  in  the  Lay  Azimuth  section  above.  

Catenary  Convergence   If   the   Catenary   statics   method   is   chosen,   then   an   iterative   catenary   calculation   is   used   to   determine   the   static   position  of  the  line.  This  calculation  is  controlled  by  a  number  of  convergence  parameters  which  can  normally  be  left   at  their  default  values.  However  sometimes  the  calculation  can  fail  to  converge.  If  this   happens,  first  check  your  data   for  errors  and  check  for  the  following  common  causes  of  convergence  failure:   x

Does   the  solution  have  a  slack  segment?  This  can  happen  in  lines  that  touch   down   on  the  seabed  almost  at  right   angles  or  in  lines  that  hang  in  a  very  narrow  U  shape.  The  catenary  calculation  cannot  handle  lines  with  slack   segments  Ȃ  try  increasing  the  number  of  segments  in  the  relevant  section  of  the  line.  

x

For  lines  that  touch  down  on  the  seabed,  is  the  Lay  Azimuth  value  specified  correctly?  It  is  the  azimuth  direction   leading  away  from  End  B  and  it  is  easy  to  get  it  wrong  by  180°.  

x

Is  the  line  buoyant,  either  deliberately  or  by  mistake.  The  catenary  calculation  has  problems  with  floating  lines  Ȃ   you  may  need  to  use  the  Spline  statics  method  instead.  

x

Does   the   line   have   a   surface-­‐piercing   buoyant   clump   attached?   If   the   clump   is   short   then   the   catenary   calculation  is  more  difficult.  

If  the  calculation  still  fails  to  converge,  then  it  is  sometimes  possible  to  obtain  convergence  by  changing  one  or  more   of  the  convergence  parameters,  as  outlined  below.  

278  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Max  Iterations  

The   maximum   number   of   iterations   that   OrcaFlex   will   make   before   treating   the   calculation   as   having   failed   to   converge.  Increasing  this  value  can  sometimes  help.   Tolerance  

The   non-­‐dimensional   accuracy   to   which   the   calculation   is   done,   before   the   calculation   is   treated   as   having   converged.  Increasing  the  tolerance  increases  the  chances  of  convergence  but  reduces  the  accuracy.   Min  Damping  

The  minimum  damping  factor  to  be  used  in  the  calculation.  Convergence  can  sometimes  be  achieved  by  increasing   this  parameter  to  a  value  greater  than  1   Ȃ  try  values  in  the  range  1.1  to  2.0.  The  minimum  damping  should  not  be  set   to  less  than  1.   Mag.  of  Std.  Error,  Mag.  of  Std.  Change  

These  parameters  control  the  maximum  size  of  the  change,  in  the  estimated  solution,  this  is  allowed  in  a  single  step.   Reducing  these  values  can  sometimes  help,  but  the  calculation  will  then  usually  require  more  iterations.   The  remaining  parameters  should  not  normally  be  changed.  For  further  information  contact  Orcina.  

Full  Statics  Convergence   The   numerical   method   used   to   solve   for   the   static   position   is   an   iterative   process   in   which   the   program   tries   to   converge   on   the   solution   in   a   series   of   steps.   This   process   is   controlled   by   a   number   of   convergence   parameters,   found  on  the  Line  data  form.   Convergence  Control  Method  

Either  Line   Search  or  Mag.  of   Std.   Error   /  Change.  We  recommend   using   the   default   value,   Line  Search,  but  if  this   fails  then  the  alternative  Mag.  of  Std.  Error  /  Change  may  succeed.   Max  Iterations  

The   calculation   is   abandoned   if   convergence   has   not   been   achieved   after   this   number   of   steps.   For   some   difficult   cases  simply  increasing  this  limit  may  be  enough.   Tolerance  

This  controls  the  accuracy  of  the  solution.  The  program  accepts  the  line  position  as  a  static  equilibrium  position  if   the  largest  out  of  balance  force  component  on  any  node  is  less  than  Tolerance  *  total  weight  in  air.   When   torsion   is   included   the   static   equilibrium   position   must   also   satisfy   the   condition   that   the   largest   out   of   balance   moment   component   on   any   node   is   less   than   Tolerance   *   total   weight   in   air   *   ODmax   where   ODmax   is   the   maximum  diameter,  contact  diameter,  stress  diameter  or  normal  drag  diameter  over  the  entire  line.   Reducing   the   Tolerance   value   will   give   a   more   accurate   static   equilibrium   position,   but   will   take   more   iterations.   OrcaFlex   may   not   be   able   to   achieve   the   Tolerance   specified   if   it   is   too   small,   since   the   computer   has   limited   numerical  precision.   Delta  

This  is  a  perturbation  size,  used  to  calculate  the  Jacobian  matrix  for  the  problem.  Delta  should  always  be  less  than   the  tolerance  specified.   Note:  

This  parameter  is  only  used  when  torsion  is  included.  

Min  Damping,  Max  Damping  

For  some  cases  it  is  necessary  to  control  the  convergence  process   by  damping  down   (i.e.   reducing)  the  step  taken  at   each   stage.   The   program   includes   an   automatic   damping   system   that   chooses   a   suitable   damping   factor   for   each   iteration,  but  the  user  can  set  the  minimum  damping  and  maximum  damping  factors  that  are  used.   Normally  the  default  values  will  suffice  but  for  difficult  cases  the  default  values  can  be  altered.  For  cases  that  appear   to   make   the   convergence   unstable   (e.g.   giving   very   bad   line   positions   on   the   screen)   try   increasing   the   Min   Damping   factor   to   a   value   greater   than   1,   say   values   in   the   range   1   to   10.   You   can   also   try   increasing   the   Max   Damping  factor,  say  to  values  in  the  range  10  to  100.  

279  

w

 

System  Modelling:  Data  and  Results,  Lines    

Note:  

Convergence  will  be  slower  with  larger  damping  values  so  you  should  use  the  smallest  values  that   result  in  statics  converging.  

Mag.  of  Std.  Error,  Mag.  of  Std.  Change  

These  parameters  are  only  available  when  the  Mag  of  Std.  Error  /  Change  convergence  control  method  is  selected.   For  some  cases  it  may  be  necessary  to  reduce  one  or  both  of  these  values  from  their  defaults.  Doing  so  is  likely  to   slow  down  the  convergence,  but  it  may  be  necessary  in  order  to  achieve  convergence!  

Spline  Starting  Shape   The  following  data  is  only  used  if  the  Spline  statics  method  is  specified.   Order  

This  sets  the  smoothness  of  the  spline  shape;  generally  order  3  is  reasonable.  If  a  higher  order  is  chosen,  a  smoother   curve  results.  The  order  cannot  exceed  the  number  of  spline  points.   Control  Points  

The  line  shape  is  specified  by  a  number  of  Control  Points.  The  first  and  last  control  points  are  automatically  placed   at   the   line   ends   A   and   B   respectively   and   OrcaFlex   generates   a   smooth   curve   between   the   first   and   last   control   points   and   passing   near   to   the   intermediate   control   points.   These   intermediate   control   points   may   be   adjusted   to   'pull'  the  curve  into  the  desired  shape.   The  first  and  last  control  points  correspond  to  line  ends  A  and  B  respectively.  The  line  is  stretched  to  the  specified   As  Laid  Tension  and  laid  out  following  the  spline  curve  starting  at  End  A  and  working  towards  End  B.   For  a  line  with  a   Free  end  the  line   is  laid  out  along  the  curve  until  End  B  is  reached.  If  the  length  around  the  curve  is   not  equal  to  the  stretched  line  length  then  the  end  will  either  fall  short  of  the  end  Estimated  Position  or  lie  beyond  it   (along  the  continuation  of  the  curve  along  its  'final'  direction).   For  a  line  with  a  Fixed  end,  Anchor  or  attached  to  some  object  the  curve  is  automatically  expanded  or  contracted  to   allow  the  end  to  lie  at  the  specified  end  position.  An  error  is  reported  if  this  process  fails.  

Prescribed  Starting  Shape   This  data  is  only  used  if  the   Prescribed  method  is  used  for  Step  1  statics.  It  can  be  found  in  the  Prescribed  Starting   Shape  page  on  the  line  data  form  and  can  be  edited  in  several  ways:   x

By  editing  the  Length  and  Turn  values  of  a  track  section  on  the  line  data  form.  OrcaFlex  then  creates  an  arc  of   the   specified   Length   and   Turn,   and   the   X   and   Y   coordinates   of   the   end   of   this   section,   and   all   subsequent   sections,  are  automatically  adjusted  to  match.  

x

By  editing  the   X  and   Y  coordinates  of  the  ends  of  a  track  section   on  the  line   data  form.  OrcaFlex  then  creates  the   (unique)   circular   arc   (or   straight   line)   that   is   a   smooth   continuation   of   the   previous   section   and   passes   through   the  new  (X,Y)   point.  The  Length  and   Turn  values  for  this  section,  and  the  X  and  Y   coordinates  for  subsequent   sections,  are  then  automatically  adjusted  to  match.  

x

By  dragging  the  end  points  of  the  track  sections  on  a  3D  view  using  the  mouse.  The  track  and  the  track  section   end  points  are  drawn  on  the  3D  views.  Dragging  a  track  section   end  point  is   equivalent  to   editing  its   X  and  Y   values,  as  described  above.  

The  individual  data  items  (see  Figure:  Plan  View  of  Example  Track)  are  as  follows:   End  A  Azimuth  

The  initial  direction  of  the  track.   Track  Sections  

The  number  of  sections  used  to  define  the  track.   Section  Length  

The  length  of  the  circular  arc  (or  straight  line  if  Section  Turn  =  0).  

280  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Section  Turn  

The  amount   by   which  the  track  azimuth  increases  over  this  section.  A  positive  value   denotes  a  turn  to   the  left,  when   viewed   from   above,   and   a   negative   value   denotes   a   turn   to   the   right.   A   value   of   zero   can   be   entered   to   specify   a   straight  track  section.   Section  Radius  

Š‡”ƒ†‹—•‘ˆ…—”˜ƒ–—”‡‘ˆ–Š‡…‹”…—Žƒ”ƒ”…ǤŠ‡”ƒ†‹—•‡“—ƒŽ•ȋͳͺͲȌȀȋɎȌǡ™Š‡”‡‹•–Š‡•‡…–‹‘Ž‡‰–Šƒ†‹•–Š‡ absolute  value  of  section  turn,  in  degrees.  For  straight   sections  (i.e.  if   Section  Turn  =  0)  the  radius  is  reported  as   Infinity.   Notes:  

This  is  a  reported  value,  not  an  editable  data  item,  and  is  hence  always  shown  in  grey.  

 

With  a  profiled  or  sloping  seabed  the  actual  track  on  the  seabed  will  have  a  slightly  different  radius   of  curvature  Ȃ  see  Laying  out  the  Line.  

Section  X  and  Y  

The   global   X   and   Y   coordinates   of   the   end   of   this   track   section.   You   can   either   edit   these   X   and   Y   coordinates   explicitly,  on  the  line  data  form,  or  else  by  dragging  the  end  point  on  a  3D  view.  If  you  edit  X  or  Y  then  OrcaFlex  fits  a   circular  arc  (starting  at   the   previous  section's  end   point)  through  the   new   end   point  and  the   Section  Length   and   Section  Turn  are  automatically  updated  to  match  this  new  arc.   Section  Z  

The  global  Z  coordinate  of  the  section  end  point  on  the  seabed.  This  is  a  reported  value,  not  an  editable  data  item,   and  is  hence  always  shown  in  grey.   Section  Arc  Length  

The  total  arc  length  to  the   end  of  the  section.   This  is  a  reported  value,  not  an   editable  data  item,  and  is  hence  always   shown  in  grey.   Section  Azimuth  

The   azimuth   direction   at   the   end   of   the   section.   This   is   a   reported   value,   not   an   editable   data   item,   and   is   hence   always  shown  in  grey.   Track  Pen  

This  controls  how  the  track  is  drawn.  You  can  switch  between  the  options  of  drawing  the  track  in  the  chosen  pen   and  not  drawing  it  at  all.   Laying  out  the  Line   The  track  data  defines  a  sequence  of  straight  lines  and  circular  arcs  in  the  horizontal  plane,  which  are  then  projected   vertically  onto  the  seabed  to  define  the  track  itself.  The  program  then  lays  the  line  out  along  the  track,  allowing  for   any  As  Laid  Tension  specified  by  the  user  on  the  line  data  form.   Because  the  line  is  modelled  as  a  series  of  straight  segments,  when  the  line  is  laid  out  along  a  curved  track  it  will   repeatedly   'cut   corners'   and   so   the   length   of  line   laid   along   a   given   curved   track   section   will   be   slightly   shorter   than   the  length  of  that  section.  The  size  of  this  discrepancy  reduces  as  more  segments  are  used.   If   End   A   is   above   the   seabed   then   the   height   above   the   seabed   varies   linearly   between   End   A   and   the   first   track   section  point,  reaching  the  seabed  at  the  end  of  the  first  track  section.  If  the  end  of  the  last  track  section  is  reached   before  all  the  line  has  been  laid  out,  then  the  rest  of  the  line  is  laid  out   in  a  straight  line   in   the   direction  of  the   end  o f   the  track.   Sloping  and  profiled  seabeds  

The  track  on  the  seabed  is  obtained  by  projecting  the  specified  circular  arcs  or  straight  sections  vertically  down  onto   the   seabed.   With   a   horizontal   seabed   this   vertical   projection   has   no   effect   on   the   shape   of   the   track.   But   with   a   sloping   seabed   the   vertical   projection   does   not   preserve   distances   and   this   causes   some   effects   that   users   should   note:   x

The   section   lengths   and   arc   lengths   that   appear   in   the   prescribed   starting   shape   data   table   are   lengths   in   the   horizontal   plane,   i.e.   before   projection   down   onto   the   seabed.   With   a   sloping   seabed   the   true   section   and   arc   lengths  on  the  seabed  will  differ,  the   difference  depending  on  the  slope  of  the  seabed.  The  actual  arc  lengths  can   be  obtained  by  running  the  static  analysis  and  looking  at  the  Full  Results  table  for  the  line.  

281  

w

 

System  Modelling:  Data  and  Results,  Lines    

x

The   section   radius   reported   in   the   prescribed   starting   shape   data   table   is   that   of   the   circular   arc   in   the   horizontal  plane,  i.e.   before   projection  down  onto  the  seabed.   When  the  circular  arc  is  projected  down  onto  a   sloping  seabed  the  resulting  track  section  is  slightly  elliptical  rather  than  circular,  so  again  the  actual  radius  of   curvature  will   differ.   The   actual   radii   of   curvature   can   b e   obtained   by  running   the   static   analysis  and   looking   at   the  Full  Results  table  for  the  line.  

User  Specified  Starting  Shape   Starting  Shape  

The  User  Specified   Starting   Shape  statics  method  places   each  node  at  the   position  specified  in  this   table.  If  torsion  is   modelled  then  node  orientations  can  also  be  specified.  

Drag  and  Wake   Drag  Formulation  

A  number   of  authors  have  proposed   formulae  to  model   how  the   drag   force  on  a  line   varies   with   the  incidence  angle.   OrcaFlex  offers  the  choice  of  the  Standard,  Pode  or  Eames  formulations.  All  of  these  use   drag  coefficients  that  are   specified  on  the  Line  Types  data  form.   For  details  of  the  formulations  see  the  Line  Theory  section.   Line  Wake  Interference   To   include   wake   interference   modelling   you   must   first   define   one   or   more  wake   interference   models.   See   the   Wake   Models  button  on  the  Line  data  form.   You   must   then   specify   which   line   sections   to   include   in   wake   modelling,   by   either   being   included   as   a   wake   generator   (an   'upstream'   section)   or   as   a   section   that   reacts   to   wake   (a   'downstream'   section),   or   both   (a   downstream  section  that  reacts  to  wake  generated  further  upstream,  but  also  generates  its  own  wake  that  further   downstream  sections  might  react  to).  For  details  see  the  Line  Wake  Interference  Data  on  the  Drag  &  Wake  page  of   the  Line  Data  Form.   Note:  

Wake  modelling   is   only  included   in   the   static   analysis   if  the  Statics   Method  is   set   to   Whole   System   Statics.   It   is   not   included   if   the   Separate   Buoy   and   Line   Statics   method   is   specified.   This   is   because  wake   effects   require   that   the   static   positions   of   the   lines  involved   are   calculated   together,   not  separately.  

How  Wake  Effects  Are  Modelled  

The  wake  models  are  steady  state  models  of  wake  effects.  Also  OrcaFlex  does  not  model  the  effect  that  wake  takes   time  to  convect  downstream.  OrcaFlex  therefore  only  attempts  to  model  the  steady  wake  effects.   Wake  is  generated  when  there  is  fluid  velocity   relative  to  the  upstream  cylinder,  so  both  fluid  motion  and  upstream   cylinder  motion  can  contribute  to  the  wake.  Therefore  the  velocity  OrcaFlex  uses  as  the  input  to  the  wake  model  is   the  steady  relative  velocity  Vs  given  by   Vs   =   [undisturbed   current   velocity   vector   at   upstream   cylinder   centre]   -­‐   [any   steady   starting   velocity   specified  for  the  model]   The   wake   effects   therefore   do   not   include   any   effects   of   wave   motion,   or   of   any   changes   in   upstream   cylinder   velocity  during  a  simulation.   Note:  

OrcaFlex  does   not  model  combined   wake  effects.  If  a  given  'downstream'   node   is  in  the  modelled   wake   of   more   than   one   'upstream'   node,   then   OrcaFlex   chooses   to   apply   the   wake   effects   of   the   upstream   node   that   gives   the   strongest   wake   effect   at   that   downstream   position   when   the   wake   effects  from  other  upstream   nodes  (which  give  weaker  wake  effects  at  that  point)  are  ignored.  So  if   you   are  modelling   riser   arrays,   for  example,   then   the   wake   effects   at   any   given   point   are   modelled   as   if   they   came   from   the   upstream   wake-­‐generating   node   that   gives   largest   wake   effect   in   isolation,  i.e.  with  the  other  upstream  nodes  were  absent.  

The   strength   of   a   given   node's   wake   decays   with   distance   downstream   and   also   with   distance   in   the   transverse   direction,   as   specified   by   the  wake   model.   But  in   addition   OrcaFlex   also   makes   the  wake   strength  decay  in  the  axial   direction   of   the   upstream   line,   so   that   the   strongest   wake   selected   comes   from   the   upstream   node   that   is   axially   closest  to  the  downstream  node.  More  details  are  given  in  the  paper  by  Wu  et  al.  

282  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Line  Wake  Interference  Data   Wake  modelling   requires  significant  computation,  so   to  avoid  slowing  simulations   unnecessarily  OrcaFlex  enables   you   to   only   include   wake   modelling   for   sections   of   line   where   you   think   wake   effects   might   be   significant.   This   is   done  on  the  Drag  &  Wake  page   of  the  line  data  form,  by  specifying  which  line  sections  generate  modelled  wake  (the   'upstream'  sections)  and  which  line  sections  react  to  modelled  wake  (the  'downstream'  sections).   The   fluid   forces   applied   to   sections   that   react   to   modelled   wake   will   include   the   strongest   wake   effect   from   the   sections  that  generate  modelled  wake.  The  following  two  data  items  specify  which  sections  generate  and/or  react   to  wake:   Wake  Generated  

This  can  be  set  to  one  of  the  defined  wake  models,  in  which  case  OrcaFlex  will  model  wake  generated  by  that  line   section   acting   as   an   'upstream'   line.   Or   it  can   be   set   to  None,   in   which   case   OrcaFlex   will   not   model   wake   generated   by  that  line  section.   Three  types  of  wake  model  are  available.   x

The  Huse  model  is  an  analytic  wake  model  that  models  the  velocity  reduction  and  hence  drag  reduction  on  the   downstream  object,  but  does  not  model  the  wake  lift  effect.  

x

The   Blevins   model   is   an   analytic   model   that   models   the   both   the   drag   reduction   and   the   wake   lift   force   that   tends  to  draw  the  downstream  object  into  the  centre  of  the  wake.  

x

The  User  Specified   model  allows  you  to  model   both  the  drag  reduction  and  wake  lift   effect  using  your  own   data,   by  specifying  the  drag  and  wake  lift  coefficients  as  a  function  of  the  position  of  the  downstream  object  relative   to  the  wake  of  the  upstream  object.  

See  the  Wake  Models  section  for  details.   Reacts  to  Wake  

If   Reacts   to   Wake   is   turned   on   then   each   node   in   the   section   will   act   as   a   'downstream'   cylinder   in   the   wake   modelling.   Those   nodes   will   then   be   subject   to   the   strongest   wake   effects   (i.e.   strongest   at   that   downstream   position)   from   any   upstream   node   that   generate   modelled   wake.   Sections   with   Reacts   to   Wake   turned   off   will   ignore  any  modelled  wake  generated  by  upstream  sections.   So   to   summarise,   the   wake   modelling   will   include   the   strongest   wake   effects   on   downstream   sections   that   have   Reacts  to  Wake  turned  on,  due  to  modelled  wake  from  upstream  nodes  with  Wake  Generated  set  to  a  wake  model.   Note:  

A  line  section  is  allowed  to  be  both  a  'downstream'  section,  i.e.  Reacts  to  Wake  is  turned  on,  and  at   the   same   time   also   act   as   an   'upstream'   section,   i.e.   generate   modelled   wake   that   further   downstream   line   sections   can   react   to.   However   the   wake   effects   of   such   a   section   will   be  based   on   the  undisturbed  steady  relative  flow  velocity,  so  the  wake  effects  of  this  wake-­‐generating  section   will   not   take   into   account   any   velocity   reduction   due   to   it   in   turn   being   in   the   wake   of   another   wake-­‐generating  section  further  upstream.  

Wake  Models   The  Wake   Model   Data   form  enables   you  to   define   one   or   more  models   of  wake   interference.  The   wake  interference   model   specified   determines   how   the   flow   velocity,   wake   drag   reduction   and   wake   lift   force   on   a   downstream   cylinder  varies  as  a  function  of  the  (x,y)  position  of  the  downstream  cylinder  centre  relative  to  the   wake  frame  of   reference  of  the  upstream  cylinder.   You  can  define   more  than   one   wake  model,  for   example   if  you  want  to  use  different   wake  models  to  model  the  wake   generation  by  different  lines.  A  wake  model  that  isn't  used  by  any  line  in  the  model  will  be  ignored,  so  you  can  define   wake  models  and  then  decide  later  which  (if  any)  to  use  to  model  wake  generation.   Wake  frame  of  reference  

The  wake  modelling  is  done  relative  to  a  wake  frame  of  reference  that  is  based  on  the   steady  relative  fluid  velocity   vector  Vs  at  the  upstream  cylinder,  and  is  given  by:   x

The  wake  frame  origin  is  at  the  upstream  cylinder  centre.  

x

The  wake  frame  x-­‐axis  is  in  the  direction  of  the  steady  relative  fluid  velocity  vector  Vs.  

283  

w

 

System  Modelling:  Data  and  Results,  Lines    

x

The  wake   frame  z-­‐direction   is  the  direction  obtained   by   projecting  the  upstream  cylinder  axial  direction   normal   to  Vs.  It  is  therefore  the  direction  normal  to  Vs  and  in  the  plane  formed  by  the  Vs  and   the  cylinder  axial  direction.   The  +ve  wake  z-­‐direction  is  towards  end  B  of  the  upstream  line.  

x

The   wake  frame  y-­‐axis  is  the  direction  that  completes  the  orthogonal  right-­‐hand  triad  of   wake  axes   x,y,z.  It   is   therefore  normal  to  the  plane  formed  by  Vs  and  the  cylinder  axial  direction.  

The  wake  model  theory  below  uses  the  following  key  variables.  Upper  case  subscript  'D'  denotes  'drag',  lower  case   subscripts   'u'   and   'd'   denote   'upstream'   and   'downstream',   and   subscript   '0'  denotes   'undisturbed',   i.e.   ignoring  any   wake  effects.   Vd(x,y)  is  the  disturbed  fluid  velocity  vector  at  downstream  position  (x,y)  relative  to  the  upstream  cylinder   wake,  allowing  for  wake  effects.   Vd0(x,y)  is  the  undisturbed  fluid  velocity  vector  at  position  (x,y).   Vu0  is  the  undisturbed  fluid  velocity  at  the  upstream  cylinder  centre.   Du,   Dd   and   CDu0,   CDd0   are   the   normal   drag   diameter   and   undisturbed   drag   coefficient   of   the   upstream   and   downstream   cylinders,   respectively,   as   specified   on   the   line   type   data   form.   Note   that   the   wake   modelling   does  not  allow  for  any  non-­‐isotropic  aspects  of  the  drag  coefficients.  If  different  drag  coefficients  have  been   specified  for  the  line  x-­‐  and  y-­‐directions  then  the  wake  modelling  uses  the  mean  of  the  two  values.   Wake  Model  Data  

Each  wake  model  is  given  a  user-­‐specified  Name.  And  each  model  can  be  one  of  three  types:  Huse  model,  Blevins   model  or  User-­‐Specified  model,  as  described  below.   Huse  Model  

This   is   an   analytic   model   proposed   by   Huse   (1993).   It   models   the   wake   velocity   reduction,   and   hence   drag   reduction,  but  does  not  give  any  wake  lift  force.   In  the  Huse  model  the  wake  fluid  flow  velocity  Vd(x,y)  at  a  given  position  (x,y)  in  the  wake  is  given  by:    Vd(x,y)  =  Vd0(x,y)  -­‐  k2Vu0[CduDu/xs]½exp[-­‐k3(y/b)2]   where   xs  =  x  +  4Du/Cdu   b  =  k1[CduDuxs]½   and  the  other  terms  are  defined  above.   k1,   k2   and   k3   are   user-­‐editable   non-­‐dimensional   model   parameters.   They   should   normally   be   left   as   the   default   values,  since  these  give  the  original  Huse  model.  (Note  that  the  parameter  k 3  is  mis-­‐printed  in  Huse,  1993  as  0.639   and   was   corrected   in   a   later   paper   to   0.693.)   Changing   these   model   parameters   from   their   default   values   will   give   a   variant  of  the  Huse  model.   Blevins  model  

The   Blevins   model   is   an   analytic   model   that   models   velocity   and   drag   reduction,   and   also   models   the   wake   lift   force   that   tends   to   draw   the   downstream   object   into   the   centre   of   the   wake.   See   the   Appendix   of   Blevins   OMAE   2005   paper  for  the  theory  of  the  model.   The  model  has  3  non-­‐dimensional  model  parameters  a1,  a2,  a3  that  are  user-­‐editable.   They  should  normally  be  left  as   the   default   values,   since  these   are   the   values   given  by  Blevins.   Changing   these   model   parameters   from   their  default   values  will  give  a  variant  of  the  Blevins  model.   User  Specified  model  

This  option  allows  you  to  define  a  model  (of  both  wake  drag  reduction  and  wake  lift  effect)  by  specifying  drag  and   lift  coefficient  factors  as  a   function  of  the  position  of  the  downstream  object   relative  to  the   wake  of  the  upstream   object.   The   wake   effects   are   specified   by   giving   a   table   of   wake   drag   and   lift   coefficient   factors   for   the   downstream   cylinder,  as  a  function  of  the   downstream  cylinder  position  relative  to  the  upstream  cylinder  wake,  as  follows.   See   above  for  notation.   x

The   Position   columns   of   the   table   define,   in   non-­‐dimensional   form,   a   number   of   downstream   cylinder   centre   positions   relative   to   the   upstream   cylinder   wake   frame   of   reference.   This   is   done   by   specifying   non-­‐

284  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

dimensionalised   distances   L/Du   (downstream)   and   T/Du   (transverse)   from   the   upstream   cylinder   centre   to  the   downstream  cylinder  centre,  where  Du  is  the  normal  drag  diameter  of  the  upstream  cylinder.   x

The   Coefficient   Factor   columns   of   the   table   define   the   wake   effects   at   the   given   (L/D u,   T/Du)   positions,   by   specifying   drag   and   lift   coefficient   factors.   Note   that   these   data   are   scaling   factors,   not   the   drag   and   lift   coefficients   themselves.   The   drag   factors   are   non-­‐negative   scaling  factors   that   are   applied   to   the   reference   drag   coefficient  CDd0  that  is  specified  on  the  line  type  data  form.  The  lift  coefficient  factors  are  signed  scaling  factors   that   are   applied   to   the   Reference   Wake   Lift   Coefficient   CLd0,   which  is   specified   on   the   wake   models   data   form.   A   +ve   lift   coefficient   factor  means   a   lift   force  in   the   +ve  wake   frame   y -­‐direction,   so  the   lift   coefficient   factor   at   a   given  T/Du  position  will   normally  have   the   opposite   sign   to   the  T/Du  value,   since  wake   lift   effects   are   normally   towards  the  centre  line  of  the  wake.  

Wake   drag   effects   are   normally  symmetric,   and  wake   lift   effects   anti-­‐symmetric,  either   side   of   the   wake   centre   line.   So  to  avoid   the   need   to  specify   in   the   table   both   +ve   and  -­‐ve   values   of   T/Du   you   can   tell   OrcaFlex   to  Reflect   Data.   In   this  case  you  must   only  specify  table   rows  for  one  half  of  the   wake  plane,  i.e.   either   for  T/D u>=  0  only,  or   for  T/Du<=   0   only.   For   a   wake   frame   position   (x,y)   that   is   on   the   other   side   of   the   wake   centre   line   OrcaFlex   will   then   automatically   use   the   drag   and   lift   coefficient   factors   defined   for   the   symmetric   position   (x,-­‐y),   but   with   the   lift   coefficient  factor  negated.   OrcaFlex   uses   linear   triangular   interpolation   to   obtain   the   drag   and   lift   coefficient   factors   to   use   for   wake   frame   positions  between  those  specified  in  the  table.  

Results   Log  Results  

This  option  is  checked  by  default  and  this  means  that  simulation  results  at  all  points  on  the  Line  are  available.  If   this   option  is  unchecked  then  no  simulation  results  are  available  for  this  Line.   OrcaFlex  stores  simulation  results  in  an  efficient  way,  only  logging  a  minimal  set  of  variables  to  the  simulation  file.   Other   results   variables   which   have   not   been   logged   are   then   derived   when   the   results   are   requested.   Usually   this   means  that  simulation  files  are  a  reasonable  size  and  we  recommend  that  this  value  is  checked.   Should  you  need  to  reduce  the  size  of  simulation  files  then  this  option  can  be  unchecked  for  those  lines  for  which   you  do  not  need  results.   Arc  length  axis,  Arc  length  axis  inverted,  Value  axis  inverted  

These  data  items  allow  you  customise  the  way  range  graphs  are  displayed.  The  Arc  length  axis  setting  allows  you  to   control   whether  the  arc  length  axis  is  horizontal  or  vertical.  The  latter   option   would  typically  be  used  for  vertical   risers.   Normally  the  axes  on  an  OrcaFlex  graph  display  increasing  values  to  the  right  (for  a  horizontal  axis)  or  upwards  (for   a   vertical   axis).   The   axis   inverted   options   allow   you   to   reverse   the   axes.   Again   this   would   typically   be   used   for   vertical   risers   to  arrange   that   up   and   down   on   the   arc   length   axis   of  the   graph   matched   up   and   down   in   the   physical   system  being  modelled.   Since  changes  to  these  settings  are  usually  motivated  by  the  physical  layout  of  the  line  in  question,  each  line  in  an   OrcaFlex  model  has  its  own  copies  of  these  settings.   These  data  items  can  also  be  set  on  the  results  form.  

Drawing   Nodes    

You  can  define  the  colour,  line  style  and  thickness  of  the  pens  used  for  drawing  the  nodes  and  sections  of  the  line.   See   How   Objects   Are   Drawn.   You   can   also  choose   to   draw  nodes   as   circular   discs   with   diameter   equal   to   the  contact   diameter.   Segments  

There  is  a  choice  for  which  pen  is  used  to  draw  the  segments.  You  may  either  specify  the  pen  explicitly  on  the  Line   Data   form,   in   which   case   it   will   be   used   for   all   segments   of   that   line.   This   allows   you   to   use   different   pens   to   distinguish  between  different  lines.  Alternatively,  you  can  choose  to  have  the  segments  drawn  using  the  appropriate   Line   Type   Pen   defined   on   the   Line   Types   form.   This   allows   you   to   use   different   pens   to   distinguish   sections   of   different  line  types.  

285  

w

 

System  Modelling:  Data  and  Results,  Lines    

Node  axes  

Node   axis   directions   (x,   y,   z)   can   be   given   individual   colours.   This   helps  distinguish  between   x   and   y  directions   thus   making   component   results   easier   to   interpret.   The   node  axes   directions   are   drawn   optionally  and   can   be   controlled   by  the  Draw  Node  Axes  preference  or  by  pressing  CTRL+ALT+Y.   Contact  

You  can  define  a  contact  pen  which  is  used  when  drawing  nodes  and  segments  which  are  in  contact  with  the  seabed,   elastic  solids  and  other  lines.  Should  you  wish  you  can  choose  to  disable  the  contact  drawing.   Prescribed  Statics  Method  (Track)  

For   Lines   with   Prescribed   Statics   Method   you   can   control   how   the   track   is   drawn.   You   can   switch   between   the   options  of  drawing  the  track  in  the  chosen  pen  and  not  drawing  it  at  all.   Spline  Starting  Shape  

For   the  Spline  Starting  Shape   you   can   switch  between   the   options   of   drawing   the   unscaled   spline   in   the   chosen  pen   and  not  drawing  it  at  all.  

VIV  Drawing   The  VIV  Drawing  page  is  visible  when  a  time  domain  VIV  model  is  used.  For  details  see  the  VIV  Drawing  topic.  

Properties  Report   The  Line  properties  report  is  available  from  the  popup-­‐menu  on  the  data  form.  It  reports  the  following:   Total  length  

The  sum  of  all  the  section  lengths.   Total  weight  in  air  

The  force  due  to  gravity  of  the  entire  line.  The  contents  are  included  but  the  weight  of  any  attachments  is  excluded.   Total  displacement  

The   weight   of   water   displaced   by   the   entire   line's   volume.   The   displacement   of   any   attachments   is   excluded.   The   reported  value  uses  the  water  density  at  the  sea  surface.   Total  weight  in  water  

Equals  Total  line  weight  in  air  -­‐  Total  line  displacement.   Full  Statics  force  accuracy,  Full  Statics  moment  accuracy  

Full  Statics  force  accuracy  is  only  reported  if  the   line  uses  the  Full  Statics  method.  Full  Statics  moment  accuracy  is   only  reported  if  the  line  uses  the  Full  Statics  method  and  torsion  is  included.   The  Full  Statics  method  finds  an  equilibrium  configuration  Ȃ  that  is  a  set  of  node  positions  for  which  the  resultant   force   and   moment   on   each   node   is   zero.   We   refer   to   the   resultant   force   and   moment   as   the   out   of   balance   load.   Because   computers   have   limited   numerical   precision   the   static   analysis   cannot   always   find   a   configuration   where   the  out  of  balance  load  is  exactly  zero.  OrcaFlex  accepts  a  position  as  a  static  equilibrium  position  if  the  largest  out  of   balance  load  component  is  less  than  the  statics  accuracy.   The   Full   Statics   force   accuracy   equals   Tolerance   *   line   typical   force   and   the   Full   Statics   moment   accuracy   equals   Tolerance   *   line   typical   moment.   The   line   typical   force   is   defined   to   be   the   total   weight   in   air.   The   line   typical   moment  is  defined  to  be  total  weight  in  air  *  total  length.   Reducing   the   Tolerance   value   will   give   a   more   accurate   static   equilibrium   position,   but   will   take   more   iterations.   OrcaFlex   may   not   be   able   to   achieve   the   Tolerance   specified   if   it   is   too   small,   since   the   computer   has   limited   numerical  precision.  

6.8.2

Line  Types  

Data   The   Line   Types   form  defines   the   properties   of   a  number   of  named   line   types,  which   can   then   be   used   to   specify  the   structure  of  the  Lines  used  in  the  model.  

286  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

The   line   types   form   must   include   all   the   line   types   referred   to   on   all   of   the   Lines   forms,   but   it   can   also   include   other   line   types   that   are   not   currently   in   use   in   the   model.   This   allows   you   to   build   up   a   library   of   standard   line   types   which  can  then  be  easily  used  when  building  Lines.   There  is  not  enough  room  on  the  screen  to  show  all  the  properties  of  all  the  line  types,  so  OrcaFlex  offers  different   view  modes:   x

Individual  mode  shows  one  line  type  at  a  time,  but  shows  you  all  its  properties.  

x

All  mode  shows  all  the  line  types,  but  different  types  of  properties  are  shown  in  different  tables.  

x

Code  Checks  mode  shows  data  used  for  post-­‐processing  code  checks.  

x

External   Function   Parameters   mode   shows   the   data   used   by   any   external   functions   referenced   by   the   Line   Type.  

The   Line   Type   Wizard   is   available   to   help   set   up   line   type   data   to   represent   commonly   used   structures   such   as   chains,  ropes  etc.   Line  Type  Name   Used  to  refer  to  the  Line  Type.   Category   Can  be  either:   x

Homogeneous  Pipe:  This   option  is  appropriate  for  a  pipe  constructed   from  a  single  homogeneous   material,  for   example   a   metal   riser,   or  when  modelling   stress   joints  and   bend   stiffeners.   The  pipe's   structural   properties   are   defined  by  specifying  Young's  modulus,  material  density  and  pipe  diameters.  When  modelling   stress  joints  and   bend  stiffeners  the  outer  diameter  can  be  specified  as  varying  with  arc  length.  

x

General:   This   option   is   used   in   all   other   situations.   The   axial,   bending   and   torsional   stiffnesses   are   directly   input  instead  of  being  calculated  from  E.  Similarly  the  mass  is  specified  as  mass  per  unit  length  as  opposed  to   being   calculated   from   a   material   density.   This   approach   allows   analysis   of   flexible   risers,   umbilicals,   hoses,   mooring  chains,  ropes,  wires,  bundles,  seismic  arrays,  power  cables,  nets  etc.  

Geometry  &  Mass  Data   Outer  and  Inner  Diameter   Used  to  define  buoyancy  and  mass  of  contents  per  unit  length  respectively.   These  data  can  also  be  used  for  other  purposes  as  follows:   x

If   the   Stress   Diameters   are   set   to   '~'   then   these   diameters   are   used   for   wall   tension   and   stress   results   calculations.  

x

If  the  Contact  Diameter  is  set  to  '~'  then  the  outer  diameter  is  used  for  contact  calculations.  

x

If  the  Drag  /  Lift  Diameters  are  set  to  '~'  then  the  outer  diameter  is  used  for  drag  calculations.  

Profiled  line  types  (homogeneous  pipe  only)  

For  homogeneous  pipes  the   outer  diameter  can  vary   with  arc  length.  To  do  this  you   first  specify  the   profile  in  a  Line   Type  Outer  Diameter  variable  data  source  which  is  then  referenced  by  the  outer  diameter  data  of  the  line  type.   This  feature  is  used  when  modelling  stress  joints  and  bend  stiffeners.  Arc  length  is  defined  relative  to  the  start  of  the   line  section  which  uses  this  line  type  and  increases  from  End  A  towards  End  B.   CG  Offset   The   x  and  y  coordinates  of  the  centre  of  gravity  (CG)  relative  to  the  centreline.  These  data  items  are  only  used  when   torsion  is  being  modelled.  Note  that  if  the  line  has  contents  then  the  contents  CG  is  assumed  to  be  at  the  centreline   and  is  not  affected  by  this  CG  Offset.   Bulk  Modulus   Specifies  the  compressibility  of  the  line  type.  If  the  line  type  is  not  significantly  compressible,  then  the  Bulk  Modulus   can  be  set  to  Infinity,  which  means  incompressible.  See  Buoyancy  Variation.  

287  

w

 

System  Modelling:  Data  and  Results,  Lines    

Material  Density  (homogeneous  pipe  only)   The  density  of  the  material.   Mass  per  Unit  Length   The   mass   of   the   line   or   pipe   structure,   excluding   contents,   per   unit   length.   For   homogeneous   pipes   the   material   density  is  used  to  calculate  the  structural  mass  and  therefore  the  mass  per  unit  length  data  item  cannot  be  edited.  

Coatings  &  Linings  Data   Coatings  and  Linings  are  available  for  homogeneous  pipe  only.   They   are   typically   used   with   steel   pipes   to   model   the   additional   mass   and   displacement   of  concrete   coatings,   plastic   linings   etc.  They  contribute   mass,  weight  and  displacement  and  also  modify  the   pipe's  inner  and  outer  diameters.   However,  they  contribute  no  additional  structural  strength  and  are  assumed   not  to  be  load  bearing.  Stress  results   are  calculated  based  on  stress  diameters  equal  to  the  underlying  pipe  diameters.   Thickness  and  Material  Density  

The  thickness  specifies  the  physical  dimension  of  the  coating  or  lining.  For  a  coating  the  buoyancy  diameter,  contact   diameter   and  drag  diameters   are   all   increased   by   twice  the   coating   thickness.   Similarly  the   pipe's   inner   diameter   is   reduced  by  twice  the  lining  thickness  which  has  the  effect  of  reducing  the  contents  mass  and  weight.   The  material  density  is  used,  together  with  the  thickness,  to  calculate  the  additional  mass  and  weight  of  the  coating   or  lining.   Multiple  layers  

The   data   on   the   Line   Type   data   form   only   allow   single   coating   or   lining   layers   to   be   specified.   It   is,   however,   also   possible   to   specify   multiple   layers   of   coatings   or   linings.   To   do   this   you   define   Coatings   or   Linings   variable   data   sources  which  are  then  referenced  by  the  thickness  data  on  the  Line  Type  data  form.   These   data   sources   specify   a   list   of   layers   defined   by   thickness   and   material   density.   The   layers   are   listed   in   the   order  of  application.  In  other  words  the  first  layer  is  immediately  adjacent  to  the  pipe,  the  second  layer  is  next  to  the   first  layer,  and  so  on.  

Limits  Data   Limit  Compression   The   program   has   two   modes   for   handling   slack   segments,   i.e.   when   the   distance   between   two   adjacent   nodes   becomes  less  than  the  original  unstretched  segment  length:   x

No   means  that  the  segment  is  treated  as  a  strut   which  can  support  unlimited  compression.  This  is  the   preferred   model  except  where  bend  stiffness  is  insignificant.  

x

Yes  means  that  the  segment  is  treated  as  an  elastic  Euler  strut   Ȃ  the  compression  is  limited  to  the  segment  Euler   load.  This  is  a  better  model  for  cases  where  the  bend  stiffness  is  insignificant,  such  as  for  chains  and  soft  ropes.  

Š‡•‡‰‡–—Ž‡”Ž‘ƒ†‹•‰‹˜‡„›Ɏ2EI/L02  where  EI  is  the  bending  stiffness  of  the  pipe  and  L 0  is  the  unstretched   length  of   the   segment.   In   all   cases,   whenever   a   segment  has   been   compressed  to   or  beyond   the   segment   Euler   load,   then  a  warning  of  this  is  given  on  the  results  form  and  in  the  statistics  table.   For  items  such  as  mooring  chain,  the  bending  stiffness  is  zero,  and  the  segment  Euler  load  is  also  zero.  In  this  case   "Limit   Compression"   should   be   set   to   "Yes"   Ȃ   this   correctly   models   a   chain   or   very   flexible   rope,   which   cannot   support  any  compression.  The  segment  Euler  load  warning  is  then  simply  a  warning  that  the  line  has  gone  slack.   For   a   line   with   non-­‐zero   bend   stiffness   the   Euler   load   warning   is   effectively   a   warning   that   the   segments   at   that   point   are   too   long   to   accurately  model   the   bending   that   is   occurring.   Effectively,  bending   is  occurring   at   a   scale   that   is  less  than  the  segment  length,  so  shorter  segments  are  needed  to  model  it  accurately.  Using  shorter  segments  in   that  area  will  give  a  larger  segment  Euler  load,  and  to  obtain  an  accurate  solution  you  should,  ideally,  use  sufficiently   short   segments   that   the   resulting   segment   Euler   load   is   not   reached.   See   Line   Compression   and   Modelling   Compression  in  Flexibles  for  details.  

288  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Allowable  Tension   The   maximum   permitted   tension   for   this   type   of   line.   This   value   is   used   to   draw   a   limit   curve   on   Tension   Range   Graphs  and  to  calculate  the   Normalised  Tension.  It  does  not  limit  the  tension  achieved  in  the  line.  If  no  limit  curve  is   wanted  then  you  may  input  "~".   Minimum  Bend  Radii  (MBR)   You  can  specify  the  minimum  permitted  radii  of  curvature   for   bending  about  the  x  and  y-­‐directions.  These  values   are  optional  Ȃ  they  are  only  used  to  draw  "allowable"  curves  on  range  graphs  and  in  the  calculation  of  Normalised   Curvature  results.  They  do  not  limit  the  bend  radius  of  the  line.  If  you  do  not  want  these  curves  then  set  the  x-­‐radius   to  "~"  (meaning  "not  applicable")  and  the  y-­‐value  to  "~"  (meaning  "same  as  x-­‐value").   Often   the   radii   for   the   x   and   y-­‐directions   are   equal   and   this   can   be   specified   by   setting   the   y-­‐radius   to   "~"   which   means  "same  as  x-­‐radius".   The  specified  values  are  used  to  draw  "allowable  curvature"  curves  on  the   x  and  y -­‐Curvature  range  graphs,  and   also   (if  the  x  and  y-­‐minimum  radii  are   equal)  on  the  Curvature  range  graph.  In  addition,  they  are  used  (together  with   the   specified  bend  stiffness)  to  derive  "allowable  bend  moment"  curves  which  are  drawn  on  the  x  and  y-­‐Bend  Moment   range  graphs,  and  also  (if  the  x  and  y-­‐values  are  equal)  on  the  Bend  Moment  range  graph.  The  MBR  is  also  used  to   calculate  Normalised  Curvature.   Notes:  

MBR  can   be  specified  as  a   variable   data  item   defined  as  a  table  relating  MBR  to  tension   Ȃ  either   wall   tension   or   effective   tension   can   be   used.   A   typical   example   of   this   would   be   a   steel   core   umbilical.  A  variable  MBR  is  used  solely  in  the  calculation  of  Normalised  Curvature.  

 

The   "allowable"   curve   may   not   be   visible   on   the   range   graph,   since   it   may   be   outside   the   range   covered  by  the  graph.  To  see  the  "allowable"  curve  in  this   case  you  will  need  to  modify  the  graph  to   increase  the  range  of  values  covered.  

Structure  Data   Young's  Modulus  (homogeneous  pipe  only)   The  Young's  modulus  of  the  material  which  can  be  constant  or  variable.   x

A  constant  value  results  in  linear  material  properties.  

x

A   variable   data   item   specifies   a   non-­‐linear   stress-­‐strain   relationship   which   results   in   a   bending   stiffness   with   non-­‐linear  elastic  behaviour.  Note  however  that  the  axial  and  torsional  stiffnesses  are  still  assumed  to  be  linear.  

For  homogeneous  pipes  the  Young's  modulus  determines  the  axial,  bending  and  torsional  stiffnesses.  These  stiffness   data  items  are  reported  on  the  data  form,  although  they  cannot  be  edited.   Bend  Stiffness   The   bend   stiffness   is   the   slope   of   the   bend   moment-­‐curvature   curve.   You   can   specify   separate   values   for   bending   about  the   x  and  y-­‐directions,  but   often  these  are  equal.  This  can   be  achieved  by  setting  the  y-­‐bend  stiffness  to  '~'   which  means  'same  as  x-­‐bend  stiffness'.   The   bend   stiffness   specified   may   be   zero,   for   example   for   chains.   It   can   also   be   very   large   values,   for   example   for   steel  pipes,  but  this  will  often  result  in  short  natural  periods  in  the  model  and  hence  require  short  simulation  time   steps.  See  Inner  and  Outer  Time  Steps.   You  can  specify  the  bend  stiffness  to  be   linear,  elastic  non-­‐linear,  hysteretic  non-­‐linear  or  externally  calculated,   as  follows.  See  Calculating  Bend  Moments  for  further  details  of  the  bending  model  used.   Linear  Bend  Stiffness  

For   normal   simple   linear   behaviour,   specify   the   bend   stiffness   to   be   the   constant   slope   of   the   bend   moment-­‐ curvature   relationship.   This   slope   is   the   equivalent   EI   value   for   the   line,   where   E   is   Young's   modulus   and   I   is   the   moment   of   area   of   the   cross-­‐section.   The   bend   stiffness   equals   the   bend   moment   required   to   bend   the   line   to   a   curvature  of  1  radian  per  unit  length.   Non-­‐Linear  Bend  Stiffness  

For   non-­‐linear   behaviour,   use   variable   data   to   specify   a   table   of   bend   moment   magnitude   against   curvature   magnitude.   OrcaFlex   uses   linear   interpolation   for   curvatures   between   those   specified   in   the   table,   and   linear  

289  

w

 

System  Modelling:  Data  and  Results,  Lines    

extrapolation  for   curvatures   beyond   those   specified  in  the   table.   The  bend   moment  must   be   zero   at   zero   curvature.   For  homogeneous  pipes  the  Plasticity  Wizard  may  be  useful  to  help  set  up  the  table.   With  non-­‐linear  bend  stiffness  you  must  also  specify  whether  the  hysteretic  bending  model  should  be  used.   x

Non-­‐hysteretic   means   that   the   non-­‐linear   stiffness   is   elastic.   No   hysteresis   effects   are   included   and   the   bend   moment  magnitude  is  simply  the  specified  function  of  the  current  curvature  magnitude.  

x

Hysteretic  means  the  bend  moment  includes  hysteresis  effects,  so  that  the  bend  moment  depends  on  the  history   of  curvature  applied  as  well  as  on  the  current  curvature.  Also  note  that  if  the  hysteretic  model  is  used  then  the   line  must  include  torsion  effects.   Warning:  

You   must   check   that   the   hysteretic   model   is   suitable   for   the   line   type   being   modelled.   It   is   not   suitable   for   modelling   rate-­‐dependent   effects.   It   is   intended   for   modelling   hysteresis   due   to   persisting   effects   such   as   yield   of   material   or   slippage   of   one   part   of   a   composite   line   structure   relative  to  another  part.  

If  you  use  the  hysteretic  bending  model  then  the  simulation  speed  may  be  significantly  slowed  if  there  are  a  large   number   of   points   specified   in   the   table   of   bend   moment   against   curvature.   You   might   be   able   to   speed   up   the   simulation,  without  significantly  affecting  accuracy,  by  removing  superfluous  points  in  areas  where  the  curve  is  very   close  to  linear.   In  addition  hysteretic  bend  stiffness  usually  requires  shorter  time  steps  than  non-­‐hysteretic  modelling,  when  using   implicit  integration.   Note:  

If  you   are   using   non-­‐linear   bend   stiffness,   then   the  mid-­‐segment   curvature   results   reported   depend   on   whether   the   bend   stiffness   is   specified   to   be   hysteretic   or   not.   If   the   bend   stiffness   is   not   hysteretic   then   the   mid-­‐segment   curvature   reported   is   the   curvature   that   corresponds   to   the   mid-­‐ segment  bend  moment  (which  is  the  mean  of  the  bend  moments  at  either  end  of  the  segment).  If  the   bend  stiffness  is  hysteretic  then  the  mid-­‐segment  curvature  cannot  be  derived  in  this  way  (because   of   possible   hysteresis   effects)   so   the   mid-­‐segment   curvature   reported   is   the  mean   of   the   curvatures   at   the   ends   of   the   segment.   This   difference   may   be   significant   if   the   bend   stiffness   is   significantly   non-­‐linear  over  the  range  of  curvatures  involved.  

For  further  details  see  non-­‐linear  bend  stiffness  theory.   The  External  results  data  is  still  under  development  and  is  not  available  for  general  use.   Externally  Calculated  Bend  Moment  

This  option  allows  the  bend  moment  to  be  calculated  by  an  external  function.  If  this  option  is  used  then  the  line  must   include  torsion   effects.  The   external  function  can  be  written  by  the  user  or  other  software  writers.  For  details  see   the  OrcaFlex  programming  interface  (OrcFxAPI)  and  the  OrcFxAPI  documentation.   Warning:  

Non-­‐linear   behaviour   breaks   the   assumptions   of   the  stress   results   and   fatigue   analysis   in   OrcaFlex.   You  should  therefore  not  use  these  facilities  when  there  are  significant  non-­‐linear  effects.  

Axial  Stiffness   The   axial   stiffness   is   the   slope   of   the   curve   relating   wall   tension   to   strain.   The   data   specifies   the   behaviour   in   the   unpressured  state,  i.e.  atmospheric  pressure  inside  and  out.   Pressure  effects,  including  the  Poisson  ratio  effect,  are   then  allowed  for  by  OrcaFlex.   You  can  either  specify  linear  or  non-­‐linear  behaviour,  as  follows:   x

For  a  simple  linear   behaviour,  specify  the  axial  stiffness  to  be  the  constant  slope   of  the  line   relating  wall  tension   to  strain.  This  slope  is  the  equivalent  EA  value  for  the  line,  where  E  is  Young's  modulus  and  A  is  the  cross  section   area.  It  equals  the  force  required  to  double  the  length  of  any  given  piece  of  line,  assuming  perfectly  linear  elastic   behaviour.  (In  practice,  of  course,  lines  would  yield  before  such  a  tension  was  reached.)  

x

For  a  non-­‐linear  behaviour,  use   variable  data  to  specify  a  table  of  wall  tension  against  axial  strain.  OrcaFlex  uses   linear   interpolation   for   strains   between   those   specified   in   the   table,   and   linear   extrapolation   for   strains   beyond   those  specified  in  the  table.  The  wall  tension  is  allowed  to  be  non-­‐zero  at  zero  strain.  

290  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Note:  

Axial   strain   is   defined   to   equal   (L   -­‐   L0)   /   L0,   where   L   and   L0   are   the   stretched   and   unstretched   length   of   a   given   piece   of   pipe.   Here   'unstretched'   means   the   length   when   unpressured   and   unstressed.   When   a   pipe   is   pressured   its   tension   at   this   'unstretched'   length   is   often   not   zero   because   of   strains   due   to   pressure   effects.   For   a   homogeneous   pipe   this   can   be   modelled   by   specifying  the   Poisson  ratio.  For  a  non-­‐homogeneous  pipe  (e.g.  a  flexible)  the  Poisson  ratio  may  not   be  able  to  capture  the  pressure  effects.  

Warning:  

Non-­‐linear  behaviour  breaks  the  assumptions  of  the  stress  results  and  fatigue  analysis.  

See  Calculating  Tension  Forces  for  details  of  the  tension  model  used.   Note:  

If   you   use   explicit   integration   for   dynamic   simulation   then   the   axial   stiffness   specified   here   will   have  a  major  effect  on  how  long  the  dynamic  simulation  will  take.  This  is  because  very  large  axial   stiffness  values  lead  to  very  small  natural  periods  for  the  nodes,  and  this  in  turn  requires  very  small   simulation  time  steps  when  using  explicit  integration.  See  Inner  and  Outer  Time  Steps.  

 

Fortunately,   the   value   of   axial   stiffness   used   is   often   not   very   important,   providing   it   is   large   enough  that  the  axial  strains  produced  are  small.  The  exception  to  this  is  where  snatch  loads  occur,   since  the  axial  stiffness  directly  affects  the  peak  tension  that  results.  It  is  therefore  normally  quite   acceptable  to  specify  a  much  smaller  axial  stiffness  value  than  applies  to  the  real  line,  so  enabling   much   faster   simulations.   We   recommend   that   artificially   low   axial   stiffness   values   are   specified,   particularly  for   early   investigative  simulations.  The  effect   of  this  can  easily  be   investigated  later  by   re-­‐running  a  selection  of  important  simulations  with  the  actual  axial  stiffness  value.  

Poisson  Ratio   This  specifies  the   Poisson  ratio  of  the   material  that  makes  up  the   wall  of  the  line  type.  It  is  used  to  model  any  length   changes  due  to  the  radial  and  circumferential  stresses  caused  by   contents  pressure  and  external  pressure.   A  Poisson  ratio  of  zero  means  no  such  length  changes.  For  metals  such  as  steel  or  titanium  the  Poisson  ratio  is  about   0.3  and  for  polyethylene  about  0.4.  Most  materials  have  Poisson  ratio  between  0.0  and  0.5.   Note:  

The   Poisson   ratio   effect   is   calculated   assuming   that   the   line   type   is   a   pipe   made   from   a   homogeneous   material.   It   is   not   really   applicable   to   complex   structures   such   as   flexibles,   whose   length   changes   due   to   pressure   are   more   complex.   However   for   such   cases   an   effective   Poisson   ratio  could  be  specified  as  an  approximation.  

Torsional  Stiffness   The  torsional  stiffness  is  used  only  if  torsion  is   included  on  the  line  data  form.   The   torsional   stiffness   specifies   the   relationship   between   twist   and   torsional   moment   (torque).   You   can   specify   linear  or  non-­‐linear  behaviour,  as  follows:   x

For  a  simple  linear  behaviour,  specify  the  torsional  stiffness  to  be  the  constant  slope  of  the  torsional  moment-­‐ twist  per  unit  length  relationship.  This  slope  is  the  equivalent  GJ  value  for  the  line,  where  G  is  the  shear  modulus   and   J   is  the   polar   moment  of   inertia.   It  equals   the   torque   which  arises   if   the   line   is   given   a  twist   of  1   radian   per   unit  length.  

x

For   a   non-­‐linear   behaviour,  use  variable   data  to   specify   a   table   of   torque   against   twist   per   unit   length.   OrcaFlex   uses  linear   interpolation  for   curvatures  between  those  specified  in  the  table,  and  linear   extrapolation  for  values   outside  those  specified  in  the  table.  The  torque  must  be  zero  at  zero  twist.  

See  Calculating  Torsion  Moments  for  details  on  how  this  data  is  used.   Note:  

When   defining   non-­‐linear   torsional   stiffness   you   should   specify   values   for   both   positive   and   negative  twist  per  unit  length.  This  allows  you,  for  example,  to  have  different  stiffnesses  for  positive   and   negative  twisting.  If  the   behaviour   is  mirrored   for  positive  and   negative  twist  then  you  must   specify  the  full  relationship  Ȃ  OrcaFlex  does  not  automatically  reflect  the  data  for  you.  

Warning:  

Non-­‐linear  behaviour  breaks  the  assumptions  of  the  stress  results  and  fatigue  analysis.  

291  

w

 

System  Modelling:  Data  and  Results,  Lines    

Drag,  Lift  &  Added  Mass  Data   Drag  Coefficients   The  drag  coefficients  for  the  normal  (x  and  y)  directions  and  axial  (z)  direction  are  specified  on  the  line  type  data   form.  For  the  x  and  y  directions  the  drag  coefficients  can  be  one  of  the  following  options:   x

A  fixed  constant  value.  

x

A  value  that  varies  with  Reynolds  number.  

x

A  value  that  varies  with  Height  above  Seabed.  

x

A  value  that  varies  with  both  Reynolds  number  and  Height  above  Seabed.  

Often  the  coefficients  for  the  x  and  y-­‐directions  are  equal  and  this  can  be  specified  by  setting  the  y -­‐coefficient  to  "~",   which  means  "same  as  x-­‐coefficient".  OrcaFlex  also  offers  a  choice  (on  the  line  data  form)  of  different  formulations   for  how  the  drag  force  components  vary  with  the  incidence  angle.   If  wake   interference  effects  are   being   modelled   for   a   given   line,   then   these   normal   drag   coefficients   on   the   line   type   data   form   specify   the   undisturbed   drag   coefficient.   And   the   drag   force   will   be   calculated   using   a   drag   coefficient   that  is  modified  from  this  value  according  to  the  wake  model  used.   For  further  details  see  the  Line  Theory  section.   Typical  drag  coefficient  values  

For  circular  cylinders,  the  drag  coefficient  for  normal  flow  depends  on  Reynolds  number  Re  and  surface  finish.  For   values   of   Re   between   2E4   and   3E5   the   drag   coefficient   takes   the   value   1.2   and   is  independent   of   surface   roughness.   Values   below   this   range   are   unlikely   to   occur   in   practice.   For   Reynolds   numbers   greater   than   3E5,   the   drag   coefficient  is  strongly  dependent  on  both  Re  and  surface  roughness.   For   very   smooth   cylinders   the   drag   coefficient   falls   rapidly   to   0.28   at   a   Re   of  about   6E5   before   recovering   to  a   value   of  0.5  for  Re  values  above  2E6.  For  rough  cylinders  the  effect  is  less  marked,  but  remains  significant.   In  view  of  this  behaviour,  the  use  of   variable  data  for  normal  drag  coefficients  is  strongly  recommended.  OrcaFlex   will   then   use   the   value   of   drag   appropriate   to   the   instantaneous   local   value   of   Reynolds   number   throughout   the   simulation.   The   functional   form   of   the   dependence   is   well   documented   in   the   open   literature   and   also   in   proprietary   data   sources,  such  as  ESDU  80025.   Note   that   some   of   these   sources   take   account   of   the   effect   of   the   amount   of   turbulence   in   the   incoming   flow   by   defining  an  effective  Reynolds  number.  Consequently,  care  is  needed  to   ensure  that  the  data  are  presented  in  a  form   that  is  consistent  with  the  definition  of  Reynolds  number  used  by  OrcaFlex.   Results  for  Reynolds  number  and  drag  coefficient  are  available  as  time  history  results.   The   above   values   apply   where   vortex-­‐induced   vibration   (VIV)   is   expected   to   be   negligible.   If   significant   VIV   is   anticipated,   then   drag   coefficients   may   be   increased   significantly.   If   this   is   the   case,   a   more   detailed   VIV   analysis   should  be  carried  out.   Axial   drag   results   from   skin   friction   only.   In   subcritical   flow   (Re   <   3.8E5),   the   drag   coefficient   for   axial   flow,   C t,   is   0.008  for  a  smooth  cylinder  and  0.011  for  a  rough  cylinder,  based  on  ESDU  data.  At  higher  Re,  ESDU  suggest  that   skin  friction  may  be  neglected,  i.e.  Ct  =  0.  In  practice,  axial  drag  is  often  negligible  and  Ct  =  0  is  often  acceptable.   Lift  Coefficient   The  lift  coefficient  is  used  to  specify  a  lift   force   which  acts  in  the   direction   normal  to   the  line  axis  and  in  the  plane  of   that  axis  and  the  seabed  normal.  It  can  be  one  of  the  following  options:   x

A  fixed  constant  value.  

x

A  value  that  varies  with  Reynolds  number.  

x

A  value  that  varies  with  Height  above  seabed.  

x

A  value  that  varies  with  both  Reynolds  number  and  Height  above  seabed.  

For  further  details  see  the  Line  Theory  section.  

292  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Drag  /  Lift  Diameters   These  are  used  when  calculating  drag  area  and  lift  area.  A  value  of  '~'  results  in  the  outer  diameter  being  used.   Note:  

OrcaFlex  calculates  the   normal  drag  /  lift  area  to   be  D nƒ†–Š‡ƒš‹ƒŽ†”ƒ‰ƒ”‡ƒ–‘„‡Ɏ aL   where   Dn  is  the  normal  drag  /  lift  diameter,  Da  is  the  axial  drag  diameter  and  L  is  the  element  length.  Note   that  different  programs  handle  these  calculations  in  different  ways.  For  example  it  is  common  for   programs   to   use   a   single   drag   /   lift   area   for   both   normal   and   axial   flow   and   in   addition   some   pr‘‰”ƒ• †‘ ‘– ‹…Ž—†‡ –Š‡ ˆƒ…–‘” ‘ˆ Ɏ ‹ –Š‡ ƒš‹ƒŽ †”ƒ‰ ƒ”‡ƒǤ ‡…ƒ—•‡ ‘ˆ –Š‡•‡ †‹ˆˆ‡”‡…‡• ›‘— must  be  careful  when  comparing  or  transferring  data  between  different  programs.  

Added  Mass  Coefficients   The  added  mass  coefficients  Ca  for  normal  (x  and  y-­‐directions)  and  axial  (z-­‐direction)  flow.  Often  the  coefficients  for   the  x  and  y-­‐directions  are   equal  and  this  can  be  specified  by  setting  the  y -­‐coefficient  to  "~"  which  means  "same  as  x-­‐ coefficient".   For  each  flow  direction,  the  inertia  coefficient,  Cm,  is  automatically  set  to  equal  1+Ca.  See  Added  Mass  for  details.  

Contact  Data   Contact  Diameter   Contact  between  lines  and  the  seabed,  elastic  solids  or  other  lines  accounts  for  the  diameter  of  the  line.  This  value   specifies  the  diameter  to  be  used.  If  a  value  of  '~'  is  specified  then  the  line   outer  diameter  is  used.   This  value  is  also  used  in  the  calculation  of  Line  Contact  Clearance  results  and  as  the  drawing  diameter  for  shaded   graphics  views.   Contact  Stiffness  and  Damping   The  stiffness  and  damping  values  used  by  the  clashing  algorithm.  See  Line  Clashing.   Damping  is  always  zero  when  using  the  implicit  integration  scheme.  

Stress  Data   Stress  Outer  and  Inner  Diameter   The   stress   diameters   are   the   inside   and   outside   diameters   of   the   load-­‐bearing   cylinder.   They   are   used   in   the   wall   tension  and  stress  results  calculations,  which  are  based  on  the  assumption  that  the  loads  in  the  line  are  taken  by  a   simple   homogeneous   cylinder.   For   simple   cases,   the   stress   diameters   can   be   set   to   '~',   in   which   case   they   will   be   taken  to  be  the  same  as  the  pipe  diameters.  For  more  complex  cases,  for  example  where  the  pipe  outside  diameter   allows  for  added  buoyancy  modules  that  are  not  load  bearing,  the  stress  diameters  can  be  set  separately.  See   Line   Results  Ȃ  Forces.   Allowable  Stress   The  maximum  allowable  stress  for  this  type  of  line.  This  value  is   only  used  to  draw  a  limit  curve  on  Stress  Range   Graphs;   it   does   not   limit   the   stress   achieved   in   the   line.   If   no   limit   curve   is   wanted   then   you   may   input   the   tilde   character  "~"  (meaning  not  applicable)  instead  of  a  number.   Stress  Loading  Factors   These   are   used   to   specify   what   proportion   of   the   loads   (tension,   bend   moment,   shear   and   torque)   are   to   be   used   when   calculating   wall   tension   and   stress   results.   The   effective   tension,   bend   moment,   shear   force   and   torque   are   multiplied   by   the   appropriate   stress   loading   factor   when   they   are   used   to   calculate   the   wall   tension   and   stress   results.   For   many   cases,   e.g.   when   modelling   a   simple   homogeneous   pipe   that   carries   all   the   loads,   these   load   factors   should   be  set  to  1,  the  default  value.   In  some  cases,  values  less  than  1  may  be  suitable.  For  example,  consider  a  case  where  the  line  models  a  composite   structure  that  consists   of  a   main  carrier   pipe  and  an  external  piggyback  pipe.   You  might  estimate  that  the  main   pipe   takes   all   of  the   tensile   and   torsional   loads,   but  only   carries   70%   of   the   bending   loads,   the   other   30%   being   taken   by   the   piggyback   pipe.   Then   to   obtain   stress   estimates   for   the   main   pipe   you   could   set   the   Stress   Outer   and   Inner   Diameters  to  '~'  and  set  the  bending  and  shear  stress  loading  factors  to  0.7.  

293  

w

 

System  Modelling:  Data  and  Results,  Lines    

Note:  

The  Allowable  Stress  and  Stress  Loading  Factors  only  affect  the  wall  tension  results,  stress  results   and   fatigue   analyses.   These   results   are   derived   after   the   simulation   has   run,   and   because   of   this   OrcaFlex  allows  these  data  items  to  be  modified  after  a  simulation  has  been  run.  

Friction  Data   Seabed  Friction  Coefficients   OrcaFlex   applies   Coulomb   friction   between   the   line   and   the   seab‡†Ǥ Š‡ ˆ”‹…–‹‘ ˆ‘”…‡ ƒ’’Ž‹‡† ‡˜‡” ‡š…‡‡†• Ɋ ™Š‡”‡‹•–Š‡•‡ƒ„‡†”‡ƒ…–‹‘ˆ‘”…‡ƒ†Ɋ‹•–Š‡ˆ”‹…–‹‘…‘‡ˆˆ‹…‹‡–Ǥ   Lines   lying   on   the   seabed   often   move   axially   more   readily   than   they   move   laterally.   To   enable   this   effect   to   be   modelled,   you   can   specif› †‹ˆˆ‡”‡– ˆ”‹…–‹‘ …‘‡ˆˆ‹…‹‡–• Ɋ ˆ‘” ‘–‹‘ ‘”ƒŽ ȋ‹Ǥ‡Ǥ Žƒ–‡”ƒŽȌ ƒ† ƒš‹ƒŽ –‘ –Š‡ Ž‹‡Ǥ ‘” ‹–‡”‡†‹ƒ–‡†‹”‡…–‹‘•‘ˆ‘–‹‘”…ƒ Ž‡š‹–‡”’‘Žƒ–‡•„‡–™‡‡–Š‡•‡–™‘˜ƒŽ—‡•–‘‘„–ƒ‹–Š‡ˆ”‹…–‹‘…‘‡ˆˆ‹…‹‡–Ɋ to  use.  If  the  axial  friction  coefficient   ‹••‡––‘̵̵̱–Š‡–Š‡‘”ƒŽˆ”‹…–‹‘…‘‡ˆˆ‹…‹‡–‹•—•‡†ˆ‘”Ɋˆ‘”ƒŽŽ†‹”‡…–‹‘•‘ˆ motion.  This  provides  a  convenient  way  of  using  the  same  friction  coefficient  for  all  directions  of  motion.   See  Friction  Theory  for  further  details  of  the  friction  model  used.   Note:  

The  friction  coefficient  for  contact  with  elastic  solids  is  specified  on  the   Solid  Friction  Coefficients   data  form.  

Typical  values  

Published  data  are  sparse.  Some  information  is  given  in  Puech  (1984)  and  Taylor  and  Valent(1984).  Both  references   distinguish   between   sliding   friction   and   starting   friction:   starting   friction   is   greater   to   represent   the   "breakout"   force.   OrcaFlex   does   not   draw   this   distinction.   In   most  cases,   the   sliding   friction   coefficient   should   be   used;   this   will   usually  be  conservative.  Both  references  are  written   in  the  context  of  the  contribution  of  chains  and  cables  to  anchor   holding   power,   so  we   assume   the   friction   values   given   are   axial.   Transverse   values  will   be   greater,   perhaps  by  50%   to  100%.   The  values  given  below  are  recommendations  from  Taylor  and  Valent.   Line  type   Seabed  Type  

Starting  Friction   Sliding  Friction   Coefficient   Coefficient  

Chain  

Sand  

0.98  

0.74  

 

Mud  with  sand  

0.92  

0.69  

 

Mud/clay  

0.90  

0.56  

Wire  rope   Sand  

0.98  

0.25  

 

Mud  with  sand  

0.69  

0.23  

 

Mud/clay  

0.45  

0.18  

Structural  Damping  Data   Rayleigh  Damping  Coefficients   A   named   Rayleigh   Damping   Coefficient   data   set.   This   data   item   can   be   set   to   "(no   damping)",   in   which   case   no   Rayleigh  damping  will  be  applied  for  this  Line  Type.   This  data  is  only  available  when  using  the  implicit  integration  scheme.  

Code  Checks  Data   The  code  check  data  can  be  found  by  setting  the  Line  Type  view  mode  to  Code  Check.   API  RP  2RD   Data  used  for  the  API  RP  2RD  stress  calculation.   Design  case  factor  Cf  

See  API  RP  2RD,  section  4.4,  table  2  and  section  5.2.3.1.   This  value  is  a  property  of  the  entire  model.  To  modify  it  using  batch  script  you  must  first  select  the  General  object.  

294  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Minimum  wall  thickness  

The  minimum  wall  thickness,  denoted  tmin  in  API  RP  2RD,  section  5.2.4.4.  A  value  of  '~'  results  in  the  nominal  wall   thickness,  (OD-­‐ID)/2,  being  used.   SMYS  

The  material  minimum  yield  strength  (SMYS),  denote†ɐy  in  API  RP  2RD,  section  5.2.3.1.  

Drawing  Data   Pen   Defines  the  colour,  line  style  and  thickness  of  the  pen  used  for  drawing  this  line  type.  See   How  Objects  Are  Drawn.   For  each  line  there  is  a  choice,  on  the   Line  Data  form,  of  whether  to  draw  the  sections  of  the  line  using  these  Line   Types  pens,  or  whether  to  define  a  specific  pen  to  use  for  all  the  sections  of  the  line.  

External  Functions   Parameters   This  data  item  specifies  the   External  Function  Parameters,  a  free   form  multi-­‐line  text   field   which  is  passed  to   any   external  function  used  by  the  Line  Type.  

Plasticity  Wizard   OrcaFlex  allows  you  to  specify   non-­‐linear   bend  stiffness  for  Line  Types.  The  data   is  given  as  a  table  of   bend   moment   against  curvature.  If  you  are  modelling  a  uniform,  homogeneous  pipe  you  can  use  the   Plasticity  Wizard  to  create   this  table.   Once  you  have  entered  the  appropriate  data  as  described  below,  click  the  Calculate  button  and  the  curvature  /  bend   moment   relationship   is   generated.   The   Bend   Stiffness   variable   data   source   that   is   created   is   initialised   to   be   hysteretic.   The  Plasticity  Wizard  is  opened  by  clicking  the  "Plasticity  Wizard"  button  on  the  Variable  Data  form.   Note:  

Before   you   can   open   the   Plasticity   Wizard   you   must   have   created   and   selected   a   Bend   Stiffness   variable  data  source.  

Plasticity  Wizard  Data   The  Plasticity  Wizard  requires  the  following  data:   Stress  Diameters  

The  inside  and  outside  diameters  of  the  load-­‐bearing  cylinder.   Direct  Tensile  Strain  

The   Plasticity   Wizard   calculates   bend   moment   curvature   relationship   by   integrating   the   stress   profile   across   the   pipe   cross-­‐section.   This   calculation   requires   a   direct   tensile   strain   to   be   specified   Ȃ   this   data   item   serves   that   purpose.   Stress-­‐Strain  relationship  

The   relationship   between   stress   and   strain   can   be   specified   by   either   Ramberg-­‐Osgood   curve   or   Stress-­‐Strain   table.   Material  E,  RP

y),  K,  n  (Ramberg-­‐Osgood  curve  only)  

Š‡•‡†ƒ–ƒ†‡ˆ‹‡–Š‡”‡Žƒ–‹‘•Š‹’„‡–™‡‡•–”‡••ȋɐȌƒ†•–”ƒ‹ȋɂȌ‹–‡”•‘ˆƒƒ„‡”‰-­‐Osgood  curve  as  follows:   ɂȋɐȌαɐȀΪȋɐȀɐy)n.   Š‡”‡ˆ‡”‡…‡•–”‡••’ƒ”ƒ‡–‡”‹•—•—ƒŽŽ›–ƒ‡–‘„‡–Š‡›‹‡Ž†•–”‡••ǡ™Š‹…Š‹•™Š›‹–‹•†‡‘–‡†ɐ y  here.  Note  that   there  is  an  alternative  parameterisation  of  the  Ramberg-­‐Osgood  equation.  It  is  straightforward  to  convert  between   the   two   forms   of   the   equation   but   please   take   care   to   ensure   that   the   data   you   input   correspond   to   the   parameterisation  used  by  OrcaFlex.  

295  

w

 

System  Modelling:  Data  and  Results,  Lines    

Stress,  Strain  (Stress-­‐Strain  table  only)  

This   table   directly   specifies   the   relationship   between   stress   and   strain.   The   table   is   interpolated   linearly   and   for   values  of  strain  outside  the  table  linear  extrapolation  will  be  used.   Maximum  curvature  for  derived  data  source   The  Bend  Stiffness  variable   data  source   is  defined  for  curvature  values  between  0  and  C max   αɂmax/Ro  where  Ro  is   the   radius  to  the  outer  fibre.  The  outer  fibre  strain  corresponding  to  Cmax  ‹•†‡‘–‡†ɂmax  and  is  defined  as  follows:   x

For   a   Ramberg-­‐•‰‘‘† …—”˜‡ ɂmax   =   max{0.05,   ͷɂȋɐyȌȔǤ ‘ –Š‡ ˜ƒŽ—‡ —•‡† ˆ‘” ɂmax   will   be   5   times   the   strain   corresponding  to  the  reference  stress  or  5%,  whichever  is  larger.  

x

For  a  Stress-­‐–”ƒ‹–ƒ„Ž‡ɂmax  is  simply  the  largest  value  of  strain  specified  in  the  table.  

Properties  Report   The  Line  Type  properties  report  is  available  from  the  popup-­‐menu  on  the  data  form.  It  reports  the  following:   Weight  in  air  

The  force  due  to  gravity  acting  on  the  line  type's  mass.   Displacement  

The   weight   of   water   displaced   by   the   line   type's   volume.   The   reported   value   uses   the   water   density   at   the   sea   surface.   Weight  in  water  

Equals  Weight  in  air  -­‐  Displacement.   Diam/Wt  Ratio  

Equals  Outer  Diameter  /  Weight  in  water.   Note:  

For  Line  Types  that  have  a  non-­‐zero  bore  you  must  specify  the  contents  density  to  be  used  in  the   calculation  of  the  above  properties,  since  this  will  affect  the  properties  that  involve  weight.  

Used  in,  Contents  Density  

The  names  and  contents  densities  of  each  line  that  uses  that  line  type.  

6.8.3

Attachments  

Attachment  Types   The  Attachment  Types  form   defines  the   properties  of  a  number  of  named  attachment  types.  Attachments  with  these   properties   can   then   be   connected   to   lines.   Attachment   Types   can   be   either   Clump   Types,   Drag   Chain   Types,   Flex   Joint  Types  or  Stiffener  Types.   The  attachment  types  form  must  include  all  the  attachment  types  referred  to  on  all  of  the  Lines  data  forms,  but  it  can   also   include   other   attachment   types   that   are   not   currently  in   use   in  the  model.   This  allows   you   to   build   up   a   library   of  standard  attachment  types  that  can  then  be  easily  used  when  building  Lines.  

Clumps   A  clump  is  a  concentrated  attachment  that  is  connected  to  a  node  on  a  Line.  It  can  be  buoyant  or  heavy  and  is  a  small   body   that   experiences   forces   (weight,   buoyancy,   drag   etc.)   exactly   as   for   a   3D   Buoy.   But   instead   of   being   free   to   move   it   is   constrained   to   move   with   the   node   and   the   forces   acting   on   it   are   transferred   to   that   node.   A   clump   therefore  adds  to  the  mass,  buoyancy  and  hydrodynamic  force  of  the  node  to  which  it  is  attached.   Clumps  only  have  3  degrees  of  freedom  Ȃ  X,Y  and  Z  Ȃ  which  are  determined  by  the  position  of  the  node  to  which  they   are   attached.   Clumps   can   be   aligned   with   the   global   axes   directions   or   alternatively   they   can   be   aligned   with   the   node  to  which  they  are  attached.   Each  clump  is  assigned  a  height  and  an   offset  from  the   node  which  are  used  to  determine  the  Z  coordinate  of   the   clump  for  the  purposes  of  evaluating  buoyancy  and  hydrodynamic  forces:  no  moment  is  applied  to  the  node  by  the   clump.   Where   the   clump   pierces   the   water   surface,   buoyancy   and   hydrodynamic   forces   are   applied   in   proportion   to   the  immersed  length  of  the  clump.  

296  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Each  clump   is  of   a   named   clump   type,   from  which  it   inherits   all   its  properties.   The   clump   types   are   specified   on  the   Attachment  Types  form  and  have  the  following  data.   Clump  Type  Name  

Used  to  refer  to  the  Clump  Type.   Mass  

Mass  or  weight  in  air.   Volume  

Used   to  calculate   buoyancy  and   added   mass   for   each   clump   of  this   type   on   a   line.   Clumps   may   be   either   net   buoyant   or  heavy  as  desired.   Height  

Used  for  drawing  the  clump  and  also  to  determine  how  much  of  the  clump  is  below  the  water  surface.   If   the   clump   is   aligned   with   global   axes   then   it   is   centred   at   the  Offset   position   above   the   node,   and   extends   for   half   its  Height  above  and  below  this  point.   If  the  clump  is  aligned   with   Line  axes  then  it  is  centred  at  the  node,  and  extends  for  half  its  Height  either  side  of  this   point  in  the  node's  axial  direction.   Offset  

A  clump  may  be  offset  vertically  from  the  line,  for   example  to  represent  a  line  supported  below  the  surface  by  floats.   The   connection   is   not   modelled   fully:   the   clump   is   always   treated   as   being   at   the   specified   offset   vertically   above   (offset  positive)  or  below  (offset  negative)  the  node  to  which  it  is  attached.   If  the  clump  is  aligned  with  Line  axes  then  the  clump  offset  is  forced  to  be  zero.   Align  with  

Determines  whether  the  clump  is  aligned  with  Global  Axes  or  Line  Axes.   This  setting  determines  the   clump's  local  directions.  If  it   is  aligned  with  global  axes  then  the  clump's  local  directions   are  the  same  as  the  global  axis  system.  If  it  is  aligned  with  Line  axes  then  its  local  directions  are  the  same  as  the  node   to  which  it  is  connected.   Drag  

Drag  forces  are  calculated  in  clump  local  directions  for  each  clump  on  a  line.   drag  force  =  PW  .  ½  .  Water  Density  .  (velocity)2  .  Cd  .  Drag  Area   where   Cd  is  Drag  Coefficient  as  specified  here,   Drag  Area  is  specified  here,   velocity  is  the  velocity  of  the  fluid  relative  to  the  clump  in  the  appropriate  direction.   Added  Mass  Coefficients  

Added  mass  in  clump  local  directions  is  given  by   Added  mass  =  PW  .  Ca  .  Water  Density  .  Volume   where   Ca  is  the  Added  Mass  Coefficient  as  specified  here.   Pen  

Defines  the  colour,  line  style  and  thickness  of  the  pen  used  for  drawing  this  clump  type.  See   How  Objects  Are  Drawn.  

Clump  Type  Properties  Report   The  Clump  Types  properties  report  is  available  from  the   popup-­‐menu  on  the  data  form.  It  reports  the  following:  

297  

w

 

System  Modelling:  Data  and  Results,  Lines    

Weight  in  air  

The  force  due  to  gravity  acting  on  the  clump's  mass.   Displacement  

The  weight  of  water  displaced  by  the  clump's  volume.  The  reported  value  uses  the  water  density  at  the  sea  surface.   Weight  in  water  

Equals  Weight  in  air  -­‐  Displacement.  

Drag  Chains   Drag  chains  are  attachments  to  a  line  that  model  straight  chains  that  hang  down  from  the  line.  They  apply  weight,   buoyancy  and  drag  forces  to  the   node  to   which  they  are  attached,   but   not  any  added   mass  effects.  For  details  see   Drag  Chain  Theory.   Drag   chains   include   two   facilities   that   can   be   important   in   modelling   towed   systems.   Firstly,   the   chain's   drag   coefficients   can   vary   with   the   incidence   angle   of   the   relative   flow;   this   enables   modelling   the   effect   that   as   the   relative   flow   increases   the   chain   hangs   at   a   greater   angle   to   the   vertical   and   so   fluid   drag   generates   more   lift,   which   is  applied  to   the   line.  Secondly,  drag   chains   interact   with  the   seabed   (in   a   simple   manner);   if  the   node   comes   closer   to  the  seabed  than  the  chain  length,  then  the  seabed  provides  a  supporting  reaction  force  and  a  friction  force,  both  of   which  are  applied  to  the  node.   Each   drag   chain   is   of   a   named   drag   chain   type,   from   which   it   inherits   all   its   properties.   The   drag   chain   types   are   specified  on  the  Attachment  Types  form  and  have  the  following  data.   Name  

Used  to  refer  to  the  Drag  Chain  Type.   Length  

Length  of  the  drag  chain.   Effective  Diameter  

Effective  diameter  of  the  drag  chain.  This  is  the  diameter  of  the  cylinder  that  has  the  same  displaced  mass  per  unit   length.   Mass  

Mass  per  unit  length.  Mass  is  assumed  to  be  uniformly  distributed  along  the  length  of  the   drag  chain.   Friction  Coefficient  

Coefficient   of   friction   for   contact   with  the   seabed.   This  coefficient   is  used   for   all   directions   of   friction.   The   value   can   be  set  to  '~',  in  which  case  the  drag  chain   will  instead  use  the  axial   friction  coefficient  of  the   node  to  which  the   drag   chain  is  attached.   Drawing  

Defines  the  colour,  line  style  and  thickness  of  the  pen  used  for  drawing  drag  chains  of  this  type.  See  How  Objects  Are   Drawn.   Drag  Coefficients  

The   fluid   drag   forces   on   the   chain   are   specified   by   giving   a   table   of   the   normal   and   axial   drag   coefficients,   as   a   ˆ—…–‹‘ ‘ˆ –Š‡ ‹…‹†‡…‡ ƒ‰Ž‡ Ƚ „‡–™‡‡ –Š‡ ”‡Žƒ–‹˜‡ ˜‡Ž‘…‹–› ˜‡…–‘” ƒ† –Š‡ †”ƒ‰ …Šƒ‹Ǥ ‘ Ƚ  =  0°   means   flow   ƒš‹ƒŽŽ›ƒŽ‘‰–Š‡†”ƒ‰…Šƒ‹ƒ†Ƚ  =  90°  means  flow  normal  to  the  drag  chain.   Coefficients   are   specified   for   a   range   of   incidence   angles   between   0°   and   90°   and   linear   interpolation   is   used   to   obtain   coefficients   for   intermediate  angles.   The   Graph   button   shows   the   resulting   coefficient   variation.   Symmetry   is   used  to  obtain  coefficients  for  angles  outside  the  range  0°  to  90°.   Note:  

‘ „‡ ”‡ƒŽ‹•–‹…ǡ –Š‡ ‘”ƒŽ †”ƒ‰ ˆ‘”…‡ •Š‘—Ž† ‹…”‡ƒ•‡ ‘‘–‘‹…ƒŽŽ› ƒ• –Š‡ ‹…‹†‡…‡ ƒ‰Ž‡ Ƚ increases  from  0  to  90.  This  turns  out  to  require  that  the  gradient  of  the  normal  drag  coefficient   …—”˜‡ †ȋȽȌ •Š‘—Ž† „‡ ‰”‡ƒ–‡” –Šƒ -­‐͸Ǥ†ȋȽȌ Ȁ –ƒȋȽȌ ˆ‘” ƒŽŽ ȽǤ ”…ƒ Ž‡š ™ƒ”• ‹ˆ –Š‡ †”ƒ‰ coefficient  data  does  not  satisfy  this.  

See  Drag  Chain  Theory  for  further  details.  

298  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Drag  Chain  Type  Properties  Report   The  Drag  Chain  Types  properties  report  is  available  from  the  popup-­‐menu  on  the  data  form.  It  reports  the  following:   Length,  Diameter  

The  data  as  specified  on  the  data  form.   Total  Mass  

The  total  mass  of  the  drag  chain  given  by  Length  ×  Mass  per  unit  Length.   Total  weight  in  air  

Equals  Total  Mass  ×  g.   Total  displacement  

Š‡™‡‹‰Š–‘ˆ–Š‡˜‘Ž—‡‘ˆˆŽ—‹††‹•’Žƒ…‡†„›–Š‡†”ƒ‰…Šƒ‹ǤŠ‡˜‘Ž—‡‹•‰‹˜‡„›‡‰–ŠέɎ 2/4  where  D  is  the   effective  diameter  as  specified  on  the  data  form.   Total  wet  weight  

Equals  Total  weight  in  air  -­‐  Total  displacement.  

Flex  Joints   Flex   joints   are   attachments   to   a   line   that   add   additional   bending   stiffness   to   a   particular   node   (specified   by   the   attachment's   z-­‐position).   This   additional   local   stiffness   acts   in   series   with   the   stiffness   of   the   underlying   line   type   bend  stiffness.  If,  for  example,  you  set  the  stiffness  for  the  flex  joint  to  zero,  then  you   effectively  make  that   node  pin-­‐ jointed.   Each   flex   joint   has   properties   determined   by   a   named   Flex   Joint   Type.   The   Flex   Joint   Types   are   specified   on   the   Attachment  Types  form  and  have  the  following  data:   Name  

Used  to  refer  to  the  Flex  Joint  Type.   Bend  Stiffness  

Specifies   the   bend   stiffness   of   the   Flex   Joint   Type.   You   can   specify   separate   values   for   bending   about   the   x   and   y-­‐ directions.  A  y-­‐bend  stiffness  value  of  '~'  is  interpreted  as  'same  as  x-­‐bend  stiffness'.   Similarly  to  line  end  connection  stiffness  the  bend  stiffness  is  specified  in  terms  of  moment  per  degree  of  deflection.   Drawing  

Defines   the   colour,   line   style   and   thickness   of   the  pen   used   for   drawing   flex  joints   of  this   type.  See   How  Objects   Are   Drawn.  

Stiffeners   Stiffeners   are   attachments   to   a   line   intended   for   use   modelling   bend   stiffeners.   Each   stiffener   has   properties   determined   by  a  named   Stiffener  Type.   The   Stiffener  Types  are  specified  on  the   Attachment   Types  form  and  have   the  following  data:   Name  

Used  to  refer  to  the  Stiffener  Type.   Line  Type  

Determines  the  physical  properties  of  the  stiffener.  Usually  this  will  be  a   profiled  homogeneous  pipe.   Because  the  stiffener  properties  are  included  in  addition  to  the  protected  line  properties  this  line  type  should  just   specify  the  properties  of  the  stiffener.   Length  

Defines   the  length   of  the  stiffener.  If  the  line  type  is  profiled  then  the  length  is   determined   by  the  profile   data  and  so   cannot  be  edited  here.  

299  

w

 

System  Modelling:  Data  and  Results,  Lines    

Connection  Arc  Length,  Relative  to  

Defines  the  point  on  the  stiffener  which  corresponds  to  the   protected  line  attachment  position.  This  correspondence   is  illustrated  in  the  following  table  which  assumes  a  stiffener  length  of  10:   Protected  line   Stiffener   Protected  line   attachment   connection   protected  arc  length   arc  length   arc  length   range   0  

0  

0-­‐10  

5  

0  

5-­‐15  

10  

5  

5-­‐15  

50  

10  

40-­‐50  

For  a  bend  stiffener  at  End   A  of  a  line  the  connection  arc  length  would  set  to  0  relative  to  End  A.  Similarly  for  a  bend   stiffener  at  End  B  of  a  line  the  connection  arc  length  would  set  to  0  relative  to  End  B.   Axial  load/inertia  transfer  

Specifies  how  the  axial  loads  and  axial  inertia  are  transferred  from  the  stiffener  to  the  protected  line.   If  at  connection  point  is  specified  then  the  axial  loads  and  axial  inertia  for  the  entire  stiffener  line  are  transferred  to   the   protected   line   at   the   connection   point.   This   option   is   suitable   when   you   wish   to   neglect   the   axial   effects   of   friction  due  to  contact  between  protected  line  and  stiffener.   If  over  full  length  is  specified  then   the  axial  loads  and  axial  inertia  are  transferred  and  shared  across  the  full  length   of  the  stiffener.  If  the  contact  friction  means  that  the  protected  line  effectively  carries  the  axial  loads  and  inertia  of   the  stiffener  then  this  option  should  be  used.   For  a  stiffener  which  is  connected  to  the  protected  line  by  means  of  a  full  length  friction  grip,  e.g.  a  Cumberland  Grip,   then  you  should  use  the  over  the  full  length  option  for  axial  load/inertia  transfer.   More  technical  details  of  how  this  data  item  is  used  are  given  in  Modelling  Bend  Stiffeners.  

6.8.4

Rayleigh  Damping  

Rayleigh   damping   is   used   to   model   structural   damping   for   Lines   and   is   only   available   when   using   the   implicit   integration  scheme.   Classical  Rayleigh  damping  

Classical  Rayleigh  damping  uses  a  system  damping  matrix  C  defined  as:   αɊΪɉ   where   Ɋ‹•–Š‡ƒ••’”‘’‘”–‹‘ƒŽ  Rayleigh  damping  coefficient.   ɉ‹•–Š‡•–‹ˆˆ‡••’”‘’‘”–‹‘ƒŽƒ›Ž‡‹‰Š†ƒ’‹‰…‘‡ˆˆ‹…‹‡–Ǥ   M  is  the  system  structural  mass  matrix.   K  is  the  system  structural  stiffness  matrix.   With  this  formulation  the  damping  ratio  is  the  same  for  axial,  bending  and  torsional  response.   Classical  Rayleigh   damping   results   in   different   damping   ratios   for   different   response   frequencies   according   to   the   following  equation:   ɌαͲǤͷȋɊȀɘΪɉɘȌ   where   Ɍ‹•–Š‡†ƒ’‹‰”ƒ–‹‘ȋƒ˜ƒŽ—‡‘ˆͳ…‘””‡•’‘†•–‘…”‹–‹…ƒŽ†ƒ’‹‰ȌǤ   ɘ‹•–Š‡”‡•’‘•‡ˆ”‡“—‡…›‹”ƒ†Ȁ•Ǥ   It   can   be   seen   from   this   that   the   mass   proportional   term   gives   damping   ratio   inversely   proportional   to   response   frequency  and  the  stiffness  proportional  term  gives  damping  ratio  linearly  proportional  to  response  frequency.   Separated  Rayleigh  damping  

In  addition  to  classical  Rayleigh  damping  OrcaFlex  offers  a  separated  Rayleigh  damping  model  where  C  is  defined  as:  

300  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

αɊΪɉaKa  ΪɉbKb  ΪɉtKt   where   ɉa  is  the  stiffness  proportional  Rayleigh  damping  coefficient  for  axial  deformation.   ɉb  is  the  stiffness  proportional  Rayleigh  damping  coefficient  for  bending  deformation.   ɉt  is  the  stiffness  proportional  Rayleigh  damping  coefficient  for  torsional  deformation.   Ka  is  the  system  structural  stiffness  matrix  corresponding  to  axial  deformation.   Kb  is  the  system  structural  stiffness  matrix  corresponding  to  bending  deformation.   Kt  is  the  system  structural  stiffness  matrix  corresponding  to  torsional  deformation.   Note  that  K  =  Ka  +  Kb  +  Kt.   The  separated  model  allows  for  different  damping  ratio  in  axial,  bending  and  torsional  response.   Data   Each   Line  Type  has  structural  damping  properties  determined  by  a   named  Rayleigh  Damping  Coefficient  data  set.   Multiple  Rayleigh  Damping  Coefficient  data  sets  can  be  defined,  each  with  the  following  data.   Rayleigh  damping  mode  

OrcaFlex  offers  four  distinct  methods  for  specifying  the  Rayleigh  damping  coefficients:   x

Stiffness  proportional.  

x

Mass  and  stiffness  proportional.  

x

Coefficients  (classical).  

x

Coefficients  (separated).  

Stiffness  proportional  and  Mass  and  stiffness  proportional  methods  

These  methods  allow  you  to  specify  structural  damping  in  terms  of  %  critical  damping  ratio.  In  addition  you  need   to  specify  response  periods  at  which  the  damping  ratio  will  be  achieved.  For  Stiffness  proportional  damping  only   one  response  period  can  be  specified.   If   Response   Period   1   is   set   to   '~'   then   OrcaFlex   will   choose   response   periods   automatically   based   on   the   wave   period,  or  peak  period   Tp   for  a   random   wave.  This   method  is   generally  recommended  and  is  particularly  useful  if   you  are  running  a  batch  of  cases  with  varying  wave  conditions.   ”…ƒ Ž‡š—•‡•–Š‡…Žƒ••‹…ƒŽƒ›Ž‡‹‰Š†ƒ’‹‰‘†‡Žƒ†”‡’‘”–•–Š‡ƒ••ƒ†•–‹ˆˆ‡••’”‘’‘”–‹‘ƒŽ…‘‡ˆˆ‹…‹‡–•Ɋ ƒ†ɉǤ   Coefficients  (classical)  and  Coefficients  (separated)  methods  

For  these  methods  you  specify  the  damping  coefficients  directly.   Damping  ratio  graph  

The   damping   ratio   graph   plots   the   damping   ratio   that   will   be   achieved   for   a   range   of   response   frequencies.   The   graph's  x-­‐axis  can  be  either  period  of  frequency.  

Guidance   Variation  of  damping  ratio  

Rayleigh   damping   is   viscous   damping   that   is   proportional   to   a   linear   combination   of   mass   and   stiffness.   The   †ƒ’‹‰ƒ–”‹š‹•‰‹˜‡„›αɊΪɉ™Š‡”‡ǡƒ”‡–Š‡ƒ••ƒ†•–‹ˆˆ‡••ƒ–”‹…‡•”‡•’‡…–‹˜‡Ž›ƒ†Ɋǡɉƒ”‡ constants  of  proportionality.   Rayleigh   damping   does   afford   certain   mathematical   conveniences   and   is   widely   used   to   model   internal   structural   damping.   One   of   the   less   attractive   features   of   Rayleigh   damping   is   that   the   achieved   damping   ratio   varies   as   response   frequency   varies.   The   stiffness   proportional   term   contributes   damping   that   is   linearly   proportional   to   response   frequency   and   the   mass   proportional   term   contributes   damping   that   is   inversely   proportional   to  response   frequency.  Mathematically,  these  frequency  d‡’‡†‡…‹‡•…ƒ„‡•‡‡‹–Š‡ˆ‘”—Žƒˆ‘”†ƒ’‹‰”ƒ–‹‘ɌαɎȋɊȀˆΪ ɉˆȌ™Š‡”‡ˆ‹•–Š‡”‡•’‘•‡ˆ”‡“—‡…›Ǥ  

301  

w

 

System  Modelling:  Data  and  Results,  Lines    

The   plot   below   illustrates   how   the   separate   mass   and   stiffness   damping   terms   contribute   to   the   overall   damping   ratio:   Mass and Stiffness

Mass term

Stiffness term

6%

Damping ratio

5% 4% 3% 2% 1% 0% 0

1

2

3

4

5

Response angular frequency (rad/s) ȝ Ȝ 

  Figure:  

Variation  Of  Damping  Ratio  With  Frequency  

Artificial  over-­‐damping  of  system  response  

Consider   a   system   which   has   two   primary   responses:   one   at   the   wave   frequency   and   the   other   at   a   much   lower   frequency,  for  example  due  to  vessel  drift.  Clearly  Rayleigh  damping  constants  must  be  chosen  carefully  to  avoid  the   mass  proportional  term  resulting  in  over-­‐damping  of  the  low  frequency  response.   It  is  common  practice  to  do  this  by  using  the  stiffness  proportional  term  only.  For  example  the  DNV  dynamic  riser   code  DNV-­‐OS-­‐F201  (Appendix  A,  K103)  makes  the  following  recommendation:  "It  should  also  be  observed  that  the   mass   proportional   damping   would   give   damping   due   to   rigid   body   motions.   The   mass   proportional   damping   is   therefore   normally   neglected   for   compliant   structures   undergoing   large   rigid   body   motions."   In   other   words   the   recommendation  for  such  systems  is  to  use  stiffness  proportional  damping  (the  red  curve  above).   This  is  a  good  argument.  However,  an  analogous  argument  can  also  be  made  about  any  high  frequency  response.  If   only   stiffness   proportional   damping   is   used   then   any   high   frequency   response   will   be   over-­‐damped.   It   is   quite   common   for   systems   to   have   responses   at   frequencies   higher   than   the   wave   frequency.   Since   high   frequency   responses   are   often   damaging   to   a   system   it   is  important   to  model   them   accurately.   Stiffness   proportional   damping   is   very   effective   at   artificially   removing   high   frequency   responses   from   an   analysis   and   this   is   a   danger   that   must   be   avoided.   Avoiding  over-­‐damping  of  system  response  

We  recommend  the  following  procedure  for  applying  Rayleigh  damping:   1.

Identify  a  number  of  critical  load  cases.  

2.

Run  these  cases  without  Rayleigh  damping.  

3.

Identify  the  minimum  and  maximum  frequencies  present  in  the  system  response,  f min  and  fmax.  

4.

Apply  mass  and  stiffness  proportional  Rayleigh  damping  with  response  periods  set  to  1/fmin  and  1/fmax.  

This   technique   will   apply   the   specified   damping   ratio   at   responses   with   frequency   f min   and   fmax.   For   frequencies   between   fmin   and   fmax   the   damping   ratio   will   be   less   than   the   specified   damping   ratio   which   ensures   that   artificial   over-­‐damping  is  avoided.   Note:  

Since   this   approach  leads   to  a   damping   ratio   less   than   the   specified  damping   ratio   for   frequencies   between   fmin   and   fmax   it   could   be   argued   that   the   response   will   be   under-­‐damped.   Generally   this   conservative  under-­‐damping  is  far  less  significant  than  the  non-­‐conservative  over-­‐damping  that  we   are  trying  to  avoid.  

302  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

If   your   system   response   is   concentrated   at   the   wave   frequency   for   all   critical   design   cases   then   it   is   safe   to   apply   stiffness  proportional  damping.   Discussion  

The   above   procedure   will   help   you   avoid   non-­‐conservative   over-­‐damping   of   high   or   low   frequency   responses   in   your  system.  However,  it  is  quite  laborious  and  time-­‐consuming.  Can  we  adopt  a  simpler  approach?   For  a  great  many  of  the  systems  that  are  analysed  by  OrcaFlex  it  turns  out  that  structural  damping  has  little  or  no   effect  on  system  performance.  For  subsea  lines  the  structural  damping  is  usually  negligible  in  comparison  with  the   damping   due   to  hydrodynamic   drag.   For   this   reason   Rayleigh   damping   can   usually  be   ignored   for   analysis   of  subsea   lines.  For  systems  which  show  significantly  different  responses  w hen  Rayleigh  damping  is  applied  you  should  check   that  this  is  not  due  to  the  over-­‐damping  issues  described  above.   For   in-­‐air   lines   (e.g.   jumper   hoses)   the   situation   is  different.   These   lines   have   no   hydrodynamic   damping   and   so   the   structural  damping  can   be  significant.  For  such  lines  it  is  very  easy  for  resonant   responses  to  be  excited  and  if   no   damping   is   modelled   then   these   responses   do   not   decay.   In   this   situation   Rayleigh   damping   can   be   very   useful.   It   is,   of  course,  important  to  apply  it  carefully  as  described  above.   One   common   phenomenon   that   is   sometimes   modelled   with   Rayleigh   damping   is   the   damping   due   to   internal   friction   between   layers   of   a  pressurised,   unbonded   flexible   riser.   This  effect   is  strongly  amplitude   dependent   and   is   poorly   represented   by   Rayleigh   damping.   The   problem   is   that   the   damping   ratio   depends   on   the   amplitude   of   response.   This   makes   the   damping   ratio   very   difficult   to   set   and   it   will   differ   for   different   load   cases.   Instead   we   recommend  that  you  use  a  hysteretic  bend  stiffness  which  gives  a  more  accurate  model  of  the  riser.  

6.8.5

Line  Results  

This   section   describes   the   line   results   that   are   available   for   the   static   and   dynamic   analyses.   These   results   are   available  using  the  Results  Selection  form.   Results  from  the  modal  analysis  and  fatigue  analysis  are  described  elsewhere  Ȃ  see  the  Modal  Analysis  and  Fatigue   Analysis  sections.   Selecting  which  Categories  of  Line  Results  are  Shown   For   Lines   there   are   a   large   number   of  results   variables   available   on   the  Results   form.   So  OrcaFlex   groups   the   results   variables  into  the  following  categories:   x

Positions.  

x

Motions.  

x

Angles.  

x

Forces.  

x

Moments.  

x

Contact.  

x

Pipe  Stress  /  Strain.  

x

End  Loads.  

To   ease   results   selection   the   Show   boxes   on   the   results   form   allow   you   to   choose   which   of   these   categories   of   variables  are  shown  in  the   Variable  list.  To  get  the  full  list  of  available  variables  simply  select  all  the  categories.   But   normally   there   are   several   categories   of   variable   that   you   do   not   currently   need,   in   which   case   de-­‐selecting   them   reduces  the  displayed  list  of  variables  to  a  more  manageable  set.   Specifying  the  Position  on  the  Line   For  line  results  you  need  to  specify  the  position  on  the  line  at  which  you  want  results.  This  is  done  by  setting  the   entries  in  a  row  in  the  Position  table  on  the  results  form.  You  are  then  offered  the  Variables  that  are  available  for   the  point  specified  by  the  currently-­‐selected  row.   Each  row  in  the  table  specifies  one  point  on  the  line.  There  are  multiple  rows  in  the  table,  so  you  can  set  up  rows   specifying   a   number   of   different   points   of   interest   and   then   easily   switch   between   them   by   choosing   which   row   you   select.  In  a  row  that  you  don't  want  to  use  you  can  set  the  Node  or  Arc  Length  column  to  '~',  meaning  'unspecified'.   Three  rows  in  the  table  are  dedicated  to  special  arc  lengths  on  the  line:  

303  

w

 

System  Modelling:  Data  and  Results,  Lines    

x

The  first  and  last  rows  in  the  Position  table  are  dedicated  to  the  line's  end  points  A  and  B.  

x

The  next  to  last  row  in  the  table  is  dedicated  to  the   Touchdown  point.  This  is  defined  to  be  the  first  node  on  the   seabed   (starting   from  the   Top  End).   If   the   results   variable   selected   is   a   segment   variable   (i.e.   is   only   available   at   mid-­‐segment  points)  then  the  value  reported  for  the  touchdown  point  is  the  mid-­‐segment  valued  in  the  segment   that  precedes  the  Touchdown  node.  When  there  are  no  nodes  on  the  seabed  then  the  results  variable  is  reported   as  N/A  (meaning  'not  available')  and  the  graph  shows  no  value.  

Arc  Length  and  Node  Columns  

The  Arc   Length   column   specifies   how  far   along   the   line   the   point   is,   measured   from   zero   at   End   A.   For   information,   if   you   set   the   Arc   Length   column   then   the   adjacent   Node   cell   is   set   to   the   number   of   the   nearest   node   to   that   arc   length.   The  Node  column  can  also  be  used  as  an  alternative  way  of  setting  the  arc  length.  You  can  set  the  Node  column  to   the  number  of  a  node  on  the  line.  The  adjacent  Arc  Length  cell  will  then  be  set  to  the  arc  length  to  that  node.  The   node   number  must  be  in  the  range  1  (the  node  at  End  A)  to  N+1  (the  node  at  End  B),   where  N  is  the  total  number  o f   segments  in  the  line.   Note:  

The   actual   arc   length   for   which   line   results   are   reported   may   not   be   exactly   the   specified   arc   length.  OrcaFlex  reports  results  for  the  'nearest  appropriate'  result  point.  See   Result  Points  below.  

R  and  Theta  Columns  

For   some   variables   (e.g.  stress   components)   you   must   also   specify   the   position   of   the   point   within   the   cross   section   through   the   specified   arc   length.   Whenever   one   of   these   variables   is   selected   in   the   Variables   list,   two   extra   columns  become  visible  in  the   Position  table.  These  extra  columns  specify  the  polar  coordinates  (R,Theta)  of  the   point  within  the  cross  section;  see  the  diagram  in  the  Pipe  Stress  Calculation  section.  The  R  column  can  only  be  set   to  either  Inner  or  Outer,  meaning  the   radii  corresponding  to  the   Stress  ID  or  Stress  OD  respectively.  Results  are  not   available  for  points  between  these  two  radii.   Clearance  Results   Clearance   results   can   be   reported   either   as   clearances   from   this   line   to   all   other   lines   or   from   this   line   to   a   specified   other   line.   You   choose   which   of   the   options   is   used   from   the   drop-­‐down   list   labelled   "Clearances   are   reported   as   clearances  from".   Result  Points   OrcaFlex   uses   a  discretised  model  and   so  results   are  only  available   at   nodes,   mid-­‐segment  points   and   line  ends;  we   call   these   points   'result   points'.   The   available   result   points   depend   on   which   variable   you   request,   they   are   documented  in  the  description  of  the  variable.   When   you   ask   for   a   variable   at   a   specified   arc   length   OrcaFlex   gives   the   value   for   the   'nearest   appropriate'   result   point.  The  phrase  'nearest  appropriate'  here  means  that  OrcaFlex  considers  the  available  result  points  that  are  in  the   same  section  as  the  arc  length  you  specified  and  then  chooses  the   one  that  is   nearest  to  the  arc  length  you  specified.   If   you   specify   an   arc   length   that   is   exactly   at   the   boundary   of  two   sections   then   OrcaFlex   uses   the   section   that   starts   at  that  arc  length.   OrcaFlex   always   labels   results   with   the   actual   arc   length   to   the   result   point   to   which   they   apply,   so   you   can   check   to   ensure  that  you  are  getting  results  at  the  result  point  you  want.  

Positions   X,  Y  and  Z  

Available  at  nodes.  The  global  coordinates  of  the  selected  node.   Layback  

Available  at  nodes.  The  horizontal  component  of  distance  between  the  selected  node  and  the   touchdown  point.   Layback   is   not  defined   (N/A)   if  there   is   no   touchdown   point,   that   is   if   there   are   no   nodes   on   the   seabed.   Likewise,   if   the  selected  node  is  after  the  touchdown  point,  then  Layback  is  not  defined.  

304  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Proportion  Wet  

Available   at   nodes.   The   proportion   of  the   part   of  the   line   that   the   node   represents,   that   is   submerged   in   the   sea.   The   value  is  in  the  range  0  to  1,  a  value  of  0  meaning  no  submersion  and  1  meaning  is  completely  submerged.  For  details   see  Line  Interaction  with  the  Sea  Surface.   Sea  Surface  Z  

Available   at   nodes.   The   global   Z   coordinate   of   the   sea   surface   directly   above   the   instantaneous   position   of   the   selected  node.   Depth  

Available  at  nodes.  The  depth  of  the  node  beneath  the  sea  surface  (=  Surface  Z   -­‐  Node  Z).   Sea  Surface  Clearance  

Available  at  nodes.  The  vertical   clearance  from  node  centre  to  the  instantaneous  sea  surface.  Negative  values  mean   that  the  node  is  submerged.   Arc  Length  

Available  at  nodes.  The  arc  length  from  End  A  to  the  selected  point.  This  is  normally  only  useful  for  the   touchdown   point,  since  for  other   points  it  is  constant.  For  the  touchdown  point  it  gives  the  arc  length  from  End  A  to  the  first   node  on  the  seabed,  or  zero  if  there  is  no  touchdown.   Expansion  Factor  

Available  at  mid-­‐segments.  The  expansion  factor  of  the  selected  segment.   Transverse  VIV  Offset  

Available  at  nodes.  This  is  the  node's  offset  from  the   non-­‐VIV  node  position,  in  the  transverse  VIV  direction.   Estimated  Transverse  A/D  

Available   only   at   nodes   that   use   one   of   the   time   domain   VIV   models   from   the   VIV   Toolbox.   The   amplitude   of   the   oscillation  in  the  transverse  VIV  direction  divided  by  the  node's  VIV  diameter.  

Motions   Velocity,  GX-­‐Velocity,  GY-­‐Velocity,  GZ-­‐Velocity,   Acceleration,  GX-­‐Acceleration,  GY-­‐Acceleration,  GZ-­‐Acceleration  

Available   at   nodes.  The  magnitude   and   components   (with  respect  to   global   axes)   of   the   velocity   and   acceleration   of   the  node.   Warning:  

The  velocity  results  are  derived  by  numerically  differentiating  the  logged  positions  of  the  node  with   respect   to   time,   using   the   central   difference   scheme.   The   acceleration   results   are   derived   by   a   further   such   numerical   differentiation.   Because   of   this   the   accuracy   of   the   results   (especially   the   accelerations)   will   depend  on  the   log  sample   interval.  If   the  log  sample   interval  is  large  then  the   results  will  not  show  higher  frequency  components  of  velocity  and  acceleration.  If  the  log  sample   interval  is  small  then  the  results  may  be  inaccurate  due  to  loss  of  precision  due  to  subtraction.  

Acceleration  (incl.  g),  x-­‐Acceleration  (incl.  g),  y-­‐Acceleration  (incl.  g),  z-­‐Acceleration  (incl.  g)  

Available   at   nodes.   The   magnitude   and   components   (with   respect   to   node   axes)   of   the   vector   a   -­‐   g   where   a   is   the   acceleration  of  the  node  and  g  is  the  acceleration  due  to  gravity,  a  vector  pointing  vertically  downwards.   These  results  can  be  used  to  compare  against  accelerometer  readings.   Relative  Velocity,  Normal  Relative  Velocity,  Axial  Relative  Velocity  

Available   at   nodes.   Relative   Velocity   is   the   velocity   of   the   fluid   relative   to   the   node,   i.e.  Vfluid  -­‐  Vnode.   The   results   reported  are  the  magnitude  of  the  relative  velocity  and  its  normal  and  axial  components  (relative  to  the  line).  For   the  axial  component,  a  positive  value  means  that  the  fluid  is  moving  (relative  to  the  line)  towards  End  B.   The   fluid   velocity   used   is   the   velocity   of   the   principal   fluid   affecting   the   node.   The   principal   fluid   is   defined   as   follows:   1.

If   proportion   dry  >   0.5   and   the  Include   wind   loads  on   Lines  option   is  enabled   in   the  Environment  data   then   the   principal  fluid  is  the  air.  

305  

w

 

System  Modelling:  Data  and  Results,  Lines    

2.

Otherwise  the  principal  fluid  is  the  sea.   Note:  

For  a  node  that  is  above  the  water  surface  OrcaFlex  reports  a  relative  velocity  based  on  the  fluid   velocity  at  the  surface  point  vertically  below  the  node.  

Warning:  

The  relative  velocity  results  are  derived  using  the  node  velocity  results,  so  see  the  accuracy  warning   given  above.  

Strouhal  Frequency  

Available   at   nodes.   The   Strouhal   Frequency   is   defined   to   be   St.V/D   where   St=0.2,   V   is   the   normal   component   of   relative  velocity  and  D  is  the  normal  drag  diameter.   Reynolds  Number  

Available   at   nodes.   The   Reynolds   number   is   a   measure   of   the   flow   regime.   OrcaFlex   offers   a   number   of   different   options  for  the  calculation  of  Reynolds  number,  specified  on  the  Environment  data  form.   x-­‐Drag  Coefficient,  y-­‐Drag  Coefficient,  z-­‐Drag  Coefficient,  Lift  Coefficient  

Available  at  nodes.  These  are  the  drag  and  lift  coefficients  used  in  the  calculation.   For   constant   coefficients   then   these   results   report   the   values   given   in   the   user's   data,   except   for   a   node   at   the   junction  between  two  sections  with  different  coefficients,  where  an  effective  average  value  is  used.   If  the  line's  drag  or  lift  coefficients  vary  with  Reynolds  number  or  Height  above  Seabed  then  these  results  report  the   computed  value  that  was  used.   If  the  line  uses  a  wake  oscillator  VIV  model  with  inline  drag  amplification  then  the  amplification  factor  is  included  in   these  results.  The  inline  drag  amplification  factor  is  also  available  as  a  separate  result.   Wake  Velocity  Reduction  Factor,  Wake  Cd,  Wake  Cl  

Available  at  nodes  only,  for  lines  which  include  sections  that  react  to  wake  effects.   Wake  Velocity  Reduction  Factor  is  the   factor  applied  to   the  velocity  at  the  node  as  a  result  of  upstream   wake  effects.   Wake  Cd  and  Wake  Cl  are  the  drag  and  lift  coefficients  respectively,  used  to  calculate  the  hydrodynamic  forces  at  the   node  as  a  result  of  any  upstream  wake  effects.   Note:  

Wake   Cl   is   positive   when   the   lift  force   is   applied   in   the   y   direction   of   the   upstream   wake's  frame   of   reference   and   negative   when   the   lift   force   is   applied   in   the   -­‐y   direction   of   the   upstream   wake's   frame  of  reference.  

Angles   Azimuth,  Declination  and  Gamma  

Available  at  mid-­‐segment  points  and   line   ends.  These  angles  report  the  local  orientation  of  the  line   relative  to  global   axes.  The  gamma  angle  is  defined  as  for  line  ends  Ȃ  see  Line  End  Orientation.   Declination   is   in   the   range   0°   to  180°.  Range   jump   suppression   is   applied   to   Azimuth   and   Gamma   (so   values   outside   the  range  -­‐360°  to  +360°  might  be  reported).   Ez-­‐Angle,  Exy-­‐Angle,  Ezx-­‐Angle,  Ezy-­‐Angle  

Available   at   mid-­‐segment   points   and   line   ends.   The   direction   angles   of   the   mid-­‐segment   point,   relative   to   the   end   axes  of  the  nearest  line  end.  See  End  Direction  Results.   Ez-­‐Angle  is  in  the  range  0°  to  180°.   Range  jump  suppression  is  applied  to  Exy-­‐Angle,  Ezx-­‐Angle  and  Ezy-­‐Angle  (so   values  outside  the  range  -­‐360°  to  +360°  might  be  reported).   Twist  

Available  at  mid-­‐segment  points.  The  twist  per  unit  length  experienced  by  the  segment.   Fluid  Incidence  Angle  

Available  at  nodes.  The  angle  between  the  relative  velocity  direction  and  the  line  axial  direction.  A  value  in  the  range   0°  to  90°.  

306  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

No-­‐Moment  Azimuth,  No-­‐Moment  Declination  

Available  at  line   ends.  The  azimuth  and  declination  angles,   relative  to  global  axes,  of   the   no-­‐moment   direction  at   the   end,  allowing  for  any  motion  of  the  object  to  which  the  line  is  attached.  These  results  are  only  available  if  the  end   orientation  angles  are  defined.   No-­‐Moment  Declination  is  in  the   range  0°  to  180°.   Range  jump  suppression  is  applied  to  No-­‐Moment  Azimuth   (so   values  outside  the  range  -­‐360°  to  +360°  might  be  reported).   End  Force  Azimuth,  End  Force  Declination  

Available  at  line  ends.  The  azimuth  and  declination  of  the  end  force  vector,  relative  to  global  axes.   End  Force  Declination  is  in  the  range  0°  to  180°.   Range  jump  suppression  is  applied  to  End  Force  Azimuth  (so  values   outside  the  range  -­‐360°  to  +360°  might  be  reported).   End  Force  Ez-­‐Angle,  End  Force  Exy-­‐Angle,  End  Force  Ezx-­‐Angle,  End  Force  Ezy-­‐Angle  

Available   at   line   ends.   The   direction   angles   of  the   end   force   vector,   with   respect   to   the   frame   of   reference   of  the   line   end.  See  End  Direction  Results.  These  results  are  only  available  if  the  end  orientation  angles  are  defined.   End  Force  Ez-­‐Angle  is  in  the   range  0°  to  180°.   Range   jump  suppression  is  applied  to  the  other  3   end  force  angles  (so   values  outside  the  range  -­‐360°  to  +360°  might  be  reported).   VIV  Stagnation  Point,  VIV  Neg  Separation  Point,  VIV  Pos  Separation  Point  

Available  only  at  nodes  that  use  one  of  the   vortex  tracking  VIV  models  from  the  VIV  Toolbox.  The  values  reported   are  the  angular  positions  of  the  stagnation  and  separation  points.  

Forces   Effective  Tension  and  Wall  Tension  

Available   at   mid-­‐segment   points   and   line   ends.   The   structural   force   along   the   line   axis.   Positive   values   denote   tension  and  negative  values  denote  compression.   The   reported   wall   tension   is   the   total   wall   tension,   T w,   scaled   by   the   tensile   stress   loading   factor.   By   default   this   loading  factor  equals  1.   For   details   of  the   difference   between   the   effective   tension,   T e,   and   the   wall   tension,   Tw,   see   the   Line   Pressure   Effects   section.   In   particular   see   the   warning   in  that   section   if  the   line   type   stress   diameters   differ   from   the   outer   and   inner   diameters.   Normalised  Tension  

Defined  to  be  Effective   Tension  divided   by   Allowable   Tension.  Available  at   mid-­‐segment  points  and   line   ends;   not   available  if  the  Allowable  Tension  is  set  to  '~'.   Contents  Density  

Available  at  nodes.  This  result  is  most  useful  when  the  free-­‐flooding  or  slug  flow  contents  methods  are  in  use.   Shear  Force,  x-­‐Shear  Force,  y-­‐Shear  Force,  Shear  Force  component,  In-­‐plane  Shear  Force,  Out-­‐of-­‐plane  Shear  Force  

Available   at   mid-­‐segment   points   and   line   ends.   The   magnitude   of   structural   force   normal   to   the   line   axis,   its   components   in   the   local   x   and   y-­‐directions,   its   component   in   the   user   specified   local   direction   theta   and   its   components  in  the  in-­‐plane  and  out-­‐of-­‐plane  directions.   The  in-­‐plane  direction  is  normal  to  both  the  line's   axis  (at  the  specified  arc  length)  and  the  vertical  direction.   The   out-­‐of-­‐plane  direction  is  normal  to  both  the  line's  axis  and  the  in-­‐plane  direction.   If  the  line's  axis  is  vertical  then  these  directions  are  ill-­‐defined  and  therefore  no  values  can  be  reported.   Vortex  Force  Magnitude,   Inline  Vortex  Force,  Transverse  Vortex  Force,   GX-­‐Vortex  Force,  GY-­‐Vortex  Force,  GZ-­‐Vortex  Force  

Available   only   at   nodes   that   use   one   of  the  time   domain   VIV   models   from   the   VIV   Toolbox.   The   magnitude   of  the   lift   and   drag   force   per   unit   length   of   line,   and   its   components   in   the   VIV   directions   and   global   axes   directions.   For   details,  see  the  documentation  of  the  relevant  time  domain  VIV  model.  

307  

w

 

System  Modelling:  Data  and  Results,  Lines    

Inline  Drag  Amplification  Factor  

Available   only   at   nodes   that   use   one   of   the   wake   oscillator   VIV   models   from   the   VIV   Toolbox.   The   instantaneous   value  of  the  Inline  Drag  Amplification  Factor.  

Moments   Bend  Moment,  x-­‐Bend  Moment,  y-­‐Bend  Moment,  Bend  Moment  component,  In-­‐plane  Bend  Moment,  Out-­‐of-­‐plane  Bend   Moment  

Available  at  mid-­‐segment  points  and   line   ends.  The   magnitude  of  bend  moment,  its  components  in  the  local  x  and   y-­‐ directions,  its  component  in  the  user  specified  local  direction   theta  and  its  components  in  the  in-­‐plane  and  out-­‐of-­‐ plane  directions.   The  in-­‐plane  direction  is  normal  to  both  the  line's  axis  (at  the  specified  arc  length)  and  the  vertical  direction.   The   out-­‐of-­‐plane  direction  is  normal  to  both  the  line's  axis  and  the  in-­‐plane  direction.   If  the  line's  axis  is  vertical  then  these  directions  are  ill-­‐defined  and  therefore  no  values  can  be  reported.   Curvature,  x-­‐Curvature,  y-­‐Curvature,  Curvature  component,  In-­‐plane  Curvature,  Out-­‐of-­‐plane  Curvature  

Available   at   mid-­‐segment   points   and   line   ends.   The   magnitude   of   curvature,   its   components   in   the   local   x   and   y-­‐ directions,  its  component  in  the  user  specified  local  direction   theta,  the  in-­‐plane  and  out-­‐of  plane-­‐components  and   its  components  in  the  in-­‐plane  and  out-­‐of-­‐plane  directions,  as  defined  above.   When  pre-­‐bend  is  modelled  curvature  results  are  reported  relative  to  the  pre-­‐bent  curvature.   Note:  

When  using  non-­‐linear  bend  stiffness,  the  reported  mid-­‐segment  curvature  depends  on  whether  the   bend   stiffness   is   specified   to   be   hysteretic   or   not.   For   details   see   the   note   in   the   Non-­‐linear   Bend   Stiffness  section.  

Warning:  

Curvature   results   are   accurate   only   if   the   segment   length  is   sufficiently   short.   T he   accuracy   can   be   estimated  by  performing  a  sensitivity  study  on  segment  length  in  the  area  of  interest.  

Normalised  Curvature  

Available  at  mid-­‐segment  points  and  line  ends.  Defined  to  be  Curvature  divided  by  Allowable  Curvature.   If   minimum   bend   radius   (MBR)   is   specified   as   varying   with   wall   tension   then   this   variation   is   taken   into   account   when  calculating  Normalised  Curvature.   Bend  Radius,  x-­‐Bend  Radius,  y-­‐Bend  Radius,  Bend  Radius  component,  In-­‐plane  Bend  Radius,  Out-­‐of-­‐plane  Bend  Radius  

Available  at  mid-­‐segment  points  and  line  ends.  The  magnitude  of  bend  radius,  its  components  in  the  local  x  and  y-­‐ directions,  its  component  in  the  user  specified  local  direction   theta  and  its  components  in  the  in-­‐plane  and  out-­‐of-­‐ plane  directions,  as  defined  above.   When  pre-­‐bend  is  modelled  these  results  are  reported  relative  to  the  pre-­‐bent  curvature.   Notes:  

Bend  radius  is  defined  to  be  1  /  curvature.  If  the  curvature  is  0  then  a  value  of  'Infinity'  is  reported.  

 

When  using  non-­‐linear  bend  stiffness,  the  reported  mid-­‐segment  curvature  depends  on  whether  the   bend   stiffness   is   specified   to   be   hysteretic   or   not.   For   details   see   the   note   in   the   Non-­‐linear   Bend   Stiffness  section.  

Warning:  

Bend  radius  results  are  accurate  only  if  the  segment  length  is  sufficiently  short.  The  accuracy  can   be  estimated  by  performing  a  sensitivity  study  on  segment  length  in  the  area  of  interest.  

Torque  

Available   at   mid-­‐segment   points   and   line   ends   only,   and   available   only   for   lines   with   torsion   included.   The   component  of  structural  moment  along  the  line  axis.  

Contact   Note:  

As  well  as  the  results  variables  documented  below  OrcaFlex  also  provides  a  Line  Clashing  Report.  

308  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Line  Centreline  Clearance,  Line  Contact  Clearance  

Available  at  mid-­‐segment  points  and  line   ends.   The  line  clearance  variables  report  shortest  distances   between  lines.   These   distances   can   be   either   the   shortest   distance   between   centrelines   (Line   Centreline   Clearance)   or  the   shortest   distance   between   outer   edges   (Line   Contact   Clearance).   The   results   selection   form   lets   you   choose   to   report   clearances  either:   x

from  this  line  to  all  other  lines  or  

x

from  this  line  to  a  specified  other  line.  

In  the  text  below  we  refer  to  the  other  lines  used  in  the  clearance  calculations  as  the  clearance  lines,  be  they  all  other   lines  or  a  single  specified  other  line.   Line  Centreline  Clearance  reports  the  centreline  clearance  from  this  line  to  the  clearance  lines.  More  precisely,  the   clearance   reported  for  a  segment  is  the  shortest  distance  from  the  centreline   of  the   segment  to   the  centreline  of  any   segment  on  the  clearance  lines.  Note  that  the  clearance  reported  therefore  does  not  allow  for  the  radii  of  the  lines   involved.  When  clearance  is  reported  for  a  line  end  it  reports  the  shortest  distance  from  the  centreline  of  the  point   at  the  end  node  to  the  centreline  of  any  segment  on  the  clearance  lines.   Line  Contact  Clearance  is  similar  but  it  reports  the  clearance  between  line  outer  edges  allowing  for  their   contact   diameters.   Note:  

Line  Contact  Clearance  can  report  negative  values.  This  means  that  the  segments  in  question  are   penetrating   each   other.   If   clashing   is   being   modelled   for   both   segments   then   a   clash   force   will   result  from  this  penetration.  

The   line   clearance   variables   are   useful   for   checking   for   clashing   between   lines.   They   are   available   in   both   range   graph  and   time  history  form.  The  range  graph,  for  a  given  period  of  the  simulation,  enables  you   to  see  where  on   the   line  clashing  may   be  a  problem.  You  can  then   examine  the  time  history  of  line  clearance   for  that   point  on  the  line,  to   see  when  closest  approach  occurs.  You  can  then  use  the  replay  to  examine  which  other  line  is  coming  closest.   It   is   sometimes   worth   choosing   carefully   which   line   to   check   for   clearance.   An   example   is   checking   for   clashing   between   a   single   mooring   line   and   one   or   more   of   a   number   of   closely  spaced   flowlines.   Let   us   assume   that   you   are   reporting  clearances  from  all  other  lines.  The  clearance  graphs  for  the  flowlines  will  include  clearance  to  the  other   flowlines,   between   which  clashing   may  not  be   a   concern.   The   mooring   line   clearance   is  probably  more   useful,  since   it  only  includes  clearance  to  the  flowlines.   Line   clearance   only   checks   against   other   lines,   not   against   edges   of   vessels,   buoys,   etc.   However   you   can   check   clearance   against   part   of   a   vessel,   for   example,   by   attaching   a   dummy   single-­‐segment   line   to   the   vessel,   spanning   across  the  area  of  interest.  The  line  clearance  graphs  for   that  dummy  line  will  then  show  how  close  other  lines  come   to  that  area  of  the  vessel.   Notes:  

For  mid-­‐segment  points  the  segment  used  is  the  one  containing  the  selected  arc  length.  

 

Line  clearance  results  are  only  available  if  there  are  at  least  2  lines  in  the  model.  

Warning:  

For  complex  models,  building  and  updating  clearance  graphs  can  be  slow.  Having  "live"  clearance   graphs  open  while  a  simulation  is  running  can  significantly  slow  down  the  simulation.  

Seabed  Clearance  

Available  at  nodes.  The  clearance  is  the  shortest  distance  between  the  node  and  any   point  on  the  seabed,  allowing   for   the   contact   diameter.   The   value   reported   is   for   the   node   that   is   nearest   the   specified   arc   length.   A   negative   value   indicates  that  the  node  is  in  contact  with  the  seabed.   This  result   is  not   available   for  3D  seabeds  because   it   is  difficult   to   calculate.   Instead   you   should   use   Vertical   Seabed   Clearance.   Vertical  Seabed  Clearance  

Available   at   nodes.   The  clearance   is  the   vertical   distance   between   the   node   and   the  seabed,   allowing   for   the  contact   diameter.  The  value  reported  is  for  the  node  that  is  nearest  the  specified  arc  length.  A  negative  value  indicates  that   the  node  is  in  contact  with  the  seabed.  

309  

w

 

System  Modelling:  Data  and  Results,  Lines    

Line  Clash  Force  

Available  at  mid-­‐segment  points.  The  magnitude  of  the  clash  force  between  this  segment  and  other  lines.  Please  note   that   this   variable   is  only  available   if   clash  checking   has   been   included   for   the   lines   concerned.  See  Line   Clashing  for   details.   Line  Clash  Force  is  given  for  the  segment  containing  the  selected  arc  length  and  results  are  available  in  the  form  of   time   histories   and   range   graphs.   If   multiple   clashes   occur   simultaneously   on   the   same   segment   then   the   value   reported  is  the  magnitude  of  the  vector  sum  of  the  clash  forces  involved.   Line  Clash  Impulse  

Available  at  mid-­‐segment  points.  The  integral  of  Line  Clash  Force  with  respect  to  time.   Line  Clash  Energy  

Available   at   mid-­‐segment   points.   Clash   energy   is   calculated   by   integrating   the   magnitude   of  clash   force   with   respect   to  depth  of  penetration.   Suppose   that   this   segment   is   denoted   by   S1   and   segment   S2   is   another   segment   which   S1   is   in   contact   with.   The   clash   energy  for   the   spring/damper   representing   contact   between   these   2   segments   is  calculated   by  integrating   the   magnitude   of   clash   force   with   respect   to   depth   of   penetration.   This   then   is   the   potential   energy   in   the   spring/damper.   If  multiple  clashes  occur  simultaneously  on  the  same  segment  then  the  value  reported  is  the  sum  of  all  individual   clash  energies  between  this  segment  and  other  segments.   Solid  Contact  Force  

Available  at  nodes.  The  magnitude  of  the  force  per  unit  length  due  to  contact  with  elastic  solids.   Seabed  Normal  Penetration/D  

Available  at  nodes.  The  component  of  seabed  penetration  normal  to  the  seabed,  divided  by  the  contact  diameter.   Seabed  Normal  Resistance,  Seabed  Normal  Resistance/D  

Available   at   nodes.   Seabed   Normal   Resistance   is   the   component   of   seabed   resistance   normal   to   the   seabed,   where   seabed   resistance   means   the   seabed   reaction   force   per   unit   length   of   line.   Seabed   Normal   Resistance/D   is   the   Seabed  Normal  Resistance  divided  by  the  contact  diameter.   The  forces  due  to  both  seabed  stiffness  and  seabed  damping  are  included.   Warning:  

The  damping  force  depends  upon  node  velocity.  This  is  derived  by  numerically  differentiating  the   logged  positions  of  the  node  with  respect  to  time,  using  the  central  difference  scheme.  Because  of   this  the  accuracy  of  the  results  will   depend  on  the  log  sample  interval.  See  Motions  results  for  more   details.  

Pipe  Stress  /  Strain   Stress   and   strain   results   are   available   at   mid-­‐segment   points   and   at   line   ends.   For   terminology   see   Pipe   Stress   Calculation.   The  stress  calculations  make  the  following  assumptions:   x

At   each   point   along   the   line   all   the   loads   are   taken   by   a   single   simple   cylinder   of   the   specified   Stress   OD   and   Stress  ID  and  made  of  a  homogeneous  material.  

x

The  stresses  included  are  those  due  to  tension,  bending,  shear  and  hoop  stress.  

x

The   loads   (tension,   bend   moment,   shear   and   torque)   which   are   used   in   stress   calculations   are   scaled   by   the   stress  loading  factors  before  being  used.  

x

Internal  pressure  in  the  line  generates  wall  tension  in  the  line  as  it  would  do  in  a  sealed  cylinder.  

x

Shear   stress   is   assumed   to   be   uniformly   distributed   across   the   cross   section.   Although   this   is   not   strictly   the   case,  the  shear  stress  is  normally  negligible  so  this  simplifying  assumption  is  reasonable.  

x

The   hoop   stress   due   to   static   internal   and  external   pressure   at   the   current   Z -­‐level   is   included,   and   is   calculated   using   the   standard   Lamé   equation   for   thick   walled   cylinders.   However   the   effect   of   dynamic   variations   in   pressure,  for  example  from  the  passage  of  the  wave,  are  not  included.  

310  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Limitations  of  stress  calculations  

The  assumptions  described  above  mean  that  the  stress  calculations  are  only  valid  for  pipes  such  as  steel  or  titanium   risers,  not  for  composite  flexible  risers,  ropes  chains,  etc.   If   the   pipe   has   non-­‐linear   stiffness   then   the   program   cannot,   in   general,   accurately   calculate   pipe   stresses.   The   program   uses   the   same   formulae   for   stress   calculation   as   it   does   for   linear   stiffness.   For   example   the   bending   stress   is   calculated   as   Mr/Ixy.   An   exception   to   this   is   made   for   a   homogeneous   pipe   with   non-­‐linear   stress-­‐strain.   In   this   situation  stress  results  can  be  calculated  accurately  by  using  the  stress-­‐strain  data.   If  the  line  type   stress  diameters  differ  from  the   outer  and  inner   diameters  then  see   the  warning  in  the   Line  Pressure   Effects  section.   The   program   does   not,   and   indeed   cannot,   allow   for   the   complex   stress   concentrations   that   can   occur  at   joints   or   at   the  top  and  bottom  of  a  riser.   Direct  Tensile  Strain  

Available   at   mid-­‐segment   points   and   line   ends.   This   is   the   component   of   axial   strain   due   to   wall   tension   (which   includes  the  effects  of  internal  and  external  pressure).   Max  Bending  Strain  

Available   at   mid-­‐segment   points   and   line   ends.   This   is   the   axial   strain   due   to   bending   at   the   outer   fibre   on   the   outside  of  the  bend.   Worst  ZZ  Strain  

Available   at   mid-­‐segment   points   and   line   ends.  This  equals   whichever   of   Direct   Tensile   Strain   ±   Max   Bending   Strain   has  the  larger  absolute  value.   ZZ  Strain  

Available   at   mid-­‐segment   points.   This   equals   Direct   Tensile   Strain   +   Bending   Strain.   ZZ   Strain   varies   across   the   cross-­‐section  and  so  its  value  is  reported  at  a  specified  (ǡɅ)  position.   Internal  and  External  Pressure  

Available   at   mid-­‐segment   points   and   line   ends.   The   internal   and   external   static   pressures,   P i   and   Po.   See   Line   Pressure  Effects  for  details.   Pressures  in  OrcaFlex  are  gauge  pressures,  not  absolute  pressures.  That  is,  they  are  relative  to  atmospheric  pressure   and  so  can  be  as  low  as  minus  1  atmosphere  (-­‐101.325  kPa).   Net  External  Pressure  

Available  at  mid-­‐segment  points  and  line  ends.  Defined  as  Pi  -­‐  Po.   Direct  Tensile  Stress  

Available  at  mid-­‐segment  points  and  line  ends.  This  is  the  axial  stress  due  to  wall  tension  (which  includes  the  effects   of  internal  and   external  pressure).  It   is  constant  across  the  cross-­‐section  and  equals  Tw/A.  A  positive  value   indicates   tension;  a  negative  value  indicates  compression.   Max  Bending  Stress  

Available  at  mid-­‐segment  points  and  line  ends.  This  is  the  maximum  value  that  the  Bending  Stress  takes  anywhere   in  the  section.  It  is  given  by   Max  Bending  Stress  =  (C2.M.ODstress/2)  /  Ixy   and  this  maximum  occurs  at  the  extreme  fibre  on  the  outside  of  the  bend.   For  a  homogeneous  pipe  with  non-­‐linear  stress-­‐strain   ƒš‡†‹‰–”‡••αɐȋɂzz)  -­‐  ‹”‡…–‡•‹Ž‡–”‡••αɐȋɂzz)  -­‐  Tw/A   ™Š‡”‡ɐȋȉȌ‹•–Š‡•’‡…‹ˆ‹‡†stress-­‐strain   relationship  ƒ†ɂzz  is  the   axial   strain   at   the  extreme   fibre   on   the   outside   of   the  bend  data.   Worst  Hoop  Stress  

Available  at  mid-­‐segment   points  and  line   ends.  The   Hoop  Stress  is  due  to  internal  and  external  pressure.  It  varies   across   the   section   and   can  be   positive   (tension)   or   negative   (compression),   and  by  the   Worst   Hoop   Stress   we  mean  

311  

w

 

System  Modelling:  Data  and  Results,  Lines    

the  hoop  stress  of  greatest   magnitude.  It  is   obtained   by  finding  the  point  in  the  cross-­‐section   where  the  unsigned   magnitude  of  the  Hoop  Stress  is  largest;  this  must  be  either  at  the  inside  or  outside  f ibre  of  the  stress  area.  The  Hoop   Stress  at  this  point  is  called  the  Worst  Hoop  Stress.   Max  xy-­‐Shear  Stress  

Available   at   mid-­‐•‡‰‡– ’‘‹–• ƒ† Ž‹‡ ‡†•Ǥ Š‡ ˜ƒŽ—‡ ȋɐRZ2   Ϊ ɐCZ2)½   is   called   the   xy-­‐Shear   Stress.   This   varies   across  the  cross-­‐section,  and  OrcaFlex  reports  the  maximum  value  that  occurs  anywhere  in  the  cross-­‐section.  This  is   the  Max  xy-­‐Shear  Stress  and  it  is  given  by   Max  xy-­‐Shear  Stress  =  (C4ǤɒǤstress/2)  /  Iz  +  C3.S  /  A   von  Mises  Stress,  Max  von  Mises  Stress  

Available   at   mid-­‐•‡‰‡–’‘‹–•ƒ†Ž‹‡‡†•ǤŠ‡˜‘‹•‡••–”‡••ǡɐvm,  is  a   stress   measure   that   is  often   used  as   a   yield  criterion.  It  is  a  combination  of  all  the  components  of  the  stress  matrix  and  in  terms  of  principal  stresses  it  is   given  by:   ɐvm  αȏȓȋɐ1-­‐ɐ2)2  Ϊȋɐ2-­‐ɐ3)2  Ϊȋɐ3-­‐ɐ1)2}/2]½   ™Š‡”‡ɐ1ǡɐ2  ƒ†ɐ3  are  the  principal  stresses,  i.e.  the  eigenvalues  of  the  3  by  3  stress  matrix.   The  von  Mises  Stress  varies  across  the  cross-­‐section,  so  its  value  is  reported  at  a  specified  (ǡɅ)  position.   The  Max   von   Mises  Stress   is   an   estimate   of  the   maximum   value   of   the   von   Mises   Stress   over   the   cross-­‐section.   The   way  it  is  calculated  depends  on  whether  the  line   includes  torsion  or  not,  as  follows.   x

If  torsion  is  not  included,  then  OrcaFlex  assumes  that  the  torque  is  zero.  In  this  case  the  maximum  value  of  the   von  Mises  stress  must  occur  in  the  plane  of  bending.  OrcaFlex  also   assumes  that  the  maximum  occurs  at  either   the  inner  or  outer  fibre.  (This  is  a  commonly-­‐used  assumption  that  is  almost  always  valid,  since  if  the  internal   pressure  stress  contribution  is  dominant  then  the  maximum  will  be  at  the  inner  fibre,  whereas  if  bending  stress   is  dominant  then  it  will  occur  at  the  outer  fibre.)  OrcaFlex  therefore  calculates  the  von  Mises  stress  at  4  points   (R  =  ±IDstress/2  and  ±ODstress/2,  in  the  plane  of  bending)  and  reports  the  largest  value.  

x

If  torsion  is  included,  then   the  maximum  value  of  the  von  Mises  stress  can,   in  general,  occur  anywhere  in  the   pipe   wall.   So   OrcaFlex   calculates   the   von   Mises   stress   at   a   grid   of   points   across   the   pipe   wall   and   reports   the   largest   value   found.   Curre–Ž›ǡ–Š‡‰”‹†…‘’”‹•‡•͵͸Ʌ-­‐values   (i.e.   every  10°   around   the   pipe   circumference)   at   each  of  5  R-­‐values  across  the  pipe  wall.  

API  RP  2RD  Stress,  API  RP  2RD  Utilisation  

Available  at  mid-­‐segment  points  and  line  ends.   API  RP  2RD  StressǡɐAPI,  is  a  von-­‐Mises  type  stress  defined  in  section  5.2  of  API  RP  2RD  as:   ɐAPI  αƒšȏȓȋɐpr-­‐ɐ’Ʌ)2  Ϊȋɐ’Ʌ-­‐ɐpz)2  Ϊȋɐpz-­‐ɐpr)2}/2]½   where   ɐpr  =  -­‐  (Po.ODstress  +  Pi.IDstress)  /  (ODstress  +  IDstress)   ɐ’Ʌ  =  (Pi  -­‐  Po)ODstress/2tmin  -­‐  Pi   ɐpz  =  Tw/A  ±  M(ODstress  -­‐  t)/2Ixy   tmin  is  the  minimum  wall  thickness   t  is  the  nominal  wall  thickness,  (ODstress  -­‐  IDstress)/2   Š‡ƒš‹–Š‡ˆ‘”—Žƒˆ‘”ɐAPI  accounts  for  the  ˆƒ…––Šƒ––Š‡ά•‹‰‹–Š‡ˆ‘”—Žƒˆ‘”ɐpz  ƒ‡•ɐpz  double-­‐valued.   API  RP  2RD  Utilisation,  UAPI,  is  reported  as  a  percentage  and  is  defined  to  be:   UAPI  αɐAPI  /  (CfCaɐy)   where   Cf  is  the  design  case  factor   Ca  is  2/3   ɐy  is  the  material  minimum  yield  strength  (SMYS)   The  strength  check  for  API  RP  2RD  code  is  therefore  equivalent  to  the  inequality  U API  ζͳǤ  

312  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

RR  Stress,  CC  Stress,  ZZ  Stress,  RC  Stress,  RZ  Stress,  CZ  Stress  

Available  at  mid-­‐segment  points  and  line  ends.  These  are  the  individual  stress  components  at  a  point  in  the  cross-­‐ section.  The  point  is  specified  by  its  polar  coordinates  ( ǡɅ)  within   the  cross  section.   See   Pipe  Stress  Calculation  and   Pipe  Stress  Matrix  for  details.  

End  Loads   The   line  end   load   results   are   based   on   the  end  force   and   end   moment   vectors   at   the   line  end.   There   are   3   groups  of   end  load  results:   x

Standard  results  like  Effective  Tension,  Bend  Moment,  etc.  are  available  at  line  ends  as  well  as  at  mid-­‐segment   points.  For  example  to  obtain  the  end  tension  at  End  A  you  can  ask  for  the  Effective  Tension  (or  Wall  Tension)  at   End  A.  

x

Magnitude  and  other  components  of  the  end  force  and  end  moment  vectors.  

x

Bend  Restrictor  Load,  which  is  a  special  end  load  result  useful  for  bend  restrictor  design.  

Sign  Convention  

When   considering   the   sign   of   end   load   components   the   question   arises   as   to   whether   the   load   reported   is   that   applied  by  the   line  to   its  connection  or   vice   versa.  The  OrcaFlex   convention  is  that   the   load   reported   at   any  point   is   that  applied  by  the  B  side  of  that  point  to  the  A  side.  So  at  End  A  we  report  the  end  load  applied   by  the  line  to  its   connection  (e.g.  a  vessel),  but  at  End  B  we  report  the  end  load  applied   to  the  line  by  its  connection.  This  is  in  keeping   with  the  OrcaFlex  convention  for  specifying   the  no-­‐moment  direction.   Treatment  of  Links  and  Winches  attached  to  the  end  node  

Normally,  the  end  force  and  end  moment  are  the  total  load  acting  between  the  end  node  and  the  object  to  which  it  is   connected.  This  includes  forces  from  any  links  or  winches  attached  to  the  end  node.   However   if   the   line   end   is   free,   or   has   been   released,   then   it   is   not   connected   to   any   object.   In   this   case   the   end   moment  is  zero  and   the  end   force  is  taken  to   be  the  total   force  acting  between  the  line   end  and  any  links   or   winches   attached  to  the  end   node.  If  there  are  no  attached  links  or  winches,  or  they  h ave  been  released,  then  the   end  force  is   zero.   Standard  Results   Effective  Tension,  Wall  Tension,   Shear  Force,  x-­‐Shear  Force,  y-­‐Shear  Force,   Bend  Moment,  x-­‐Bend  Moment,  y-­‐Bend  Moment,   Curvature,  x-­‐Curvature,  y-­‐Curvature  

These   results   variables   are   available   at   the  line   end   nodes,   as   well   as   at   mid-­‐segment   points.   Whether   you   are   given   end   values   or   mid-­‐segment   values   depends   on   the  point   at   which   you   ask   for   the   results.   If   you   ask   for   these   results   at   EndA   or   EndB,   or   at   an   arc   length   that   is   closer   to   a   line   end   than   to   the   nearest   mid-­‐segment   arc   length,   then   the   values  at  the  line  end  will  be  given.  Otherwise  the  values  for  the  nearest  mid-­‐segment  point  will  be  given.   For  mid-­‐segment  values  see  Line  Results:  Forces,  Line  Results:  Moments  and  Line  Results:  Pipe  Stresses.   At  a  line  end  they  report  the  components  of  the  end  loads  in  the  local  node  directions  of  the  end  node,  as  follows:   x

Effective  tension  is  the  component  of  the  end  force  vector  in  the  end  node  axial  direction  (=  Nz  direction).  

x

Wall  tension  is  derived  from  the  effective  tension  at  the  line  end,  using  the   pressure  effects  formula.  

x

Shear  is  the  component  of  the  end  force  vector  normal  to  the  end  node  axial  direction.  

x

x-­‐Shear  and  y-­‐Shear  are  the  components  of  the  end  force  vector  in  the  end  node  Nx  and  Ny  directions.  

x

Torque  is  the  component  of  the  end  moment  vector  in  the  end  node  axial  direction.  

x

Bend  moment  is  the  component  of  the  end  moment  vector  normal  to  the  end  node  axial  direction.  

x

x-­‐Bend  Moment  and  y-­‐Bend  Moment  are  the  components  in  the  end  node  Nx  and  Ny  directions.  

x

Stress  results  are  based  on  the  end  load  components  in  the  end  node  axes  directions.  

Differences  between  End  Loads  and  End  Segment  Loads  

The  end  values  of  these  results  differ  from  the  corresponding  values  for  the  end  segment  for  two  reasons.  

313  

w

 

System  Modelling:  Data  and  Results,  Lines    

Firstly,  they  include  the  loads  (weight,  buoyancy,  drag  etc.)  on  the  last  half  segment  adjacent  to  the  end.   Secondly,  they  are  components  in   the   local  node  directions  (Nx,Ny,Nz)  at  the   end  node,   whereas  the   end  segment   values   are   components   with   respect   to  the   segment   directions   (Sx,Sy,Sz).   The   end   node   is   often   not   aligned   with   the   end  segment  because  end  connection  stiffness  turns  it  towards  the  end  orientation  direction.  For  example:   x

If  the  end  connection  stiffness  is  zero,  or  if  the  line  end  is  free  or  has  been  released,  then  the  end  node  directions   are   aligned   with   the   end   segment   directions.   The   end   node   values   then   differ   from   the   end   segment   values   only   by  the  loads  on  the  end  half  segment.  

x

If   the   end   connection   stiffness   is  Infinity  (and   the   end   is  not   free   or  released)  then   the   end   node   directions   stay   aligned   with   the   line  end   axes   Ex,   Ey,   Ez.   The   end   node   values   are   then   usually   in   different   directions   to   the   end   segment  values.  

x

For   intermediate   values   of   end   connection   stiffness,   the   end   node   directions   will   be   somewhere   between   the   two.  They  will   tend   to   be   nearer  to   the  end  fitting   directions   if  the  end   connection   stiffnesses   are   stronger   than   the  line  bend  stiffness  and  torsional  stiffness,  but  nearer  to  the  end  segment  directions  if  it  is  weaker.  

End  Load  Magnitude  and  Components   End  Force,  End  Moment,   End  GX-­‐Force,  End  GY-­‐Force  and  End  GZ-­‐Force,  End  GX-­‐Moment,  End  GY-­‐Moment,  End  GZ-­‐Moment,   End  Lx-­‐Force,  End  Ly-­‐Force  and  End  Lz-­‐Force,  End  Lx-­‐Moment,  End  Ly-­‐Moment,  End  Lz-­‐Moment,   End  Ex-­‐Force,  End  Ey-­‐Force  and  End  Ez-­‐Force,  End  Ex-­‐Moment,  End  Ey-­‐Moment,  End  Ez-­‐Moment  

These  results  report  the  magnitude  of  the  end  force  and  end  moment  vectors,  and  their  components  in  the  following   directions:   x

The  directions  of  the  global  axes  GX,  GY,  GZ.  

x

The  directions  of  the  local  axes  Lx,  Ly,  Lz  of  the  object  to  which  the  line  end  is  connected.  For   example  if  the  line   end  is  connected  to  a  vessel,  the  Lx,  Ly,  Lz  are  the  directions  of  the  vessel  axes.  

x

The  directions  of  the  line  end  axes  Ex,  Ey,  Ez.  See  Line  End  Orientation.  

For  a  line  with  a  stiffener  attached  results  are  reported  separately  for  the  protected  line  and  its  stiffener.  However,  it   is  sometimes  necessary  (e.g.  when  designing  end  fittings)  to  report  combined  end  loads  including  the  load  from  both   the  protected  line  and  its  stiffener.   End  load  results  are  available  for  the  protected  line  which  include  the  stiffener  end  load,  in  addition  to  the  protected   line  end  load.  These  results  are  all  prefixed  with  "Total",  e.g.   Total  End  Load,  Total  End  Moment,  Total  End  GZ-­‐ Force  etc.   Bend  Restrictor  Load  

This  is  defined  as  Bend  Restrictor  Load  =  End  Force*(1   -­‐  cos(End  Force  Ez-­‐Angle)).  Another  commonly  used  name   for  this  variable  is  "pseudo-­‐curvature".  

6.8.6

Drag  Chain  Results  

For  details  on  how  to  select  results  variables  see  Selecting  Variables.   For  Drag  Chains  the  following  results  variables  are  available.   Azimuth  and  Declination  

The  azimuth  and  declination  of  the  drag  chain,  relative  to  global  axes.   Supported  Length  and  Hanging  Length  

The  supported  length  is  the  length  deemed  to  be  supported   by  the  seabed.  The  hanging  length  is  the  length  of  the   rest  of  the  drag  chain.  The   supported  length   plus  the  hanging  length   equals  the  total  length  of  the   drag  chain.   See   Drag  Chain  Seabed  Interaction  for  details  on  how  these  values  are  calculated.   Drag  Force  

The  magnitude  of  the  drag  force  acting  on  the  drag  chain.  This  includes  both  the  axial  and  normal  components  of  the   drag  force.   Axial  Drag  Force,  Normal  Drag  Force  

The  components  of  drag  force  axial  and  normal  to  the  drag  chain.  

314  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Horizontal  Drag  Force,  Vertical  Drag  Force  

The  horizontal  and  vertical  components  of  the  drag  force.  For  the  vertical  drag  force  a  positive  value  indicates  an   upwards  force.   See  Drag  Chain  Theory  for  details  on  how  the  drag  force  is  calculated.  

6.8.7

Flex  Joint  Results  

For  details  on  how  to  select  results  variables  see   Selecting  Variables.   For  Flex  Joints  the  following  results  variables  are  available.   Bend  Moment,  x-­‐Bend  Moment,  y-­‐Bend  Moment  

The  magnitude  of  bending  moment  at  the  Flex  Joint  and  its  components  in  the  local  x  and  y-­‐directions.  

6.8.8

Line  Setup  Wizard  

Intended   principally   for   mooring   analyses,   this   wizard   adjusts   line   configuration   to   achieve   specified   tension,   declination   or   layback.   The   wizard   is   available   when   the   current   simulation   is   in   Reset   state   and   is   opened   by   clicking  the  Calculation  |  Line  Setup  Wizard  menu  item.  

  Figure:  

The  Line  Setup  Wizard  

Calculation  Mode  

The  Wizard  has  two  modes  of  operation:  Calculate  Line  Lengths  or  Calculate  Anchor  Positions.  

315  

w

 

System  Modelling:  Data  and  Results,  Lines    

The  Calculate  Line  Lengths  option  operates  by  varying  the  length  of  the  specified  section  of  each  included  line  until   the   target   top   value   is   achieved.   Because   the   section   length   is   altered   by   the   Calculate   Line   Lengths   option   we   recommend   that   you   specify   the  segmentation   using   Target   Segment   Length   rather   than   Number   of  Segments.   Thus   if  the  wizard  lengthens  the  line  then  the  program  will  automatically  add  more  segments  to  meet  the  Target  Segment   Length.   The  Calculate  Anchor  Positions  option  operates  by  varying  the  position  of  the  Bottom  End  of  each  included  line   until   the   target   value   is   achieved.   The   Bottom   End   position   is   constrained   to   be   on   a   line   in   the   Lay   Azimuth   direction  so  you  must  set  these  data  items  before  using  the  wizard.   Note:  

These   calculations   use   a   numerical   method   which   is   not   100%  robust.   We   recommend   that   you   set   up  your  model  so  that  your  top  tension  /  top   declination   values  are  quite  close  to  your   exact  target   values  before  using  the  wizard.  If  you  do  this  the  wizard  is  more  likely  to  be  able  to  find  a  solution.  

Target  Values  

To   use   the   wizard   you   must   first  specify  the  target  tensions  or   declinations.   The   wizard  presents  a  list   of  all  Lines  in   the  model.   The  Include  this  Line  option  determines  which  Lines  are  included  in  the  calculation  which  allows  you  to  exclude   certain  Lines.  For  example,  you  may  be  modelling  both  moorings  and  flow  lines  in  the  same  OrcaFlex  file.  Typically   you  would  only  include  the  moorings  in  this  calculation.   The  Target  Variable  option  allows  you  to  switch  between  the  following  options:   x

End  A  Tension  or  End  B  Tension.  

x

End  A  Horizontal  End  Force  or  End  B  Horizontal  End  Force.  

x

End  A  Declination  or  End  B  Declination.  

x

Layback,   defined   to   be   the   horizontal   component   of   distance   between   the   Top   End   of   the   line   and   the   touchdown  point.  

x

No   Target,   which   means   that   the   line   will   be   included   in   the   static   calculation   but   that   its   data   is   not   to   be   modified.  This  is  particularly  useful   if   you   are   analysing   buoyed   systems  where   the  upper   and   lower   catenaries   are  modelled  with  different  Lines.  

Finally   you  specify   Tension,   Horizontal   End   Force,   Declination   or   Layback   values,   as   applicable,   in   the  Target   Value   field.   Line  section  to  be  modified  

Specifies  which  section  on  the  line  is  to  have  its  length  modified  when  using  the  Calculate  Line  Lengths  mode.   Convergence  Parameters  

The  Wizard  performs  an  iterative  calculation  and  these  parameters  can  be  used  to  help  convergence.   The   calculation   is   abandoned   if   convergence   has   not   been   achieved   after   the   number   of   steps   specified   by   Max   Iterations.  For  some  difficult  cases  simply  increasing  this  limit  may  be  enough.   The   non-­‐dimensional   Tolerance   parameter   determines   when   the   calculation   is   deemed   to   have   converged.   The   calculation  has  converged  once  the  following  conditions  are  satisfied:   x

Calculated  Value  <  Tolerance  ×  Typical  Force  for  tension  and  end  force  targets.  Typical  Force  for  a  line  is  defined   as  to  be  the  total  dry  weight  of  the  line.  

x

Calculated  Value  <  Tolerance  for  declination  targets.  

x

Calculated  Value  <  Tolerance  ×  Target  Value  for  a  layback  target.  

The  Min  Damping  and  Max  Damping  parameters  can  sometimes  be  used  to  help  difficult  problems  converge.  Try   increasing  the  Min  Damping  factor,  say  values  in  the  range  1.5  to  10.  You  can  also  try  increasing  the  Max  Damping   factor,  say  to  values  in  the  range  10  to  100.  

6.8.9

Line  Type  Wizard  

The  Line  Type  Wizard  is  a  tool  that  helps  you  set  up  a   Line  Type  that  represents  one  of  the  following  commonly  used   structures:   x

Chain.  

316  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

x

Rope/Wire.  

x

Line  with  Floats.  

x

Homogeneous  Pipe.  

x

Hose.  

x

Umbilical.  

What  the   Wizard  does   is  ask  you  for   the   basic  data  of  the  structure   Ȃ  e.g.  the  bar  diameter   for  a  chain   Ȃ  and  then   calculate   for   you   as   much   of   the   line   type   data   as   it   reasonably   can   for   representing   that   structure.   The   Wizard   leaves  you  to  set  other  data  Ȃ  e.g.  friction  coefficients  Ȃ  where  there  is  no  formula  on  which  to  base  the  data.   Warning:  

The  values  generated  by  the  Wizard  are  offered  in  good  faith,  but  due  to  variations  in  properties   between  products  they  cannot  be  guaranteed.  Please  use  suppliers'  data  where  this  is  available.  

How  The  Line  Type  Wizard  Works   The  Wizard  works  on  the  currently  selected  line  type  on  the   line  types  form,  so  you  should  first  create,  name  and   select   the   Line   Type   that   you   want   to   set   up.   You   can   then   open   the   Wizard   using   the   Wizard   button   on   the   Line   Types  form.   The   first   time   you   use   the  Wizard   on   a   given   line   type   you  must   be   in  reset   state,   since   you   will   be   setting   data.   You   then   tell   the   Wizard   the   category   of   structure   that   you   want   to   model   (chain,   rope   etc.)   and   the   data   for   that   structure  (e.g.  chain  bar  diameter).  This  information  is  called  the  Wizard  data,  and  from  it  the  Wizard  derives  line   type  data  to  correspond  to  that  Wizard  data.  If  necessary  you  can  then  manually  adjust  the  derived  line  type  data.   Once  you  have  used  the   Wizard  to  set  up  data  for  a  given  line  type,  then  the  Wizard  remembers  the   Wizard  data  you   gave   it.   If   you   re-­‐open   the   Wizard   when   in   reset   state   then   you   can   edit   the   Wizard   data   and   the   Wizard   will   calculate  corresponding  new  derived  line  type  data.  Any  manual  adjustments  will  need  to  be  done  again.   You   can   also   re-­‐open   the  Wizard   when   in   other   states   (e.g.   in  static   state   or   when   a   simulation   is  active)   but  only   in   order  to  view  the  Wizard  data.  You  cannot  edit  Wizard  data  or  re-­‐derive  line  type  data  except  in  reset  state.   Note:  

Remember  that  the  current  line  type  data  might  not  correspond  to  the  current  Wizard  data,  since   you  might  have  manually  edited  the  line  type  data  after  it  was  derived  by  the  Wizard.  

Using  the  Line  Type  Wizard   The  Wizard  has  three  stages,  with   Next  and  Back  buttons  so  that  you  can  move  between  stages  to  set  up  the  data   you  want.   Stage   1   displays   the   name   of  the   selected   Line   Type   and   asks   you   to   specify   the   special   category   that   you   want.   You   can  then  click  Next  to  proceed  to  the  second  stage.   Stage   2  presents   3   frames  of   information.   The   top   left  frame   asks   you   for   the  basic   data   of   the   special   category   you   have   selected.   The   bottom   left   frame   displays   the   resulting   derived   Line   Type   data   Ȃ   you   should   check   that   the   values  are  reasonable.   The  right  hand  frame  displays  other  properties  of  the  resulting  Line  Type,  which  are  often  useful  as  a  check.  In  some   cases   these   depend   on   contents   density,   in   which   case   you   can   specify   the   contents   density   to   be   used   for   the   calculation  of  properties.  If  there  are  any  errors  then  a  message  will  be   displayed.  When  everything  is  correct  you   can  click  Next  to  proceed  to  the  last  stage.   Stage  3  displays  all  of  the  Line  Type  data.  Bold  text  is  data  that  has  been  derived  for  you  by  the  Wizard,  based  on  the   special  line  type  data  you  specified.  Non-­‐bold  text  is  data  that  has  not  been  set  by  the  Wizard  Ȃ  this  data  will  be  as   you  last  set  it.  You  can  adjust  any  of  the  data  at  this  stage,  overriding  the  values  derived  by  the  Wizard  if  you  wish.   You   can   also   still   go   back   to   previous   stages   of   the   Wizard   if   further   modifications   are   required.   When   everything   is   correct  you  can  click  the  Finish  button,  in  which  case  the  new  data  will  be  written,  overwriting  the  previous  data  for   that   line   type.   Alternatively,   you   can   Cancel   to   leave   the   line   type   unchanged,   but   then   any   newly   entered   special   category  data  will  also  be  lost.  

6.8.10

Chain  

A  chain  can  be  modelled  in  OrcaFlex  by  using  a   Line  Type  with  its  various  properties  set  to  suitable  values.  This  note   derives  the  values  to  use  for  anchor  chain  of  nominal  (i.e.  bar)  diameter  D,  as  shown  in  the  Figure:  Chain  Geometry.   The  properties  of  an  equivalent  line  type  are  given  below.  

317  

w

 

System  Modelling:  Data  and  Results,  Lines    

 

Studless  

Studlink  

OD  

1.80D  

1.89D  

ID  

0  

0  

Contact  diameter  

3.35D  

3.6D  

Mass/Length  

19.9D2  

21.9D2  te/m  for  D  in  m  

Axial  stiffness  

0.854x108D2  

1.01x108D2  kN  for  D  in  m  

Bend  stiffness  

0  

0  

Limit  Compression  

yes  

yes  

Normal  drag  coefficient  

1.0  

1.0  

Normal  drag  diameter  

2.10D  

2.26D  

Axial  drag  coefficient  

0.4  

0.4  

Axial  drag  diameter  

ͲǤͷͶȀɎ  

ͲǤ͸ͲȀɎ  

Normal  added  mass  coefficient   1.0  

1.0  

Axial  added  mass  coefficient  

0.08  

0.07  

Stress  diameters  

'~'  

'~'  

Allowable  stress  

'~'  

'~'  

Friction  coefficient  

typically  0.4  -­‐  0.8  depending  on  the  seabed  

Reference  

Puech  A,  1984.  

Geometry     D = Nominal Diameter   AFACE   AEDGE  

3.35D (studless) 3.6D (studlink)  

6D  

3.35D (studless) 3.6D (studlink)  

  Figure:  

Chain  Geometry  

Data   Chains  are  widely  used  in  a   variety  of  offshore  applications,  most  obviously  in  mooring.  The  Line  Type   Wizard  helps   derive  a  line  type  to  represent  a  chain  based  on  the  following  input  data.   Bar  Diameter  

The  diameter  of  the  metal  bar  that  forms  the  links.   Link  Type  

Can  be  either  studlink  or  studless.  

318  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Mechanical  Properties   Catalogue  Data   When  modelling  mooring  chain  the  Line  Type  Wizard  aims  to  derive  data  for  a  line  type  whose  characteristics  are   equivalent  to  that  of  a  chain.   Warning:  

The  values  generated  by  the  Wizard  are  approximate  only  and  are  intended  as  first  estimates  for   preliminary   use.   They   are   offered   in   good   faith,   but   due   to   variations   in   properties   between   products  they  cannot  be  guaranteed.  Please  use  suppliers'  data  where  this  is  available.  

In   deriving   these   some   of   the   available   catalogue   data  will   prove   useful   and   we   outline   here   the   relevant   aspects.   The  Mooring   Chain   figure  shows  the   geometry  of   a  pair  of   chain   links.   The   values   are   given   in  terms   of   the   nominal   bar  diameter  of  the  chain  (D),  assumed  to   be  in  metres,  and  are  given  for  both  a  studless  chain  and,   where  different,   for   a   studlink   chain.   The   geometry   given   in   the   figure   is   based   on   catalogue   data   available   from   the   chain   manufacturer  Scana  Ramnas  (1990  &  1995),  as  is  the  following  expression  for  mass  per  metre:   Mass  per  metre  (M)  =  19.9D2  te/m  (studless)  or  21.9D2  te/m  (studlink).   The  catalogue  also  gives  the  following  value  for  the  Young's  Modulus  of  the  chain  that  has  been  deduced  from  stress-­‐ strain  relationships  in  which  the  cross-­‐sectional  area  of  two  bars  is  taken  to  be  the  load  bearing  area:   E  =  5.44  x  107  kN/m2  (studless)  or  6.40  x  107  kN/m2  (studlink).   Minimum  Breaking  Loads  

For  information,  the  properties  window  displays  minimum  breaking  loads  that  depend  on  the  nominal  diameter  and   chain   grade.   They   are   derived   using   the   following   relationship,   which   was   obtained   from   the   manufacturer's   catalogue:   Min  Breaking  Load  =  c.D2.(44  -­‐  80D)  kN   where  c  is  a  grade-­‐dependent  constant,  given  in  the  catalogue  data  as  Grade  2:  1.37e4,  Grade  3:  1.96e4,  ORQ:  2.11e4,   R4  -­‐  2.74e4.   Studless  and  Studlink  chains  with  the  same  nominal  diameters  are  stated  to  withstand  the  same  break-­‐  and  proof-­‐ loads.   Derived  Data   It  will   be  useful   to  know  the   centreline  length  of   bar   needed  to  make  a  single  link.   We  can  obtain  this   by  noting  that,   for  a  long  chain,  there  is  one  chain  link  every  4D  length  of  chain.  Hence,  the  number  of  links  per  metre  of  chain  is  N  =   1/(4D),  and  thus  for  a  single  link:   Mass  per  link  =  M  /  N  =  79.6D3  te  (studless)  or  87  6D3  te  (studlink).   ••—‹‰–Šƒ––Š‡…Šƒ‹‹•ƒ†‡ˆ”‘•–‡‡Žǡƒ†—•‹‰ɏs  as  density  of  steel  (=  7.8  te/m3),  this  then  leads  to:   Volume  per  Ž‹αȋȀȌȀɏs  =  10.2D3  m3  (studless)  or  11.2D3  m3  (studlink).   But,  by  considering  the  geometry  of  a  link,  we  also  have   ‘Ž—‡αǤɎ2/4,   where  L  is  centreline  length  of  bar  needed  to  make  a  single  link  (including  the  stud  in  the  case  of  the  studlink  chain).   Hence:   α‘Ž—‡ȀȋɎ2/4)  =  13.0D  m  (studless)  or  14.3D  m  (studlink).  

Outer,  Inner  and  Contact  Diameter   The  Line  Type  Wizard  sets  up  diameters  for  a  chain  as  follows:   Outer  Diameter  

The   effective   outer   diameter   of   the   equivalent   line   is   obtained   using   a   similar   argument   to   that   deployed   in   obtaining  the  overall  length  of  bar  per  link.  Firstly,  note  that  the  volume  per  metre  can  b e  expressed  as  both:   ‘Ž—‡’‡”‡–”‡αȀɏs   and  also  as   ‘Ž—‡’‡”‡–”‡αɎ2/4  

319  

w

 

System  Modelling:  Data  and  Results,  Lines    

where   OD   is   the   equivalent   diameter   for   a   line   with   constant   volume   along   its   length.   Equating   these   expressions   leads  to:   —–‡”‹ƒ‡–‡”αȏͶȀȋɎɏs)]½  =  1.80D  m  (studless)  or  1.89D  m  (studlink).   Inner  Diameter  

Chains  do  not  have  any  contents,  so  the  Inner  Diameter  is  set  to  zero.   Contact  Diameter  

The  contact  diameter  is  set  to  the   chain  link  envelope  diameter.  That  is  Contact  Diameter  =  3.35D  m  (studless)  or   3.6D  m  (studlink).  

Axial  and  Bending  Stiffness   The  Line  Type  Wizard  sets  up  Axial  and  Bending  Stiffness  and  Limit  Compression  for  a  chain  as  follows:   Axial  Stiffness  

As  detailed  in  Mechanical  Properties  of  Mooring  Chains  we  have  values  for  the  Young's  Modulus  for  both  studlink   and  studless  chains  from  catalogue  data.  Taking  A  to  be  the  combined  cross-­‐sectional  area  of  two  bars,  that  is:   αʹȋɎ2  /  4)  m2   leads  to:   EA  =  0.854  x  108  D2  kN  (studless)  or  1.01  x  108  D2  kN  (studlink).   Bending  Stiffness  

For  both  studlink  and  studless  chains  the  bending  stiffness  is  set  to  zero  as  the  chains  are  assumed  to  bend  when   subjected  to  very  small  moments.   Limit  Compression  

In  conjunction  with  a  zero  value  for  bend  stiffness,  Limit  Compression  is  set  to  'yes'.  

Axial  Added  Mass  Coefficient   The  Line  Type  Wizard  sets  up  Axial  Added  Mass  Coefficient  for  a  chain  as  follows.   As  for  axial  drag,  the  parts   attracting  added  mass  in  axial  flow  are  the  projecting  lobes  only   Ȃ  see  the  figure.   Each   pair   of   lobes   are  simply  a   link   with  the   middle   section   (and   stud   if   present)  removed,   and   can   be   viewed   roughly  as   an  ellipsoid  split  down  the  centre  with  the  following  dimensions:   length  6D,  width  D  and  height  2.35D  (studless)  or  2.60D  (studlink).   J  N  Newman  (1977)  (page  147,  Fig  4.8)  gives  added  mass  coefficients  for  spheroids.  We  approximate  the  ellipsoid  as   a   spheroid   with   a   lenbgth   of   6D   (the   link   length)   and   a   mean   width   of   (D+2.35D)/2   =   1.675D   for   studless),   or   (D+2.60D)/2  =  1.80D  for  studlink.  This  gives  an  aspect  ratio  (width/length)  of  about  0.3  in  both  cases.   For  this  aspect  ratio  and  axial  flow  Newman  gives  an  added  mass  coefficient  of  0.1.  This  is  for  use  with  a  reference   volume   equal   to   the   spheroid   volume,   which  in   this   case   is  the   volume   of   the   lobes   only.  But  OrcaFlex   uses   the   total   displaced  volume  as  the  reference  volume,  so  a  suitable  scaling  of  Newman's  coefficient  is  needed  to  allow  for  this   difference  in  reference  volume.   For   a   studless   link   the   non-­‐lobe   length   of   bar   is   2D   (the   middle   section)   out   of  a   total   of  13D,   so   the   lobes   represent   11D/13D  =  84.6%  of  the  total  link  volume.  For  a  studlink  chain  the  non-­‐lobe  length  of  bar  is  2D  +  1.6D  (the  stud)  our   of   a   total   of   14.3D,   so   the   lobes   represent   10.7D/14.3D   =   74.8%   of   the   total   link   volume.   So   we   scale   Newman's   added  mass  coefficient  0.1  by  these  factors  and  round  to  give  OrcaFlex  axial  added  mass  coefficients  of   Caa  =  0.08  (studless)  or  0.07  (studlink).  

Axial  Drag   The  Line  Type  Wizard  sets  up  the  axial  drag  coefficient  and  axial  drag  diameter  for  a  chain  as  follows.   Generally,  axial  drag  is  very  low  for   smooth  pipes,  being   due  to  skin  friction  only.  However,  for  a  chain  there  is  some   projected  area  present  even  in  axial  flow  and  we  consider  the  drag  force  due  to  this  effect.  We  ignore  the  effect  of   skin  friction  in  the  derivation  outlined  below.  

320  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

As   in   the   calculation   for   normal   flow   we   consider   two   adjacent   links   and   calculate   their   projected   area.   The   projected  area,  normal  to  the  flow,  for  axial  flow  consists  of  the  four  "lobes"  only,  since  the  central  part  is   effectively   shielded  from  the  flow  Ȃ  see  the  figure.   Hoerner  (1965),  page  5-­‐8,  Fig  14c,  gives  Cda  =  0.32  for  a   hemispherical  rivet  head   projecting  from  a  plane.  The  lobes   here  are  similar  Ȃ  more  elongated  in  the  flow  direction  (implying  a   lower  Cda)  but  on  a  less  smooth  body  (implying  a   higher  Cda).  Hence,  we  assume:   Cda  =  0.40.   The   reference  drag  area  that  corresponds  to  this  is  the  axial  projected  area  of  the  lobes.  Each  lobe   is  (3.35D  -­‐  D)/2  =   1.175D  long  for  studless,  or  (3.6D   -­‐  D)/2  =  1.30D  long  for  studlink,  consisting  of  a  semicircle  of  diameter  D  on  the   end   of   a   rectangle   of   width   D   and   length   1.175D   -­‐   0.5D   =   0.675D  (studless),   or   1.3D   -­‐   0.5D   =   0.8D  (studlink).   And   each  pair  of  links  gives  4  lobes,  so  the  total  axial  projected  area  p er  metre  of  chain  (=  1/8D  pairs  of  links)  is  given  by:   ͶǤȋɎ2/8  +  0.675D2)  /  (8D)  =  0.54D  for  studless   ͶǤȋɎ2/8  +  0.8D2)  /  (8D)  =  0.60D  for  studlink.   ‘™‡˜‡”ˆ‘”ƒš‹ƒŽ†”ƒ‰”…ƒ Ž‡š—•‡•ƒ”‡ˆ‡”‡…‡†”ƒ‰ƒ”‡ƒ‡“—ƒŽ–‘Ɏ a,   where   Da   is   the   axial   drag   diameter.   So   the   axial  drag  diameter  for  OrcaFlex  is:   Da  αͲǤͷͶȀɎȋ•–—†Ž‡••Ȍ   Da  αͲǤ͸ͲȀɎȋ•–—†Ž‹ȌǤ  

Normal  Drag   The  Line  Type  Wizard  sets  up  the  Normal  Drag  Coefficient  for  a  chain  as  follows:   We   first   calculate   the   drag   force   on   a   chain   in   normal   flow,   for   which   we   require   a   value   for   its   projected   area   (normal   to   the   flow).  To   calculate   this   we  must   consider  the   chain   as   a   collection   of  pairs   of   adjacent   links,   one   face   on   to   the   flow,   with   projected   area   AFACE,   and   one   edge   on,   with   projected   area   AEDGE   Ȃ   see   Figure.   The   overall   projected  area  per  metre  will  be  a  multiple  of  the  sum  of  these  two  areas.   AFACE  =  L  D  -­‐  2D2  =  11.0  D2  m2  (studless)  or  12.3D2  m2  (studlink)   and   AEDGE  αͷΪʹȋɎ2/4)/2  =  5.79  D2  m2.   There  are  1/(4D)  links  per  metre  and  hence  1/(8D)  such  pairs  of  links  per  metre.  Hence,  the  total  projected  area  per   metre  (normal  to  the  flow)  is  given  by  the  following  expression:   ANORMAL  =  (AFACE  +  AEDGE)  (1/(8D))  =  2.10D  m  (studless)  or  2.26D  m  (studlink).   So,  we  are  now  able  to  calculate  the  drag  force  per  metre  length  of  chain  as:   Drag  fo”…‡αΦɏ˜2  Cdn  ANORMAL   for  a  given  drag  coefficient  Cdnǡ™Š‡”‡ɏ‹•–Š‡†‡•‹–›‘ˆ•‡ƒ™ƒ–‡”ƒ†˜‹•–Š‡ˆŽ‘™˜‡Ž‘…‹–›Ǥ ‘”‹””‡‰—Žƒ”•Šƒ’‡† bluff  bodies  such  as  chain  links,  of  either  type,  a  suitable  value  for  Cd n  is  1.0.   The  drag  force  per  metre  length  as  calculated  by  OrcaFlex  is  given  by:   ”ƒ‰ˆ‘”…‡αΦɏ˜2  Cdn  Dn   where  Dn  is  the  normal  drag  diameter.  Equating  the  two  equations  for  drag  force  leads  to:   Dn  =  ANORMAL  =  2.10D  m  (studless)  or  2.26D  m  (studlink).  

Normal  Added  Mass  Coefficient   The  Line  Type  Wizard  sets  up  the  Normal  Added  Mass  Coefficient  for  a  chain  as  follows:   When  a  line  is  accelerated  in  water  it  requires  an  impulse  in  excess  of  that  needed  for  the  same  acceleration  in  air.   This   is   due   to   the   extra   force   required   to   displace   the   water   in   the   vicinity   of   the   submerged   part   of   the   line.   An   added  mass  term  is  used  to  reflect  this  and  it  is  found  to  be  proportional  to  the  volume  of  displaced  fluid:   ††‡†ƒ••αƒǤɏǤ‘Ž   where   ɏ‹•†‡•‹–›‘ˆ™ƒ–‡”ǡ  

321  

w

 

System  Modelling:  Data  and  Results,  Lines    

Vol  is  the  displaced  volume.   The   parts   of   a   line   displacing   the   fluid   are   said   to   be   attracting   added   mass.   For   asymmetrical   bodies   the   parts   attracting  added  mass   will  differ  in  different   directions.  Hence,  we  consider  the   effect  due   to  fluid  flow   exerting  a   force  in,  first,  the  normal  and  then  the  axial  directions.   For  a  circular  cylinder  in  flow  normal  to  its  axis:   Can  =  1.0.   The  situation  for  a  chain  is   more  complicated  as,  for   flow  normal  to  a  link,  parts  of  the  link  are  shielded  from   the   flow  but  there  is  also  some  entrapped  water  within  each  edge-­‐on  link.  An  accurate  calculation  is  very  problematic   and  is  unlikely  to  give  a  value  for  the  normal  added  mass  coefficient  far  distant  from  1.0.  Hence  we  assume:   Can  =  1.0.  

Stress  Diameters  and  Allowable  Stress   These   are   set   to   '~'   because   the   entire   structure   is   load   bearing.   Please   note,   however,   that   stress   results   are   not   meaningful  for  a  chain.  

6.8.11

Rope/Wire  

D = Nominal rope diameter

D

Wire with Fibre core

Fibre rope Figure:  

Wire with Wire core

 

Rope/Wire  Geometry  

Ropes   and   wires   have   many  applications   in   the   offshore   industry   including   towing,   mooring   and   winching.  The  Line   Type   Wizard   can   be   used   to   derive   Line   Type   data   to   represent   five   different   types:   Nylon   (8-­‐strand   Multiplait);   Polyester   (8-­‐strand   Multiplait);   Polyethylene   (8-­‐strand   Multiplait);   6x19   Wire   Rope   with   Fibre   Core;   and   6x19   Wire   Rope  with  Wire  Core.   Most   of   the   calculations   of   the   derived   line   properties   are   based   on   data   from   a   catalogue   published   by   Marlow   Ropes   Ltd   (1995).   All   quantities   are   expressed   as   a   function   of   the   rope's   nominal   diameter   D.   Note   that   this   documentation   uses   the   SI   units   system,   so   D   is   in   metres   in   this   documentation,   but   the   program   automatically   adjusts  the  formulae  to  match  the  units  specified  by  the  user.   Warning:  

The  values  generated  by  the  Wizard  are  approximate  only  and  are  intended  as  first  estimates  for   preliminary   use.   They   are   offered   in   good   faith,   but   due   to   variations   in   properties   between   products  they  cannot  be  guaranteed.  Please  use  suppliers'  data  where  this  is  available.  

Data   The   Line   Type   Wizard   can   be   used   to   create   line   types   representing   a   variety   of   ropes   and   wires.   The   input   data   required  consists  of  the  following:   Rope/Wire  Nominal  Diameter  

The  overall  diameter  of  the  rope  or  wire.  The  majority  of  the  derived  line  type  data  are  functions  of  this  diameter.   Warning:  

The  line  type  outer  diameter  derived  by  the  wizard  is  less  than  this  nominal  diameter,  in  order  to   give   the   correct   buoyancy.   You   need   to   allow   for   this   when   setting   the   line   type   drag   and   added   mass  coefficients,  since  the  coefficients  correspond  to  the  derived  line  type  outer  diameter,  not  the   nominal  diameter.  

322  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Construction  

Can  be  one  of:   x

Nylon  (8-­‐strand  Multiplait).  

x

Polyester  (8-­‐strand  Multiplait).  

x

Polyethylene  (8-­‐strand  Multiplait).  

x

6x19  Wire  with  Fibre  Core  (8-­‐strand  Multiplait).  

x

6x19  Wire  with  Wire  Core  (8-­‐strand  Multiplait).  

The  construction  affects  both  the  mass  per  unit  length  of  the  line  type  and  the  strength  of  the  line  type.  

Mass  per  unit  length   The  Line  Type  Wizard  sets  up  Mass  for  a  Rope/Wire  as  follows:   The   quantity   Mass   per   unit   length   is   available   from   catalogue   data   for   ropes.   The   nominal   rope   diameter   and   nominal   mass   are   available   for   a   variety   of   rope   constructions.   A   simple   statistical   analysis   of   the   available   data   leads  to  the  following  expressions:   Mass  Per  Metre  =  0.6476  D2  te/m  (for  Nylon  ropes).   Mass  Per  Metre  =  0.7978  D2  te/m  (for  Polyester  ropes).   Mass  Per  Metre  =  0.4526  D2  te/m  (for  Polypropylene  ropes).   Mass  Per  Metre  =  3.6109  D2  te/m  (for  Wire  ropes  with  fibre  core).   Mass  Per  Metre  =  3.9897  D2  te/m  (for  Wire  ropes  with  wire  core).  

Outer  and  Inner  Diameters   The  Line  Type  Wizard  sets  up  outer  and  inner  diameters  for  a  Rope/Wire  as  follows.   The  inner  diameter  is  set  to  zero  for  all  rope  construction  types.  The  line  type  outer  diameter,  OD,  is  set  as  follows:   OD  =  0.85  D  (for  Nylon  ropes).   OD  =  0.86  D  (for  Polyester  ropes).   OD  =  0.80  D  (for  Polypropylene  ropes).   OD  =  0.82  D  (for  Wire  ropes  with  fibre  core).   OD  =  0.80  D  (for  Wire  ropes  with  wire  core).   where  D  is  the  specified  rope  diameter.   These   outer   diameters   are   effective   diameters   that   give   the   line   type   a   displaced   volume   per   unit   length  that   equals   the  estimated  displaced  volume  per  unit  length  of  the  rope/wire.  The  line  type  then  has  the  appropriate  buoyancy.   Note   that   this   effective   diameter   is   less   than   the   specified   rope   diameter,   because   there   are   gaps   between   the   fibres   and  so  not  all  of  the  specified  nominal  diameter  contributes  to  buoyancy.   The  above  formulae  for  the  line  type  OD  were  derived  by  equating  the  line  type  displaced  volume  per  unit  length,   Ɏ2ȀͶǡ–‘–Š‡†‹•’Žƒ…‡†˜‘Ž—‡’‡”‡–”‡ǡȀɏǡ™Š‡”‡‹•–Š‡”‘’‡Ȁ™‹”‡mass  per  unit  length  ƒ†ɏ‹•–Š‡ƒ˜‡”ƒ‰‡ density  of  the  material.   Š‡ ˆ‘ŽŽ‘™‹‰ ƒ˜‡”ƒ‰‡ ƒ–‡”‹ƒŽ †‡•‹–‹‡• ɏ ȋ‹ –‡Ȁ 3)   were   assumed:   Nylon   1.14;   Polyester   1.38;   Polypropylene   0.91;   Wire   with   fibre   core   6.87;   Wire   with   Wire   core   7.85.   The   average   material   density   for   the   Wire   with   fibre   core   was   estimated   by  assuming  a   ratio   of   6:1  between   the   wire   and   fibre   volume,   with  the   fibre   taken  to   have   the   same   density  as  (fresh)  water.  

Axial  and  Bending  Stiffness   The  Line  Type  Wizard  sets  up  Axial  and  Bending  Stiffness  and  Limit  Compression  for  a  Rope/Wire  as  follows   Axial  Stiffness  

The  expressions  for  axial  stiffness  are  calculated  in  different  ways  for  the  two  groups  of  fibre  ropes  and  wire  ropes.  

323  

w

 

System  Modelling:  Data  and  Results,  Lines    

For   Fibre   Ropes   we   use   the   catalogue   data.   Load/extension   characteristics   depend   on   previous   load   history,   whether   the   rope  is  wet  or  dry,  and   the   rate   of   application   of   the   load.   To  reflect   the   likely  working  environment   of   the  rope  we  use  data  associated  with  ropes  that  have  been  tested  under  the  following  conditions:   x

the  rope  has  been  pre-­‐worked  Ȃ  loaded  to  50%  of  breaking  load  and  then  rested  for  24  hours  (this  causes  the   rope  to  bed  down  so  that  its  elastic  behaviour  is  more  consistent  and  repeatable)  

x

subjected  to  slowly  varying  loads  (for  loads  varying  at  wave  frequency,  stiffness  should  be  about  twice  the  value   shown)  

x

a  wet  rope  Ȃ   pre-­‐soaked  in   water  (this   is  most  significant  for  Nylon   ropes   which  suffer  a  loss  in   performance   when  wet)  

x

we   use   figures   for   the   average   performance   when   the   mean   extension   is   10%   (by   taking   the   tangent   of   the   stress-­‐strain  curve  at  10%).  

Incorporating   all   of   the   factors   indicated   above   we   can   produce   values   of   axial   stiffness   for   a   range   of   rope   diameters.   Once   again   using   simple   statistical   techniques   we   obtain   the   following   expression   for   axial   stiffness   of   fibre  ropes:   Axial  Stiffness  =  1.18  x  105  D2  kN  (for  Nylon  ropes).   Axial  Stiffness  =  1.09  x  106  D2  kN  (for  Polyester  ropes).   Axial  Stiffness  =  1.06  x  106  D2  kN  (for  Polypropylene  ropes).   Axial   stiffness   for  Wire   Ropes   is  calculated   directly,  rather   than   estimated   from  empirical   relationships.  We   assume   a  value  for  Young's  Modulus,  for  the  6x19  strand  group,  of:   E  =  1.03  x  108  kN/m2  (for  Wire  ropes  with  fibre  core).   E  =  1.13  x  108  kN/m2  (for  Wire  ropes  with  wire  core).   and  work  on  an  assumed  metallic  area  of:   αͲǤͶͷͷȋɎ2/4)  m2  (for  both  wire  ropes).   Both  of  these  quantities  have  been  obtained  from  the  HER  Group  Marine  Equipment  &  Wire  Rope  Handbook.  Note   that   for   wire   ropes   with   a   wire   core   the   additional   axial  stiffness   is   accounted   for   in   the   enhanced   Young's   modulus.   This  leads  to:   Axial  Stiffness  =  3.67  x  107  D2  kN  (for  Wire  ropes  with  fibre  core).   Axial  Stiffness  =  4.04  x  107  D2  kN  (for  Wire  ropes  with  wire  core).   Bending  Stiffness  

For  all  rope  construction  types  the  bending  stiffness  offered  by  the  Wizard  is  zero.  For  systems  where  bend  stiffness   is  a  significant  factor  you  should  override  this  value  with  the  true  value  obtained  from  the  rope  supplier.   Limit  Compression  

In  conjunction  with  a  zero  value  for  bend  stiffness  Limit  Compression  is  set  to  yes.  

Stress  Diameters  and  Allowable  Stress   These   are   set   to   '~'   because   the   entire   structure   is   load   bearing.   Please   note,   however,   that   stress   results   are   not   meaningful  for  complex  structures  such  as  ropes  or  wires.  

Minimum  Breaking  Loads   The   properties   window   in   the   line   type   wizard   displays   approximate   minimum   breaking   load   (MBL)   values   for   ropes  and  wires.  These  may  be  useful  for  setting  the  Allowable  Tension  data  item  for  the  line  type.   The   MBL   values   displayed   are   calculated   using   the   following   functional   formulae,   where   D   is   rope/wire   nominal   diameter  in  metres:   Nylon  ropes  (dry)  

163950.D2  kN  

Nylon  ropes  (wet)  

139357.D2  kN  

Polyester  ropes  

170466.D2  kN  

Polypropylene  ropes  

105990.D2  kN  

324  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Wire  ropes  with  fibre  core  

584175.D2  kN  

Wire  ropes  with  wire  core  

633358.D2  kN  

These  formulae  were  derived  from  manufacturer's  catalogue  data,  which  consist  of   minimum  (dry)  strength  against   nominal   diameter   for   each   of   the   five   rope/wire   constructions.   The   formulae   were   derived   using   least   squares   fitting,   and   they   were   found   to   give   a   good   fit   to   the   manufacturer's   data,   except   that   they   tend   to   underestimate   MBL  for  small  diameter  non-­‐wire  ropes.   Note:  

6.8.12

Nylon  ropes   lose  some  strength   when   wet;  the  formula   given  for   wet   nylon  ropes   is  based  on  the   manufacturer's  statement  that  they  can  lose  up  to  15%  of  their  (dry)  strength  when  wet.  

Line  with  Floats  

You   can   model   floats   or   buoyancy   modules   attached   to   a   line   by   using   buoyant   Clumps   attached   at   the   relevant   points.  However  when  a  number  of  floats  are  supporting  a  length  of  line  it  is  often  easier  to  model  the  buoyancy  as  if   it  were  smeared,  i.e.  spread  out  evenly,  along  that  part   of  the  line.  This  allows  the  length  and  segmentation  of  the   buoyed  section  to  be  varied  easily  without  having  to  add  and  remove  individual  floats.   To  use  this  smeared  properties  approach  you  need  to  do  the  following:   x

Create  a  new  line  type.  

x

Set  the  new  line  type's  properties  to  be   equivalent  to  those  of  the  original  pipe+floats.  This  is  done   by  spreading   each  float's   buoyancy,  drag,  etc.   uniformly  over   the   length  of   pipe   from   S f/2  before   the   float   centre  to   Sf/2   after   the  float  centre,  where   Sf  is  the  float  pitch,  i.e.  the  spacing  between  float  centres  (see   diagram  below).   The   result   is  a  uniform  circular  section  line  which  will  experience  the  same  forces  per  unit  length  as  the  original  line  plus   floats.  The  line  type  wizard  will  automatically  set  up  this  equivalent  line  type  for  you.  

x

Set   up   a   line  section  to  model   the   length  of   line   supported  by  the   floats.  The   section's   line   type   should   be   set   to   the  equivalent   line   type   and   its  length  should   be   N   x   S f,  where   N   is  the  number   of  floats   and   Sf  is  the  float   pitch.   Note  that  this  length  is  a  little  more  than  the  length  between  the  start  of  the  first  float  and  the  end  of  the  last   one,   since   each   float   is   effectively   being   smeared   equally   both   ways   from   its   centre;   see   the   diagram   below,   which  show  the  situation  when  N=3.  

We   describe   below   how   the   Line   Type   Wizard   derives   the   properties   of   the   equivalent   line   type.   Note   that   this   approach  is  also  suitable  for  modelling  a  regularly  weighted  section  of  line.   Warning:  

The  values  generated  by  the  Wizard  are  based  on  current  best  practice,  but  more  specific  project   data  should  be  used  where  this  is  available.  

Floats

Dp

Sf

Sf

Df

Sf  

Figure:  

Geometry  of  Line  and  Floats  

We  first  define  the  notation  to  represent  the  underlying  line  onto  which  the  floats  are  to  be  attached,  which  we  refer   to  as  the  Base  Line  Type.  We  then  specify  the  quantities  required  to  represent  the  floats.   The  following  properties  are  all  deemed  to  be  unaffected  by  the  addition  of  floats  to  the  base  line  and  so  are  set  to   have  the  same  values  as  those  of  the  base  line.  

325  

w

 

System  Modelling:  Data  and  Results,  Lines    

x

Axial  and  Bending  Stiffness  

x

Limit  Compression  

x

Allowable  Tension  

x

Minimum  Bend  Radius  

x

Torsional  Stiffness  

Data   Adding   floats   to   a   line   to   produce   extra   buoyancy   is   a   common   requirement.   The   Line   Type   Wizard   helps   you   to   quickly  derive  such  a  line  type  by  specifying  both  the  existing  underlying  base  line  type,  onto  which  the  floats  will  be   added,  and  various  properties  of  the  floats:   Base  Line  Type  

The  line  type  on  which  the  floats  are  mounted.   Float  Diameter  

The  outside  diameter  of  each  float.  It  must  be  greater  than  the  outside  diameter  of  the  underlying  base  line  type.   Float  Length  

The  axial  length  of  each  float.   Float  Pitch  

The  average  distance  between  the  centres  of  successive  floats.   Float  Material  Density  

The  density  of  the  material  forming  the  floats,  excluding  additional  items  such  as  fixing  material.   Float  Hardware  Mass  

This  accounts  for  the  extra  mass  due  to  the  addition  of  the  floats  above  that  due  to  the  material  density  and  covers   such  items  as  the  clamping/fixing  mechanisms.   Float  Normal  Drag  Coefficient  

The  drag  coefficient  associated  with  the  float  for  flow  normal  to  the  line.   Float  Axial  Skin  Drag  Coefficient  

The  drag  coefficient  associated  with  the  floats,  due  to  the  floats'  skin  friction,  for  flow  along  the  axis  of  the  line.   Float  Axial  Form  Drag  Coefficient  

The  drag  coefficient  associated  with  the  float,  due  to  the  projected  annulus  area  of  the  end  of  the  float,  for  flow  along   the  axis  of  the  line.   Float  Normal  Added  Mass  Coefficient  

The  added  mass  coefficient  for  flow  normal  to  the  line.   Float  Axial  Added  Mass  Coefficient  

The  added  mass  coefficient  for  flow  along  the  axis  of  the  line.   The  Line  Type  data  that  are  derived,  and  the  associated  underlying  expressions,  are  detailed  in  Modelling  Lines  with   Floats.  

Properties  of  Base  Line  Type   For  modelling  lines  with  floats  the  line  without  floats  is  referred  to  as  the  base  line  type  and  the  following  notation  is   used.  The  line  without  floats  is  assumed  to  be  of  circular  cross-­‐section  and  have  the  following  characteristics:   x

ODp  Ȃ  outer  diameter.  

x

IDp  Ȃ  inner  diameter.  

x

Mp  Ȃ  mass  per  unit  length.  

326  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

x

Dnp  Ȃ  normal  drag  diameter.  

x

Dap  Ȃ  axial  drag  diameter.  

x

Cdnp  Ȃ  drag  coefficient  in  normal  flow.  

x

Cdap  Ȃ  drag  coefficient  in  axial  flow.  

x

Canp  Ȃ  Added  mass  coefficient  in  Normal  flow  (commonly  taken  as  1.0  for  circular  section).  

x

Caap  Ȃ  Added  Mass  coefficient  in  Axial  flow  (commonly  taken  as  zero).  

Properties  of  the  Floats   For   modelling   lines   with   floats   the   following   notation   is   used   for   the   floats.   The   floats   are   assumed   to   be   short   cylinders  fitted  co-­‐axially  on  the  line  at  constant  spacing:   Lf  

length  

Df  

diameter  

ɏf  

float  density  

Sf  

float  pitch  

mfh  

float  hardware  mass  (e.g.  fixing  clamps,  bolts,  etc.)  

Cdnf  

drag  coefficient,  normal  flow  

Cdaf1   drag  coefficient,  axial  flow  due  to  form   Cdaf2   drag  coefficient,  axial  flow  due  to  skin  friction   Canf  

added  mass  coefficients  in  normal  flow  

Caaf  

added  mass  coefficient  in  axial  flow  

With  the  above  information  we  can  calculate  the  volume  occupied  by  an  individual  float  as:   Vf  αɎȀͶȋf2  -­‐  ODp2)  Lf   which  leads  to  the  mass  of  the  float  being  calculated  as  follows:   Mf  =  VfǤɏf  +  mfh.  

Outer,  Inner  and  Contact  Diameter   The  Line  Type  Wizard  sets  up  diameters  for  a  Line  with  Floats  as  follows:   Outer  Diameter  

The   Outer   Diameter   (OD)   of   the   equivalent   line   is   calculated   by   equating   two   equivalent   expressions   for   the   volume   per  unit  length  of  the  line:   ‘Ž’‡”—‹–Ž‡‰–ŠαɎȀͶǤD2  (equivalent  line)   ‘Ž’‡”—‹–Ž‡‰–ŠȋȌαɎȀͶǤp2  +  Vf  /Sf  (line  with  floats)   This  leads  to:   —–‡”‹ƒ‡–‡”ȋȌαȋͶȀɎȌ½   Inner  Diameter  

The  Inner  Diameter  is  unaffected  by  the  addition  of  floats  and  so  is  set  to  be  the  same  as  that  of  the  base  line.   Contact  Diameter  

The  Contact  Diameter  is  set  to  equal  the  float  diameter  Df.  This  allows  float  clearance  to  be  reported  using  the  Line   Contact  Clearance  result.  

Mass  per  Unit  Length   The  line  type  mass  per  unit  length  is  calculated  by  allowing  for  the  fact  that  there  is  one  float  for  every  S f  length  of   the  section  and  hence  (1/Sf)  floats  per  unit  length,  giving:   Mass  per  unit  length  =  Mp  +  Mf  /  Sf  

327  

w

 

System  Modelling:  Data  and  Results,  Lines    

Normal  Drag  Coefficient   The  Line  Type  Wizard  sets  up  the  Normal  Drag  Coefficient  for  a  Line  with  Floats  as  follows:   The  drag  force  per  unit  length  of  the  equivalent  line  when  flow  is  normal  to  the  line's  axis  can  be  expressed  as:   Drag  Forcen  αΦɏ˜2  Cdn  Dnp   in   which   the   reference   drag   area   per   unit   length,   normal   to   the   flow,   is   given   by   the   normal   drag   diameter   of   the   base  line,  Dnpǡƒ†™Š‡”‡ɏ‹•–Š‡†‡•‹–›‘ˆ•‡ƒ™ƒ–‡”ƒ†˜‹•–Š‡ˆŽ‘™˜‡Ž‘…‹–›Ǥ   We   can   also  express   the   drag   force   per   unit   length  experienced  by  the  equivalent   line   as   the   sum   of   the   drag   forces   experienced  by  the  floats  and  the  drag  forces  experienced  by  the  part  of  the  line  not  hidden  by  the  floats:   Drag  Forcen  =  Drag  Forcen-­‐FLOATS  +  Drag  Forcen-­‐EXP  LINE   αΦɏ˜2  [Cdnf.Drag  Arean-­‐FLOATS  +  Cdnp.Drag  Arean-­‐EXP  LINE]   in  which  the  reference  drag  area  for  the  floats  in  normal  flow  is  given  by:   Drag  Arean-­‐FLOATS  =  Df  Lf/Sf   and  the  reference  drag  area  for  the  exposed  line  in  normal  flow  is   given  by:   Drag  Arean-­‐EXP  LINE  =  Dnp  (Sf-­‐Lf)/Sf.   Equating  the  two  expressions  for  drag  force  leads  to:   Cdn  =  [Cdnf.Drag  Arean-­‐FLOATS  +  Cdnp.Drag  Arean-­‐EXP  LINE]  /  Dnp.   In   case   the   base   line   type   uses   drag   which   varies   with   Reynold's   number,   a   variable   data   table   is   created   which   specifies   the   drag   variation   with   Reynold's   number   for   the   equivalent   line.   The   formula   above   is   then   applied   to   the   drag  coefficients  in  each  row  of  the  table.  

Axial  Drag  Coefficient   The  Line  Type  Wizard  sets  up  the  Axial  Drag  Coefficient  for  a  Line  with  Floats  as  follows.   To   derive   the   drag   coefficient   when   flow   is   axial   to   the   line   we   adopt   a   similar   approach   to   that   used   above   for   normal  flow.   When   considering   the   equivalent   line,   with   the   additional   buoyancy   smeared   along   it's   outer   surface,   the   drag   force   per  unit  length,  when  flow  is  axial  to  the  line,  is  due  solely  to  skin  friction  and  can  be  expressed  as:   Drag  Forcea  αΦɏ˜2  Cda  ȋɎap)   in  which  the  reference  drag  area  per  unit  length  is  the  circumference  of  the  base  line  (calculated  u sing  the  axial  drag   diameter  Dap)  and  where  r  is  the  density  of  seawater  and  v  is  the  flow  velocity.   As   in   the   case   for   flow   normal   to   the   line,   we   can   also   express   the   drag   force   per   unit   length   experienced   by   the   equivalent  line  as  the  sum  of  the  drag  forces  experienced  by  the  floats  and  the  drag  forces  experienced  by  the  part  of   the  line  not  hidden  by  the  floats.  However,  the  drag  forces  experienced  by  the  floats  are  slightly  more  complicated  in   axial  flow  as  there  will  be  a  drag  force  due  to  the  exposed  annulus  on  the  end  of  each  float  and  a  drag  force  due  to   skin  friction.   Drag  Forcea  =  Drag  Forcea-­‐FLOATS  +  Drag  Forcea-­‐EXP  LINE   αΦɏ˜2  [Cdaf1.Drag  Area1a-­‐FLOATS  +  Cdaf2.Drag  Area2a-­‐FLOATS  +  Cdap.Drag  Areaa-­‐EXP  LINE]   in  which  the  reference  drag  area,  due  to  the  annulus,  for  the  floats  in  axial  flow  is  given  by:   Drag  Area1a-­‐FLOATS  αɎȀͶȋf2-­‐ODp2)/Sf   the  reference  drag  area,  due  to  the  skin,  for  the  floats  in  axial  flow  is  given  by:   Drag  Area2a-­‐FLOATS  αɎfLf/Sf   and  the  reference  drag  area,  due  to  the  skin,  for  the  exposed  line  in  axial  flow  is  given  by:   Drag  Areaa-­‐EXP  LINE  αɎap  (Sf-­‐Lf)/Sf.   Equating  these  two  expressions  leads  to:   Cda  =  [Cdaf1.Drag  Area1a-­‐FLOATS  +  Cdaf2.Drag  Area2a-­‐FLOATS  +  Cdap.Drag  Areaa-­‐EXP  LINEȐȀȋɎap).  

328  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Added  Mass  Coefficients   The  Line  Type  Wizard  sets  up  Normal  and  Axial  Added  Mass  Coefficients  for  a  Line  with  Floats  as  follows:   Normal  Added  Mass  Coefficient  

Added  mass   coefficients   are   calculated   in   a   similar  way  to   the  drag   force   coefficients.   For   flow  normal   to   the   axis   of   the  line  the  added  mass  per  unit  length  is  given  by:   Added  Massn  αɏɎȀͶ2  Can   ‹™Š‹…Š–Š‡”‡ˆ‡”‡…‡˜‘Ž—‡‹•–Š‡˜‘Ž—‡‘ˆ–Š‡‡“—‹˜ƒŽ‡–Ž‹‡ƒ†™Š‡”‡ɏ‹•–Š‡†‡•‹–›‘ˆ•‡ƒ™ƒ–‡”Ǥ   We  can  also  express  the  added  mass  term  of  the  equivalent  line  as  the  sum  of  the  added  masses  due  to  the  floats  and   due  to  the  underlying  line:   Added  Massn  αɏȋƒnf  AMVolFLOATS  +  Canp  AMVolEXP  LINE)   in  which  the  reference  volume  per  unit  length  for  the  floats  (and  the  portion  of  line  they  cover)  is  given  by:   AMVolFLOATS  αɎȀͶf2  Lf/Sf   and  the  reference  volume  per  unit  length  for  the  exposed  part  of  the  line  is  given  by:   AMVolEXP  LINE  αɎȀͶp2  (Sf-­‐Lf)/Sf   Equating  the  expressions  for  added  mass  leads  to:   Can  =  (Canf  AMVolFLOATS  +  Canp  AMVolEXP  LINEȌȀȋɎȀͶ2).   Axial  Added  Mass  Coefficient  

The   added   mass   coefficients   follow   in   a   similar   way  to   above.   The   reference   volumes   for   the  equivalent   line   and   for   the   floats   and   exposed   part   of   the   underlying   base   line   are   taken   to   be   the   same   in   axial   flow   as   in   normal   flow.   Hence,  we  can  take  the  above  expression  for  the  added  mass  coefficient  in  normal  flow  and  replace  the  coefficients   for  normal  flow  with  those  for  axial  flow:   Caa  =  (Caaf  AMVolFLOATS  +  Caap  AMVolEXP  LINEȌȀȋɎȀͶ2).  

Stress  Diameters  and  Allowable  Stress   The  stress  diameter  and  allowable  stress  are  set  to  be  the  values  used  by  the  base  line,  since  it  is  the  base  line  which   is  load  bearing.  

6.8.13

Homogeneous  Pipe   O

N



2¶   Figure:  

Homogeneous  Pipe  

The   Line   Type   Wizard   for   Homogeneous   Pipe   creates   data   for   a   General   Category   Line   Type   with   properties   appropriate  to  a  pipe  constructed  from  a  single  homogeneous  material,  for  example  a  metal  riser.  

329  

w

 

System  Modelling:  Data  and  Results,  Lines    

Note:  

If   you   are  modelling   a   homogeneous   pipe   then  it   is  normally   better   to   do   so  using   a   Homogeneous   Pipe  Category  Line  Type.  This  allows  you  to  specify  material  density,  Young's  modulus  etc.  directly.  

The  derived  properties  are  based  on  the  following  input  data:   ɏ  

material  density  

E  

Young's  Modulus  

ɋ  

Poisson  Ratio  

OD  

Outer  Diameter  

ID  

Inner  Diameter  =  OD-­‐2t  where  t  is  the  wall  thickness  

The  properties  of  the  derived  equivalent  line  type  are  given  below.   Mass  per  Unit  Length  

ƒ••’‡”—‹–Ž‡‰–ŠαɏȋɎȀͶȌȋ2  -­‐  ID2Ȍ™Š‡”‡ɏ‹•–Š‡material  density  specified.   Outer  and  Inner  Diameters  

The  line  type  outer  and  inner  diameters  are  set  to  the  pipe  diameters  specified  by  the  user.   Axial  Stiffness  

The  line  type  axial  stiffness  is  given  by:   Axial  Stiffness  =  EA   where  E  is  the  Young's  Modulus  and  A  is  the  cross  sectional  area,  hence:   š‹ƒŽ–‹ˆˆ‡••αȋɎȀͶȌȋ2  -­‐  ID2).   Bending  Stiffness  

The  line  type  bending  stiffness  is  given  by:   Bending  Stiffness  =  EI   where  I  is  the  second  moment  of  area,  about  an  axis  in  the  plane  of  the  cross-­‐section  through  the  centroid  (e.g.  NN'),   and  leads  to:   ‡†‹‰–‹ˆˆ‡••αȋɎȀ͸ͶȌȋ4  -­‐  ID4).   Limit  Compression  

As  the  bending  stiffness  is  significant  this  is  set  to  'no'.   Torsional  Stiffness  

The   line   type   torsional   stiffness   is   set   as   follows.   The   torque   experienced   by   a   pipe   of  length   l   when   twisted   through   ƒƒ‰Ž‡Ʌ‹•‰‹˜‡„›ǣ   ‘”“—‡αȋ ɅȀŽȌ   where   J  is  the  second  moment  of  area  about  the  axial  axis  OO'  (often  called  the  polar  moment  of  inertia)  and  G  is  the   Shear  Modulus  (sometimes  called  the  modulus  of  rigidity).  For  homogeneous  pipes  J  =  2I.  The  quantity  G  is  related  to   –Š‡‘—‰̵•‘†—Ž—•ȋȌƒ†‘‹••‘ƒ–‹‘ȋɋȌ‘ˆ–Š‡ƒ–‡”‹ƒŽ–Š”‘—‰Š–Š‡ˆ‘ŽŽ‘™‹‰”‡Žƒ–‹‘•Š‹’ǣ  

αȀȓʹȋͳΪɋȌȔǤ   The  Torsional  Stiffness,  representing  the  Torque  resisting  a  twist  of  1  radian,  per  unit  length,  is  therefore  given  by:   ‘”•‹‘ƒŽ–‹ˆˆ‡••α

αȀȓʹȋͳΪɋȌȔȋɎȀ͵ʹȌȋ4  -­‐  ID4).   Stress  Outer  and  Inner  Diameters  

The  line  type  stress  diameters  are  set  to  '~',  since  they  are  the  same  as  the  pipe  diameters.   Stress  Loading  Factors  

These  are  set  to  one,  the  default  value,  as  a  simple  homogeneous  pipe  carries  all  the  loads.  

Data   The  Line  Type  Wizard  helps  build  a  line  type  to  represent  a  homogeneous  pipe,  based  on  the  following  data:  

330  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Material  

The   Wizard   provides   3   standard   materials   for   a   homogeneous   pipe:   Steel;   Titanium   and   High   Density   Polyethylene.   For   these   standard   materials   OrcaFlex   automatically   sets   Material   Density,   Young's   Modulus   and   Poisson  Ratio.   There  is  also  an  option  to  enter  User  Specified  as  the  Material.  In  this  case  you  must  set  Material  Density,  Young's   Modulus  and  Poisson  Ratio.   Material  Density  

This  is  the  density  of  the  material  used  in  the  construction  of  the  pipe.   Outer  Diameter,  Wall  Thickness  

These  data  specify  the  geometry  of  the  pipe  cross-­‐section.   Young's  Modulus  

The  ratio  of  the  tensile  stress  to  the  tensile  strain.   Poisson  Ratio  

The   amount   of   lateral   strain   experienced   by   a   material   subjected   to   tensile   strain   as   a   negative   proportion   of   the   tensile  strain.   The   Line   Type   data   that   are   derived,   and   the   associated   underlying   expressions,   are   detailed   in   Modelling   Homogeneous  Pipes.  

6.8.14

Hoses  and  Umbilicals  

The  Line  Type  Wizard  estimates  typical  properties  for  hoses  and  umbilicals  based  on  project  data.   Warning:  

The  values  generated  by  the  Wizard  are  approximate  only  and  are  intended  as  first  estimates  for   preliminary   use.   They   are   offered   in   good   faith,   but   due   to   variations   in   properties   between   products  they  cannot  be  guaranteed.  Please  use  suppliers'  data  where  this  is  available.  

There  are  three  categories  of  hose  available:   x

High  pressure   which  covers  high  pressure  flexible  risers  and  flowlines  of  unbonded  construction  with  inside   diameters  in  the  range  2  to  15  inches  (50  to  380mm).  

x

Low   pressure   which   covers   low   pressure   floating   hoses   of   bonded   rubber   construction   with   inside   diameter   from  2  to  20  inches  (50  to  500  mm).  

x

Fold-­‐flat  which  covers  low  pressure,  fold-­‐flat  hoses  with  steel  reinforcement;  inside  diameter  around  6  inches   (150  mm).  

The  umbilical  properties  assume  construction  with  steel  wire  armour  and  thermoplastic  hoses  and  outside  d iameter   up  to  250mm.   The   properties  derived  by  the   Wizard  are  obtained  from  empirically  estimated   relationships   with  the  diameter  of   the   hose/cable.   They   have   been   estimated   from   a   limited   amount   of   data   covering   only   the   range   of   diameters   indicated  above.  For  simplicity,  only  those  relationships  of  the  form:   Y  =  aXb,   where  b  is  an  integer,  were  considered.    In  the  details  below  the  diameter  is  assumed  to  be  in  metres  and  the  SI  units  system  is  applied  throughout.   The  amount  of  data  available  for  low  pressure  hoses  and  fold-­‐flat  hoses  is  very  small.   There   is  quite   a   bit   more   data   for   high   pressure   hoses   and   umbilicals   but   it   is   found   to   have   quite   a   large   spread.   To   demonstrate   this   spread,   the   ratio   of   the  observed  value   to   the   fitted   value,  expressed   as   a   percentage,   is  calculated   and  the  largest  and  smallest  of  these  is  given.   The   OrcaFlex   stress   analysis   is   not   applicable   to   complex   structures   such   as   hoses   and   umbilicals.   Any   available   stress  or  wall  tension  results  should  therefore  be  ignored.  

331  

w

 

System  Modelling:  Data  and  Results,  Lines    

Data  for  Umbilicals   The   Line   Type   Wizard   can   help   build   a   line   type   to   represent   umbilical.   Umbilical   cables   have   many   applications   including  the  carrying  of  electrical  communication  wires  and  hydraulic  connectors  to  submersibles.  The  Line  Type   data  quantities  that  the  wizard  derives  have  been  estimated  from  a  limited  amount  of  project  data.  The  single  item   of  input  data  is:   Umbilical  Diameter  

The  outer  diameter  of  the  umbilical.  Each  derived  line  type  property  is  a  function  of  this  diameter.  

Data  for  Hoses   The  Line   Type  Wizard  helps   you  build  a   line  type  to   represent  a  hose,   based  on  the  following  data.  A  limited  amount   of  available  project  data  has  been  collated  and  used  to  derive  purely  empirical  relationships  between  the  diameter   of  types  of  hose  and  certain  line  type  data  quantities.  The  input  data  consists  of:   Hose  Inner  Diameter  

Each  derived  line  type  property  is  a  function  of  the  hose  inner  diameter.   Hose  Type  

The  Hose  Type  can  be  one  of  high  pressure,  low  pressure  or  fold-­‐flat.  These  categories  roughly  cover  the  available   project  data.  

Outer  and  Inner  Diameters   The  Line  Type  Wizard  sets  up  Outer  and  Inner  Diameters  for  hoses  and  umbilicals  as  follows:   Hoses  

The  inner  diameter  (ID)  is  specified  by  the  user  and  the  outer  diameter  (OD)  is  a  function  of  the  inner  diameter:   OD  =  1.40  ID  m  (for  High  Pressure)  [90%  150%],   OD  =  1.28  ID  m  (for  Low  Pressure),   OD  =  1.34  ID  m  (for  Fold-­‐Flat).   Umbilicals  

The  inner  diameter  (ID)  is  set  to  zero  and  the  outer  diameter  (OD)  is  specified  by  the  user.  

Mass  per  unit  length   The  Line  Type  Wizard  sets  up  mass  for  hoses  and  umbilicals  as  follows:   Hoses  

For  each  type  of  hose  the  mass  per  metre  has  been  estimated  as  a  function  of  inner  diameter  giving:   Mass  per  metre  =  0.7523  ID  te/m  (for  High  Pressure)  [55%  145%],   Mass  per  metre  =  0.3642  ID  te/m  (for  Low  Pressure),   Mass  per  metre  =  0.1844  ID  te/m  (for  Fold-­‐Flat).   Umbilicals  

For  the  umbilicals  the  mass  per  metre  has  been  estimated  as  a  function  of  outer  diameter  giving:   Mass  per  metre  =  1.8  OD2  te/m  (for  Umbilical)  [35%  170%].  

Axial  and  Bending  Stiffness   The   Line   Type   Wizard   sets   up   Axial   and   Bending   Stiffness   and   Limit   Compression   for   hoses   and   umbilicals   as   follows:   Axial  Stiffness  

For  each  type  of  hose  the  axial  stiffness  has  been  estimated  as  a  function  of  inner  diameter  giving:   Axial  Stiffness  =  2.80  x  106  ID  kN  (for  High  Pressure)  [40%  160%],  

332  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

Axial  Stiffness  =  3.40  x  104  ID  kN  (for  Low  Pressure),   Axial  Stiffness  =  6.56  x  103  ID  kN  (for  Fold-­‐Flat).   For  the  umbilicals  the  axial  stiffness  has  been  estimated  as  a  function  of  outer  d iameter  giving:   Axial  Stiffness  =  1.44  x  106  OD  kN  (for  Umbilical)  [15%  415%].   Bending  Stiffness  

For  each  type  of  hose  the  bending  stiffness  has  been  estimated  as  a  function  of  inner  diameter  giving:   Bending  Stiffness  =  3  x  104  ID4  kN.m2  (for  High  Pressure)  [45%  300%],   Bending  Stiffness  =  6  x  102  ID3  kN.m2  (for  Low  Pressure),   Bending  Stiffness  =  1  x  103  ID3  kN.m2  (for  Fold-­‐Flat).   For  the  umbilicals  the  bending  stiffness  has  been  estimated  as  a  function  of  outer  diameter  giving:   Bending  Stiffness  =  3  x  103  OD3  kN.m2  (for  Umbilical)  [55%  240%].   Limit  Compression  

As  the  bending  stiffness  is  significant  this  is  set  to  'no'.  

6.8.15

Modelling  Stress  Joints  

A   tapered   stress   joint   is   a   riser   joint   with   a   tapered   cross-­‐section.   Stress   joints   are   used   to   spread   loads   over   a   critical  section  of  the  riser.  Stress  joints  are  typically  fabricated  from  steel  or  titanium.   Stress   joints   are   modelled   in   OrcaFlex   using   profiled   line   types.   To   model   a   stress   joints   you   need   to   take   the   following  steps:   1.

Create   a   line   type   with   the   homogeneous   pipe   category   and   set   the   inner   diameter,   material   data   (Young's   modulus,  Poisson  ratio,  material  density),  hydrodynamic  properties  etc.  

2.

Create  a  Line  Type  Outer  Diameter  variable  data  source  which  defines  the  stress  joint  profile.  

3.

Set  the  line  type  outer  diameter  to  refer  to  the  variable  data  source  of  the  profile.  

4.

Set   the  Line   Type   data   item  on   the   structure   page   of   the  line   data   form   to   refer   to   the   p rofiled   line   type   created   above.  

Segmentation  

OrcaFlex  line  segments  are  straight,  that  is   the  diameter  is  constant  over  the  length  of  the  segment.  Each  segment   in   a  profile  line  section  has  a  diameter  defined  by  evaluating  the  profile  data  at  the  mid-­‐point  of  the  segments.   Because  of  this  you  typically  need  to  use  quite  short  segments  to  model  the  taper  accurately.  However,  note  that  a   stress   joint   is   usually   placed   in   a   region   of   concentrated   stress.   Such   regions   also   require   short   segments   for   accurate   modelling   and   so   the   use   of   straight   segments   to   model   the   profile   does   not   in   practice   turn   out   to   be   a   significant  limitation.   In  order  to  determine  segment  length  we  recommend  that  you  perform  sensitivity  studies  on  segment  length.   Arc  length  convention  

Profile   arc   length   is   defined   relative   to   the   start   of   the   line   section   and   increases   from   End   A   towards   End   B.   To   illustrate  this  consider  a  stress  joint  defined  by  the  following  profile:  

333  

w

 

System  Modelling:  Data  and  Results,  Lines    

  Figure:  

Stress  joint  profile  

A  stress  joint  specified  this  way  would  commonly  be  used  in  the  first  section  of  a  line.  If,  however,  your  stress  joint  is   located   adjacent   to   End   B   of   the   line,   then   the   End   A   to   End   B   convention   means   that   the   stress   joint   would   be   incorrectly  configured.  This  is  easy  to  check  with  the  profile  graph  available  on  the  line  data  form:  

  Figure:  

Stress  joint  profile  at  End  B,  incorrectly  modelled  

The  problem  is  that  the  taper  is  now  in  the  wrong  direction.  The  thicker  end  of  the  taper  should  be  adjacent  to  End  B   of  the  line.  In  order  to  fix  this  we  simply  need  to  reverse  the  profile  data.  This  is  very  simple  to  do  using  the  Reverse   button  on  the  variable  data  form.  The  result  looks  like  this:  

  Figure:  

Stress  joint  profile  at  End  B,  corrected  

The  line  profile  graph  when  using  the  reversed  profile  now  shows  that  the  data  is  now  applied  as  intended:  

334  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

  Figure:  

6.8.16

Stress  joint  profile  at  End  B,  correctly  modelled  

Modelling  Bend  Restrictors  

We  begin  by  introducing  some  terminology.  A   bend  restrictor  is  any  device  that  controls,  restricts  or  limits  bending   on   a   line.   A   bend   limiter   is   a   bend   restrictor   that   has   no   effect   until   a   certain   curvature   is   reached,   and   then   curvature  is  prevented  from  going  above  that  value.  A   bend  stiffener  is  a  bend  restrictor  that  provides  increased   bend  stiffness  in  order  to  distribute  more  widely  the  bending.   Modelling  Bend  Limiters   Non-­‐linear   bend   stiffness   can   be   used   to   model   a   bend   limiter.   The   approach   is   to   specify   a   relationship   between   curvature  and  bend  moment  that  has:   x

Low  stiffness  for  curvature  values  lower  than  the  lock-­‐out  curvature.  

x

High  stiffness  for  curvature  values  greater  than  the  lock-­‐out  curvature.  

Typically   the   low   stiffness   value   will   be   close   to   zero   and   the   high   stiffness   value   will   be   one   or   two   orders   of   magnitude   greater   than   the   stiffness   of   the   protected   line.   Try   to   avoid   using   too   large   a   value   since   doing   so   can   result  in  numerical  instability.  It  may  also  help  to  smooth  the  transition  from  low  to  high  stiffness.   The   most   common  modelling   approach  for  bend   limiters   uses   a   single  equivalent   line   type   object   to  represent  both   the   protected   line   and   the   limiter.   The   bend   stiffness   for   this   equivalent   line   type   must   account   for   both   the   protected  line  and  the  limiter.  You  may  choose  also  to  account  for  mass,  displacement  and  hydrodynamic  properties   but  often  these  properties  are  of  lesser  importance.   An  alternative  to  the  equivalent  line  approach  is  to  model  the  limiter  as  a  separate  object  using  the  bend  stiffener   attachment  (see  below).  The  main  difference  from  an  elastomeric  stiffener  is  that  a  general  category   line  type  with   non-­‐linear  bend  stiffness  must  be  used  for  the  attachment  line  type.  The  main  advantage  of  this  approach  is  that  it   becomes   easier   to   check   that   the   data   is   specified   correctly   because   you   can   keep   the   data   for   the   protected   line   separate  from  the  data  for  the  limiter.   Modelling  Bend  Stiffeners   Bend  stiffeners  are  modelled  in  OrcaFlex  using  two  separate  lines  to  represent  the   stiffener  and  the  line   which  it   protects,   which  we   refer   to   as   the  protected   line.   The   region   of   the   protected   line  which  is  covered   by  the   stiffener   is  called  the  protected  region.   The   two   line   approach   enables   reporting   of   separate   results   for   the   protected   line   and   stiffener.   In   particular   this   makes  fatigue  analysis  of  the  protected  line   quite  simple  since  the  reported  loads  and  stresses   for  the  protected  line   do  not  include  the  contributions  of  the  stiffener.   The  protected  line  can  have  linear,   non-­‐linear   elastic  or   hysteretic  bending  properties.  The  stiffener  is  modelled   as  a   profiled  homogeneous  pipe.  The  stiffener  can  have  linear  or  non-­‐linear  elastic  material  properties.  

335  

w

 

System  Modelling:  Data  and  Results,  Lines    

Data  

Although   the   stiffener   is   modelled   as   a   separate   line   you   do   not   need   to   create   this   line   manually   Ȃ   OrcaFlex   creates   it  automatically  as  an  attachment.  The  procedure  for  setting  up  a  bend  stiffener  is  as  follows:   1.

Create   a  Line   Type  which  defines   the   material,   structural   and   hydrodynamic   properties   of   the   stiffener.   Usually   this  will  be  a  profiled  homogeneous  pipe.  

2.

Create  a  Stiffener  Type  which  uses  this  Line  Type.  

3.

Create  a  line  attachment  based  on  this  Stiffener  Type.  

4.

Set   the  line   attachment   position  and   the  Stiffener   Type   connection   arc   length  so   that   the   stiffener   is  attached   at   the  desired  location  on  the  protected  line.  

If   you   have   multiple   protected   lines   which   all   use   identical   bend   stiffeners   then   you   can   create   a   single   Stiffener   Type  which  can  be  re-­‐used  on  each  protected  line.   The  stiffener  profile  uses  the  convention  that   profile  arc  length  increases  from  End   A  towards  End  B  of  the  stiffener.   If  you  have  a  bend  stiffener  connected  at  End  B  of  a   line  then  you  will  need  to  define  the  profile  so  that  the  arc  length   0   refers   to   the   tip   of   the   stiffener.   The   Modelling   Stress   Joints   topic   illustrates   this   issue   in   some   more   detail.   Although   the   discussion   there   centres   on   stress   joints   many   of   the   points   covered   are   equally   applicable   to   bend   stiffeners.   We   strongly   recommend   that   you   use   the  Profile   Graph   available   from   the   Line   Data   form   to   check   that   the   stiffener   is  connected  at  the  correct  location  on  the  line  with  the  profile  defined  as  you  intended.   Segmentation  

The   stiffener   line   that   OrcaFlex   creates   is   modelled   with   constant   segment   length   Ȃ   that   is   every   segment   in   the   stiffener  has  the  same  length.  The  segment  length  is  determined  by  the  segment  length  of  the  protected  line  in  the   protected  region.   The  stiffener  modelling  (see  below)   requires  that   each   node  on  the  stiffener  line   is  associated   with  a   node  on   the   protected  line.  Each  stiffener  node  is  effectively  clamped  to  its  associated  protected  node.   These  constraints  have  the  following  implications  for  the  segmentation  of  the  protected  line:   1.

The  protected  region  must  have  constant  segment  length.  

2.

The  stiffener  length  must  be  an  exact  multiple  of  the  segment  length.  

One   simple  way  to   satisfy  these  requirements   is  to   model   the   protected  region   as   a   single  section  with  length  equal   to  the  stiffener  length.   Note   that   it   is   not   essential   for   the   protected   region   to   be   a   single   section.   The   protected   region   could   comprise   multiple  sections  each  using  different  line  types,  so  long  as  you  satisfy  the  two  rules  above.   Drawing  and  Results  

The  stiffener  line  is  drawn  using  the   drawing  data  of  the  protected  line  to  which  it  is  attached.  Note  that  the  stiffener   is  not  drawn  when  the  program  is  in  reset  state;  it  is  only  drawn  after  the  static  or  dynamic  analysis  has  started.   Results  are  available  for  the  stiffener  line  exactly  as  they  are  for  any  other  OrcaFlex  line.   OrcaFlex   reports   results   separately   for   protected   line   and   stiffener   line   and   this   does   need   some   explanation.   For   example,  consider  bend  moment  at  a  particular  location  in  the  protected  line  and  at  the  corresponding  location  in   the  stiffener  line.  Suppose  that  the   bending  stiffnesses   are  EI p  and  EIs   for  protected  line  and  stiffener  respectively   (we   are   assuming   linear   bend   stiffness   for   simplicity).   The   bend   moment   carried   by  the   protected   line   and   stiffener   ensemble  is  given  by  BMtotal  =  C(EIp  +  EIs)  where  C  is  the  curvature  at  this  location.  For  the  protected  line  OrcaFlex   reports  the   local   protected   line   bend   moment   BM p  =   C.EIp  and   likewise  for   the   stiffener   line   OrcaFlex   reports   BM s  =   C.EIs.  It  is  straightforward  to  see  that  BMtotal  =  BMp  +  BMs.   The  total  load  is  also  split  into  separate  protected  line  and  stiffener  loads  for   effective  tension,  wall  tension,  shear   force,  torque  and  stress   results.   However,   the   method   for   doing   this   varies   for   axial   components   as   explained   in   the   next  section.   Modelling  details  

As  mentioned  above  the  stiffener  is   modelled  as  a  separate  OrcaFlex  line   which  is  created  automatically  by  OrcaFlex   as  an  attachment.  The  stiffener  line  inherits  a  number  of  properties  from  its  protected  line,  namely:  

336  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

x

Include  Torsion.  

x

Segment  length.  

x

Statics  friction  data.  

x

Drag  formulation  and  wake  interference  data.  

x

VIV  data.  

x

Drawing  data.  

x

Results  data.  

The  stiffener  line  does  not  have  any  free  degrees  of  freedom.  Instead  each  node  on  the  stiffener  is  clamped  to  and   moves  and  rotates  with  its  associated  node  on  the  protected  line.  The  stiffener  line  calculates  its  loads  and  inertia   and  then  transfers  them  to  the  protected  line.  How  this  transfer  is  performed  is  governed  by  the   Axial  load/inertia   transfer  data  of  the  Stiffener  Type.   All   components   of   load   and   inertia   normal   to   the   stiffener   are   transferred   directly   from   each   stiffener   node   to   its   associated   protected   node.  This,  of   course,  enables   the  stiffener  to   perform   its   job   of   spreading   the  bend   loads  over   the  protected  region.   If   the   axial   load/inertia   transfer   is   specified   to   occur   at   the  connection   point   then   components   of  axial   load/inertia   are   transferred   to   the   protected   node   at   the   connection   point.   Typically   this   connection   point   is   at   the   end   of   the   protected   line   and   the   axial   loads   and   inertia   are   thus   transferred   to   the   protected   line's   end   connection.   This   modelling  option  effectively  neglects  any  axial  friction  due  to  contact  between  stiffener  and  protected  line.   If   the   axial   load/inertia   transfer   is   specified   to   occur   over   the   stiffener's   full   length   then   components   of   axial   load/inertia  are  transferred  directly  from  each  stiffener  node  to  its  associated  protected  node.  This  corresponds  to   the  assumption  that  the  axial  contact  friction  is  sufficient  that  there  is  no  axial  slipping.  The  axial  load  will  be  shared   between   protected   line   and   stiffener   as   determined   by   their   relative   axial   stiffnesses,   just   as   the   bend   moment   is   shared.   Bend  Stiffener  design  using  OrcaFlex   The  modelling  approach  described  above  applies  where  a  bend  stiffener  has  already  been  designed,  and  one  of  the   objectives  of  the  analysis  is  to  confirm  that  the  stiffener  provides  the  required  protection.  However,  in  many  cases   the  stiffener  design  does  not  yet  exist  and  the  analysis  is  needed  in  order  to  define  design  loads.  If  this  is  the  case,   then   run  a   preliminary   analysis   with   no   bend   stiffener   included.   The   line   should   be   modelled   with  a   pinned   end   (i.e.   zero  bending  stiffness  at  the  line  end  connection).   The  load  information   required  for   bend  stiffener  design  then  consists  of  paired  values  of   tension  and  angle  at   the   pinned  end.  These  can  be  extracted  in  the  form  of  an   X-­‐Y  graph  showing  Effective  Tension  against  Ez  Angle  for  the   first  segment.  In  practice,  it  is  often  sufficient  to  consider  just  three  points  on  this  graph,  corresponding  to  maximum   tension,  maximum  angle  and  maximum  bend  restrictor  load:  these  can  be  extracted  as  linked  statistics.  Recall  that   Ez  Angle  is  an  absolute  magnitude  and  therefore  always  takes  a  positive  value.  If  a  signed  value  is  required  (e.g.  to   define  out-­‐to-­‐out  load  cycles  for  fatigue  analysis),  then  use  the  Ezx  or  Ezy  angle  as  appropriate.   It  is  usually  necessary  to  combine  results  from  several  analysis  runs  in  order  to  fully  define  the  bend  stiffener  design   loading.  This  is  most  conveniently  done  by  exporting  the  Effective   Tension  vs  Ez  Angle  results  as  a  table  of  values  for   each  analysis  case,  combining  into  a  single  Excel  spreadsheet  and  using  the  plotting  facilities  in  Excel  to  generate  a   single   plot   with   all   results   superimposed.   A   simplified   set   of   load   cases   representing   the   overall   loading   envelope   can   then   be   selected   for   use   in   stiffener   design.   The   export   to   Excel   can   be   done   manually   or   automated   through   the   Results  spreadsheet.   Bend  Stiffener  design  using  OrcaBend   The   task   of   bend   stiffener   design   is   usually   left   to   the   manufacturer,   since   the   actual   stiffener   shape   selected   is   governed  in  part  by  the  manufacturing  process,  availability  of  tooling,  etc.,  as  well  as  by  the  load  cases.   The  Orcina   program  OrcaBend  has   been   developed   to   assist  this   process.   There   is   a   demonstration   version   of  OrcaBend  on   the   OrcaFlex  CD  Ȃ  see  CD:\Demo_CD\ReadMe  for  details.  For  further  information  contact  Orcina.  

6.8.17

Modelling  non-­‐linear  homogeneous  pipes  

A  non-­‐linear  stress-­‐strain  relationship  is  most  commonly  used  to  model  either:   x

non-­‐linear  behaviour  of  elastomeric  bend  stiffeners,  or  

337  

w

 

System  Modelling:  Data  and  Results,  Lines    

x

plastic  deformation  of  steel  pipes  during  installation.  

OrcaFlex   can   also   model   non-­‐linear  bend   stiffness   for  General   Category  Line   Types.  However   there   are  a   number   of   advantages  of  using  the  homogeneous  pipe  approach  if  it  is  applicable:   x

The  data  are  entered  directly  and  there  is  no  need  to  use  the   Line  Type  Wizard  for  Homogeneous  Pipe  or  the   Plasticity  Wizard  which  makes  data  checking  simpler.  

x

The  effects  of  a  diameter  profile  and  a   non-­‐linear  material  (e.g.   bend  stiffener)  are  handled  automatically  by  the   program.  

x

The  effects  of  direct  tensile  strain  variation  are  again  handled  automatically  by  the  program  (see  below).  

x

Stress  results  are  derived  using  the  stress-­‐strain  relationship.  

On  the  other  hand  if  you  wish  to  model  hysteretic  bending  behaviour  then  you  must  use  the   General  Category  Line   Types  approach.   Data   Stress-­‐Strain   Relationship   variable   data   sources   are   used   to   define   non-­‐linear   elastic   material   properties   for   homogeneous  pipes.   Stress-­‐Strain  relationship  

The   relationship   between   stress   and   strain   can   be   specified   by   either   Ramberg-­‐Osgood   curve   or   Stress-­‐Strain   table.   P –…

y),  K,  n  (Ramberg-­‐Osgood  curve  only)  

Š‡•‡†ƒ–ƒ†‡ˆ‹‡–Š‡”‡Žƒ–‹‘•Š‹’„‡–™‡‡•–”‡••ȋɐȌƒ†•–”ƒ‹ȋɂȌ‹–‡”•‘ˆƒƒ„‡”‰-­‐Osgood  curve  as  follows:   ɂȋɐȌαɐȀΪȋɐȀɐy)n.   The  reference  stress  parameter  is  usually  ta‡–‘„‡–Š‡›‹‡Ž†•–”‡••ǡ™Š‹…Š‹•™Š›‹–‹•†‡‘–‡†ɐy  here.  Note  that   there  is  an  alternative  parameterisation  of  the  Ramberg-­‐Osgood  equation.  It  is  straightforward  to  convert  between   the   two   forms   of   the   equation   but   please   take   care   to   ensure   that   the   data   you   input   correspond   to   the   parameterisation  used  by  OrcaFlex.   Stress,  Strain  (Stress-­‐Strain  table  only)  

This   table   directly   specifies   the   relationship   between   stress   and   strain.   The   table   is   interpolated   linearly   and   for   values  of  strain  outside  the  table  linear  extrapolation  will  be  used.   Model  building   OrcaFlex  uses  the  stress-­‐strain  relationship   to  generate  a  table  of   bend   moment  against  curvature  using  the  same   algorithm   as   the   Plasticity   Wizard.   Each   segment   in   the   OrcaFlex   model   uses   a   distinct   bend   moment   /   curvature   table  which  is  clearly  necessary  if  the  line  type  has  a  diameter  profile.   The  use  of  distinct  bend  moment  /  curvature  tables  also  allows  OrcaFlex  to  account  for  the  variation  of  direct  tensile   strain  within  a  line.  The  bend  moment  /  curvature  relationship  depends  upon:   x

The  stress-­‐strain  relationship  (as  described  above).  

x

The  inner  and  outer  diameter  (as  specified  in  the  Line  Type  data).  

x

The  direct  tensile  strain.  

The  direct  tensile  strain  can  have  a  significant  effect  on  the  non-­‐linear  bending  behaviour  if  it  is  large.  To  see  why   this  is  so  consider  a  steel  pipe  under  tension  such  that  the  direct  tensile  strain  equals  the  yield  strain.  When  the  pipe   is  in  this  state  then  any  small  amount  of  curvature   will  yield  the   pipe  outer  fibres.   On  the  other  hand  consider  an   unstressed   steel   pipe,   where   the   direct   tensile   strain   is   zero.   In   this   state   the   pipe   can   withstand   significant   curvature  before  the  outer  fibres  yield.   In   principle   the   bend   moment   /   curvature   relationship   could   be   recalculated   at   each   time   step   of   an   OrcaFlex   calculation.   However   this   would   incur   a   significant   performance   cost.   Instead   we   make   the   assumption   that   the   effect  of  dynamic  variation  of  direct  tensile  strain  on  the  bend  moment  /  curvature  relationship  is  small.   This   allows   us   to   use   a   constant   value   of   direct   tensile   strain   for   the   purpose   of   deriving   the   bend   moment   /   curvature   relationship.  Note  that  each  segment  in  the  model  has  a  distinct  bend  moment  /  curvature   relationship   based  on  a  distinct  value  of  direct  tensile  strain.  By  "constant"  we  mean  that  we  do  not  update  the  bend  moment  /  

338  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

curvature   relationship   during   the   dynamic   simulation.   Note   also   that   this   discussion   of   direct   tensile   strain   only   pertains   to   the   generation   of   bend   moment   /   curvature   relationships.   The   program   does,   of   course,   account   for   dynamic  variation  of  direct  tensile  strain  when  calculating  wall  tension,  effective  tension  etc.   The   program   chooses   the   value   of   direct   tensile   strain   by   first   performing   a   static   calculation   under   the   assumption   that  direct  tensile  strain  is  0.  Non-­‐linear  bend  moment  /   curvature  relationships  are  created  for  each  segment  under   this   assumption.   The   direct   tensile   strain   values   from   this   initial   static   solution   are   then   used   to   update   the   non-­‐ linear  bend  moment  /  curvature  relationships.  Finally  the  static  calculation  is  repeated  to  obtain  a  solution  which   accounts  for  the  effects  of  direct  tensile  strain.   Stress  results   The   non-­‐linear   stress-­‐strain   relationship   is   also   used   to   calculate   certain   stress   results   from   strain   values.   Non-­‐ linear   stress-­‐strain   leads   to   a   non-­‐linear   elastic   bend   stiffness,   but   the   axial   and   torsional   stiffnesses   are   still   assumed   to   be   linear.   This   means   that   the   only   stress   results   affected   are   those   that   depend   on   bending:   Max   Bending  Stress,  von  Mises  Stress,  Max  von  Mises  Stress  and  ZZ  Stress.  

6.8.18

Line  Ends  

Lines   in   OrcaFlex   run   from   End   A   to   End   B.   Travelling   from   A   to   B,   the   orientation   of   any   segment   in   the   line   is   defined  in  terms  of  Azimuth  and  Declination  angles,  relative  to  global  axes.  Azimuth  is  measured  in  the  X-­‐Y  plane,   Declination  is  measured  downwards  from  the  Z  axis.  See  No-­‐Moment  Direction.  

No-­‐moment  Direction   Associated   with   each   end   is   a   stiffness,   and   a   no-­‐moment   direction   which   is   described   in   terms   of   azimuth   and   declination.  This  too  uses  the  End  A  to  End  B  convention,  so  if  we  hang  up  a  catenary  of  line,  and  then  freeze  the   ends,  the  no-­‐moment  directions  are  as  shown  below:  

No moment direction ( Az = 0, Dec = 160 )

No moment direction ( Az = 0, Dec = 45 )

End A

End B

z

Declination Angle

y

Azimuth Angle x  

Figure:  

Directions  

If  the  line   end  is  attached  to  a  body   which  can   move  (a  Vessel   or  Buoy),  then  the   no-­‐moment  direction  is  defined   relative  to  the  body  axes  and  therefore  moves  with  the  body.  Otherwise,  it  is  defined  in  global  axes.  

339  

w

 

System  Modelling:  Data  and  Results,  Lines    

End  Stiffness   The   stiffness   associated   with   the   end   can   be   used   to   represent   an   item   such   as   a   flexjoint,   whose   stiffness   is   in   units   of  moment   per   unit  angle,   e.g.   kN.m/degree.   More   commonly,   the   line   end   is  either   free   to   rotate   or   fully   restrained.   In  the  first  case,  the  end  stiffness  is  set  to  zero;  in  the  second  case,  the  end  stiffness  is  set  to  Infinity.  Note  that  it  is   never   necessary   (or   correct)   to   'convert'   the   line   stiffness   into   an   end   stiffness:   the   program   includes   the   line   stiffness  for  you  automatically.  

Free-­‐to-­‐rotate  or  Fully-­‐restrained  Ends   In  many   practical   cases,   the   line   ends   are   neither   completely   free   nor   fully   restrained.   Nevertheless,   we   recommend   that   you   should   usually   choose   one   of   these   conditions.   When   should   you   use   one   rather   than   the   other?   The   following  notes  offer  a  brief  guide:   1.

Many   systems   modelled   using   OrcaFlex   consist  of  relatively   long   flexible   lines   where   bend   stiffness   plays   only   a   minor  role  in  determining  the  overall  forces  on  and  movements  of  the  system.  In  such  systems,  line  ends  may   safely  be  modelled  as  free-­‐to-­‐rotate.  

2.

An  exception  to  this   rule  is   systems  which  include  one   or  more  6D  buoys.  The  rotational  motions  of  the  buoy   may   then   be   influenced   by   moment   transfer   from   the   ends   of   lines   attached   to   it,   particularly   where   buoy   rotational  inertias  are  small.  In  such  cases,  the  end  connections  to  the  buoy  should  be  fully  restrained.  

3.

A   further   exception   is   systems   where   the   flexible   lines   are   relatively   short   and   stiff,   e.g.   a   large   diameter   under-­‐ buoy   hose   in   shallow   water.   Bend   stiffness,   including   end  moments,   may  have   a   significant   influence   on   overall   system  behaviour  in  such  cases,  and  the  end  connections  should  be  fully  restrained.  

4.

Where  fully  restrained  ends  are  used,  it  is  necessary  to  pay  more  attention  to  the  modelling  of  the  line  close  to   the   end.   In   particular   make   allowance   for   the   additional   stiffness   of   a   bend   stiffener,   if   one   is   fitted   and   use   shorter  segments  near  the  line  ends  so  as  to  represent  the  moments  with  sufficient  accuracy.  

5.

Roll-­‐on/roll-­‐off   contact   (e.g.   stern   rollers,   pipelay   stingers,   mid-­‐water   arches   for   riser   systems).   A   pinned   connection   at   the   average   contact   point   is   often   sufficient.   For   a   more   exact   representation,   use   one   or   more   solids  to  represent  the  supporting  surface,  but  remember  that  there  must  be  sufficient  nodes  at  the  line  end  to   interact  with  the  solid.  

End  Force  and  End  Force  Ez-­‐Angle   The   figure   below   shows   the   end   connection   of   a   flexible   line   fitted   with   a   bend   stiffener.   The   line   applies   a   load   (tension)  T  as  shown.  If  the  local  loads  (weight,  drag,  etc.)  on  the  end  part  of  the  line,  including  the  bend  stiffener,   are  small  by  comparison  with  T,  then  the  reaction  force  F  is  equal  and  opposite  to  T,  and  the  bend  moment  at  the   end  fitting  is  M  =  T.h.   OrcaFlex  reports  the  End  Force,  F,  and  the  End  Force  Ez-­‐AngŽ‡ǡɅǡƒ••Š‘™ǤŠ‡̶‘‘‡–†‹”‡…–‹‘̶‹•†‡ˆ‹‡†‹ the  input  data.  When  the  reaction  force  F  acts  in  the  no  moment  direction,  then  the  reaction  moment  M  is  zero.   It  is  clear  from  this  that   1.

End   Force   and   End   Force   Ez-­‐Angle   are   the   same   whether   the   end   condition   is   defined   as   free-­‐to-­‐rotate,   fully   restrained,  or  some  intermediate  condition;  

2.

The   bend   moment   at   the   end   fitting,   M,   is   a   function   of   the   lever   arm,   h,   which   depends   not   only   on   the   end   condition  but  also  on  the  bend  stiffness  distribution  in  the  line/bend  stiffener.  

340  

w

 

System  Modelling:  Data  and  Results,  Lines  

 

No moment direction M F

T

h T   Figure:  

End  connection  of  a  flexible  line  fitted  with  a  Bend  Stiffener  

Design  Loads  for  End  Fittings   For  design  of  end  fittings,  including  bend  restrictors,  the  principal  parameters  provided  by  OrcaFlex  are  End  Force   and  End  Force  Ez-­‐Angle.  The  moment  at  the   end  can   then   be  determined  by  a  local  (static)  analysis  which  can  be   developed  to  incorporate  as  much  detail  as  required.   This  approach  is  usually  sufficient,  except  where  End  Force  is  very  small.  This  occurs  when  the  line   tension   T  comes   close   to  zero.   The   direction   of  the   end   force   is   then   no   longer   dominated   by   the   line   tension,   and   other   loads  (shear,   local  drag  and  inertia  loads  etc.)  which  are  usually  negligible   become  important.  In  these  conditions,  the  reported   End   Force   Ez-­‐Angle   is   misleading   and   a   more   appropriate   estimate   should   be   made   from   the   system   geometry.  This   can  be  done  using  the  Ez-­‐Angle  results  variable.  Ez-­‐Angle  for  any  segment  gives  the  angle  of  that  segment  relative  to   the  No  Moment  Direction  at  the  adjacent  line  end,  including  allowance  for  the  motion  of  line  end  where  the  line  is   attached   to   a   vessel   or   buoy.   Ez-­‐Angle   for   a   point   near   the   end   of   the   bend   restrictor   is   a   reasonable   alternative   where  End  Force  Ez-­‐Angle  is  not  suitable.  

Results   When  examining  results  at  line   ends  note  that  if  a  stiff  pipe  goes  into  compression,  line  tension  becomes  negative   but  End  Force  remains  positive,  and  End  Force  Ez-­‐Angle  may  approach  180°.   Curvature   is   calculated   in   OrcaFlex   by   dividing   the   angle   change   at   any   node   by   the   sum   of  the   half-­‐segment   lengths   on   each   side   of   the   node:   bend   moment   is   curvature   multiplied   by   bend   stiffness.   At   the   end,   OrcaFlex   takes   the   angle   change   between   the   end   segment   of   the   line   and   the   no-­‐moment   direction,   and   reports   the   corresponding   curvature   and   bend   moment   based   on   the   half   length   of   the   end   segment.   Where   bend   stiffness   at   the   line   end   is   zero  (pinned  end  or  a  zero  stiffness  line),  curvature  and  bend  moment  are  reported  as  zero.  

Design  Data  for  Bend  Restrictors   We  classify  bend  restrictors  into  3  types:   x

Bellmouths:  curved  surfaces  which  support  the  flexible  and  maintain  acceptable  curvature.  

x

Bend  Limiters:  articulated  devices  which  rotate  freely  to  a  specified  curvature,  then  stop.  

x

Bend  Stiffeners:  elastomeric  devices  which  provide  a  tapered  additional  bend  stiffness.  

Different  design  information  is  required  for  each  type:   Bellmouth  

The  principal  design  requirement  is  that  bellmouth  angle  should  be  greater  than  the  maximum  value  of  End  Force   Ez-­‐Angle.   For   cases   where   the   bellmouth   is   not   radially   symmetrical,   OrcaFlex   reports   components   of   End   Force  

341  

w

 

System  Modelling:  Data  and  Results,  Lines    

Angle  in  the  local  XZ  and  YZ  planes.  End  Force  Ezx-­‐Angle  is  the  component  in  the  local  xz  plane;  End  Force  Ezy-­‐Angle   is  the  component  in  the  local  yz  plane.   Bend  Limiter  

There  are  two  design  requirements:   1.

The  limiter  length  must  be  not  less  than  a*R  where  a  is  End  Force  Ez-­‐Angle  and  R  is  the  limiter  locking  radius.  

2.

The   limiter   must   be   capable   of  withstanding   the   maximum   bend   moment   M   given   by  M   =   R*F*(1-­‐cos(a))   where   F,  a  are  simultaneous  values  of  End  Force  and  End  Force  Ez-­‐Angle.  OrcaFlex  reports  Bend  Restrictor  Load  P  =   F*(1-­‐cos(a))  as  an  aid  to  bend  limiter  design.  P  is  sometimes  called  "pseudo-­‐curvature".  

Bend  Stiffener  

The  design  process  for  a  stiffener  is  more  complex  than  for  a  bellmouth  or  a  bend  limiter,  and  the  critical  design  load   cases   are   not   always   self-­‐evident.   An   X-­‐Y   graph   of   F   against   a   (End   Force   against   End   Force-­‐Ez   Angle)   provides   a   complete   definition   of   the   loading   for   one   analysis   case,   with  each   (F,a)   pair   defining   a   load   case.   The   bend   stiffener   should   be   designed   to   prevent   infringement   of   the   permitted   curvature   for   any   (F,a)   pair.   In   practice,   it   is   often   sufficient   to   consider   just   the   three   (F,a)   pairs   corresponding   to   maximum   values   of   End   Force   F,   End   Force   Ez-­‐ Angle  a  and  Bend  Restrictor  Load  P.   Once  the  stiffener  design  is  known  the  stiffener  can  be  modelled  using  a  bend  stiffener  attachment.  

6.8.19

Modelling  Compression  in  Flexibles  

When   a   flexible   line   experiences   compression,   it   responds   by   deflecting   transversely:   the   magnitude   of   the   deflection  is  controlled  by  bend  stiffness.  Under  static  conditions,  the  behaviour  of  an  initially  straight  section  of  line   under  pure  axial  loading  is  described  by  classic  Euler  buckling  theory.  This  defines  the  maximum  compressive  load  Ȃ   the  "Euler  load"  Ȃ  which  a  particular  length  of  line  can  withstand  before  transverse  deflection  occurs.  The  Euler  load   is   a   function   of   the   length   of   the   straight   section,   the   bend   stiffness   and   the   end   conditions.   For   a   simple   stick   of   length   L,   bend   stiffness   EI,   with   pin   joints   at   each   end,   the   Euler  Ž‘ƒ† ‹• Ɏ2EI/L2.   The   Euler  load   is   derived   from   a   stability  analysis:  it  tells  us  the  value  of  axial  load  at  which  transverse  deflection   will  occur  but  nothing   about  the   post-­‐buckling  behaviour.   Under   dynamic   loading   conditions,   the   transverse   deflection   is   resisted   by   a   combination   of   inertia   and   bending.   OrcaFlex   is   fully   capable   of   modelling   this   behaviour   provided   the   discretisation   of   the   model   is   sufficient,   i.e.   provided   the   segments   are   short   enough   to   model   the   deflected   shape   properly.   Another   way   of   saying   the   same   thing  is  that  the  compressive  load  in  any  segment  of  the  line  should  never  exceed  the  Euler  load  for  the  segment.   Why   are   these   two   statements   equivalent?   Imagine   the   real   line   replaced   by   a   series   of   rigid   sticks   connected   by   rotational   springs   at   the   joints   Ȃ   this   is   essentially   how   OrcaFlex   models   the   line.   Under   compression,   the   line   deflects:  the  sticks  remain  straight  and  the  joints  rotate.  Provided  the  wavelength  of  the  deflection  is  longer  than  the   length   of   the   individual   sticks   then   the   rigid   stick   model   can   approximate   it:   shorter   sticks   give   a   better   approximation.   If  the  compressive  load  reaches  the  Euler  load  for  an  individual  stick,  then  the  real  line  which  the  stick  represents   will   start   to   deform   at   a   shorter   wavelength,  and   deflections   within   the   stick   length  become   significant.   Clearly,  this   stick  model  is  no  longer  adequate.  By  replacing  each  long  stick  by  several  short  ones,  we  can  make  the  Euler  load  for   each  stick  greater  than  the   applied  compressive  load.  Each  stick  will  then  remain  straight,  but  we  now  have  more   sticks  with  which  to  model  the  deflected  shape.   This   gives   us   a   convenient   way   of   checking   the   adequacy   of   our   model:   provided   the   compressive   load   in   each   segment   always   remains   less   than   the   Euler   load   for  that   segment,   then  we   can   have   confidence   that   the   behaviour   of  the  line  in  compression  is  adequately  modelled.  OrcaFlex  makes  this  comparison  automatically  for  all  segments   and  reports  any  infringements  in  the  Statistics  tables.  The  segment  Euler  load  is  also  plotted  in  tension  range  graphs   (as  a  negative  value  Ȃ  compression  is  negative)  so  that  infringements  are  clearly  visible.   If   the   segment   Euler  load   is   infringed   during   a   simulation,   then   we   have   to   decide   what   to   do   about   it.   If   infringement  occurs  only  during  the  build-­‐up  period,  perhaps  as  a  result  of  a  starting  transient,  then  we  can  safely   ignore   it.  If   it  occurs   during   the   main   part   of  the   simulation,   then   we   should   examine   the   time   histories   of  tension   in   the  affected  areas.  Where  infringements  are  severe  and  repeated  or  of  long  duration  the  analysis  should  be  repeated   with   shorter   segments   in   the   affected   area.   However   it   may   be   acceptable   to   disregard   occasional   minor   infringements  of  short  duration  on  the  following  grounds:   x

Transverse  deflection  caused  by  compression  takes  some  time  to  occur  because  of  inertia.  

342  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

x

The   segment   Euler  load   used   in   OrcaFlex   as   a   basis   for   comparison   is   the   lowest   of   the   various   alternatives,   and   assumes  pinned  joints  with  no  bend  stiffness  at  each  end  of  the  segment.  This  is  a  conservative  assumption.  

x

Whether   or   not   to  disregard   an   infringement   is   a   decision   which   can   only   be   taken   by   the   analyst   in   the   context   of  the  task  in  hand.  

Limit  Compression  Switch   For  each  line  type,  the  data  includes  a  Limit  Compression  switch.   The   usual   setting   is   "No".   This   means   that   each   segment   of   this   line   type   is   treated   as   a   strut   capable   of   taking   whatever  compressive  loads  arise  in  the  course  of  the  simulation.   For   some   special   cases,   such   as   chains   and   soft   ropes   with   little   bend   stiffness,   this   is   not   the   most   useful   model   and   OrcaFlex  offers  an  alternative.  Lines  of  this  sort  cannot  take  compression  at  all,  so  the  "Limit  Compression"  switch   can  be  set  to  "Yes".  OrcaFlex  then   does  not  allow  compressive  loading  greater  than  the  segment  Euler  load  (which  is   zero  if  the  bend  stiffness  is  zero).   Note:  

6.9

In   either   case,   if   the  segment   Euler  load   is   reached   then   a  Warning   is  given   on   the   result  form   and   in  the  statistics  table.  

6D  BUOYS  

6D   Buoys   are   objects   having   all   six  degrees   of   freedom  Ȃ   3   translational   (X,   Y   and   Z)   and   3   rotational   (Rotation   1,   2   and  3).  The  forces  acting  on  a  buoy  are  mass,  buoyancy,  added  mass  and  damping  and  drag  in  the  three  principal   buoy  directions.  Corresponding  moments  are  applied  for  the  rotational  degrees  of  freedom.  Buoys  can  be  surface-­‐ piercing,   and   have   a   notional   height;   this   allows   the   hydrostatic   and   hydrodynamic   forces   to   be   proportioned   depending  on  the  depth  of  immersion.   6D   Buoys   can   have   wings   attached   to   them.   A   wing   is   a   rectangular   surface,   attached   to   the   buoy   at   a   specified   position  and  orientation,  which  experiences  lift  and  drag  forces,  and  a  moment,  due  to  the  relative  flow  of  the  sea   past  the  wing.   Lines  attached  to  a  6D  Buoy  can  thus  experience  both  moment  effects  and  translations  as  the  buoy  rotates  under  the   influence  of  hydrodynamics  and  applied  loads.  Lines  can  be  attached  to  an  offset  position  on  a  buoy  Ȃ  this  allows  the   direct  study  of  line  clashing,  including  the  separation  introduced  by  spaced  attachment  points.   Three  types  of  6D  Buoy  are  available,  the  differences  being  the  way  in  which  the  geometry  of  the  buoy  is  defined.   Lumped  Buoys  

The   first   type,  Lumped   Buoys,   are   specified   without  reference   to   a   specific   geometry.  This  necessarily  restricts  the   accuracy  with  which  interactions  with  the  water  surface  are  modelled.  Where  a  lumped  buoy  pierces  the  surface  it  is   treated  for  buoyancy  purposes  as  a  simple  vertical  stick  element  with  a  length  equal  to  the  specified  height  of  the   buoy   (thus   buoyancy   changes   linearly   with   vertical   position   without   regard   to   orientation).   This   model   does   not   provide  the  rotational  stiffness  that  would  be  experienced  by  most  surface  piercing  buoys.   Interactions  with  the  seabed  and  with  shapes  are  also  modelled  in  a  fairly  simple  manner,  and  friction  effects  are  not   included.  Arbitrary  hydrodynamic  and  physical  properties  are  modelled  by  deriving  equivalent  terms.   Spar  Buoys  

The   second   type,   called   Spar   Buoys,   are   intended   for   modelling   axi-­‐symmetric   buoys   whose   axis   is   normally   vertical,  particularly  where  surface  piercing  effects  are  important  (such  as  for  a  CALM  buoy).   Spar  Buoys  are  modelled  as  a  series  of  co-­‐axial  cylinders  mounted  end  to  end  along  the  local  z-­‐axis  (see  Spar  Buoy   and  Towed  Fish  Properties).  This  allows  you  to  provide  some  information  about  the  buoy  geometry,  by  specifying   the   number   of   cylinders   and   their   lengths   and   diameters.   A   conical   or   spherical   shape   can   be   approximated   as   a   series  of  short  cylinders  of  gradually  increasing  or  diminishing  diameter.   Spar   Buoys   model   surface-­‐piercing   effects   in   a   much   more   sophisticated   way   than   Lumped   buoys.   Effects   such   as   heave  stiffness  and  righting  moments  in  pitch  and  roll  are  determined  by  calculating  the  intersection  of  the  water   surface  with  each  of  the  cylinders  making  up  the  buoy,  allowing  for  the  instantaneous  position  and  attitude  of  the   buoy   in   the   wave.   However   note   that   OrcaFlex   does   not   calculate   radiation   damping   (a   linear   damping   term   resulting  from  the  creation  of  surface  waves  as  the  buoy  oscillates)  or  impact  loads  (slamming).   Because   they  are   modelled  as   a   stack   of   concentric   cylinders,   Spar   Buoys   are  often  not   suitable   for  fully  submerged   objects  with  more  complex  geometry.  

343  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

As  with  Lumped  Buoys,  the  modelling  of  seabed  interaction  is  simplistic  and  friction  effects  are  not  included.   Hydrodynamic   loads   on   Spar   Buoys   are   calculated   using   Morison's   equation.   Added   mass   and   drag   forces   are   applied  only  to  those  parts  of  the  buoy  which  are  in  the  water  at  the  time  for  which  the  force  is  calculated.  For  partly   immersed   cylinders,   added   mass   and   drag   are   proportioned   according   to   the   fraction   of   the   cylinder   which   is   immersed.   The   use   of   Morison's   equation   implies   that   the   buoy   diameter   is   small   compared   to   the   wavelength   (usually  the  case  for  CALM  buoys  and  the  like)  but  means  that  some  load  terms  are  not  represented.   Towed  Fish  

The   third   type,   called   Towed   Fish,   are   intended   for   modelling   bodies,   such   as   towed   fish,   whose   principal   axis   is   normally  horizontal.  Towed  Fish  buoys  are  identical  to  Spar  Buoys  except  that  the  stack  of  cylinders  representing   the  buoy  is  laid  out  along  the  x-­‐axis  of  the  buoy,  rather  than  along  the  z-­‐axis.  

6.9.1

Wings  

6D  buoys  can  have  a  number  of   wings  attached;  these   are  useful  for   representing  lift  surfaces,  diverters   etc.  Each   wing  has  its  own  data  and  results  available.   A  wing  is  a  rectangular  surface,  attached  to  the  buoy  at  a  specified  position  and  orientation,  which  experiences  lift   force,   drag   force   and   drag   moment,   due   to  the   relative   flow   of  fluid   past   the   wing.   These   drag   loads   depend   on  user-­‐ specified  coefficients  that  depend  on  the  incidence  angle  of  the  relative  fluid  flow.   The  fluid  referred  to  here  can  be  the  sea,  the  air,  or  both,  as  follows.   x

Whenever   the   wing   is   completely   below   the   instantaneous   water   surface,   then   the   lift   and   drag   loads   are   calculated  using  the  sea  density,  velocity  and  incidence  angle.  

x

Whenever   the   wing   is   completely   above   the   water   surface,   and   if   you   have   selected   to   include   wind   loads   on   wings   (on   the   Wind   page   on   the   Environment   data   form),   then   instead   air   lift   and   drag   loads   are   calculated   and   applied,  using  the  same  formulae  and  coefficients,  but  using  the  air  density,  velocity  and  incidence  angle.  

x

When   the  wing   is  partially  submerged,   OrcaFlex   calculates   what   proportion   of   the   wing   rectangle  area   is  below   the   instantaneous   water   surface,   i.e.   its   'proportion   wet'   PW.   OrcaFlex   then   calculates   the   water   lift   and   drag   loads  as  if  the  wing   was  fully  submerged,  but  then  scales  them   by  PW  before  they  are   applied.  In  addition,  if  you   have  selected  to  include  wind  loads  on  wings,  then  OrcaFlex  also  calculates  the  air  lift  and  drag  loads  (as  if  the   wing   was   not  submerged)   and   scales   them   by   1-­‐PW,   i.e.   the   'proportion   dry',   before   they   are   applied.   When   this   happens,  therefore,  both  water  and  air  lift  and  drag  loads  are  applied,  each  appropriately  scaled.  The  wing   lift,   drag  and  moment  results  then  report  the  water  loads  whenever  the  wing  is  more  than  half  submerged  and  the   air  loads  whenever  it  is  less  than  half  submerged.  

Wings   do   not   have   any   mass,   added   mass   or   buoyancy   associated   with   them.   Therefore   any   mass,   added   mass   or   buoyancy  due  to  wings  should  be  added  into  the  properties  specified  for  the  buoy  itself.   The   drag   force   on   a   wing   is   the   force   applied   in   the   direction   of   relative   flow.   The   lift   force   is   the   force   at   90°   to  that   direction.   The   moment   represents   the   moment   (about   the   wing   centre)   that   arises  due   to   the   fact   that   the   centre   of   pressure  may  not  be  at  the  wing  centre.  These  loads  are  applied  at  the  wing  centre  and  are  specified  by  giving  lift,   †”ƒ‰ ƒ† ‘‡– …‘‡ˆˆ‹…‹‡–• ƒ• ƒ ˆ—…–‹‘ ‘ˆ –Š‡ ‹…‹†‡…‡ ƒ‰Ž‡ Ƚ „‡–™‡‡ –Š‡ ”‡Žƒ–‹˜‡ ˜‡Ž‘…‹–› ˜‡…–‘” ȋˆŽ‘™ velocity  relative  to  wing)  and  the  wing  plane.  

344  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

+ve lift Flow Velocity V (relative to wing) D

Wy Chord

Wz

Principal Wing Axis

W Leading edge

Wx

Span -ve lift   Figure:  

Wing  Model  

Each  wing  has  its  own  set  of  local  wing  axes,  with  origin  W  at  the  wing  centre  and  axes  Wx,  Wy  and  Wz.   x

Wy  is  normal  to  the  wing  surface  and  points  towards  the  positive  side  of  the  wing,  i.e.  the  side  towards  which   positive  lift  forces  act.  

x

Wx  and  Wz  are  in  the  plane  of  the  wing.  The  wing  is  therefore  a  rectangle  in  the  Wxz  plane,  centred  on  W.  

x

Wz  is  the  principal  axis  of  the  wing.  It  is  the  axis  about  which  the  wing  can  easily  be  pitched,  by  adjusting  the   wing  gamma  angle.  

x

Wx  is  in  the  plane  of  the  wing,  normal  to  the  axis  Wz,  so  that  (Wx,Wy,Wz)  form  a  right-­‐hand  triad.  

x

We   normally   choose   Wz   and   Wx   so   that   Wx   is   towards   the   leading   edge   of   the   wing.   With   this   arrangement,   increasing  the  wing  gamma  angle  moves  the  leading  edge  in  the  direction  of  positive  lift.  

We  refer  to  the  wing's  length  in  the  Wz  direction  as  its  span  and  its  width  in  the  Wx  direction  as  its  chord.   If   the   wing   is   not   completely   submerged,   then   the   forces   and   moments   applied   by   OrcaFlex   are   scaled   down   according   to   the   proportion   of   the   wing   area   that   is   below   the   surface.   However,   note   that   the   true   effects   of   breaking  surface,  for  instance  planing  and  slamming,  are  much  more  complex  than  this  and  are  not  modelled.  

6.9.2

Common  Data  

All  types  of  6D  Buoy  use  a  local  buoy  axes  coordinate  system.  The  origin  of  the  buoy  axes  can  any  point  chosen  by   the   user,   but   the   buoy   axes   directions   should   be   in   the   directions   of   the   principle   axes   of   structural   inertia   of   the   buoy  Ȃ  see  Mass  Moments  of  Inertia  below.   Name  

Used  to  refer  to  the  6D  Buoy.   Type  

Three   types   of   buoy   are   available:   Lumped   Buoys,   Spar   Buoys   and   Towed   Fish.   The   following   data   items   are   common  to  all  types.   Connection  

A   6D   Buoy   can   either   be   Free,   Fixed   or   connected   to   a   Vessel,   6D   Buoy   or   a   Line   (provided   that   line   includes   torsion).  

345  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

x

If  the  buoy  is  Free  then  it  is  free  to  move  in  response  to  wave  loads,  attached  lines  etc.  In  this  case  the  buoy's   Initial  Position  and  Attitude  are  specified  relative  to  global  axes.  

x

If  the  buoy  is  Fixed  then  it  cannot  move.  Its  Initial  Position  and  Attitude  are  specified  relative  to  global  axes.  

x

If   the   buoy   is   connected   to  a  Vessel,   6D   Buoy  or   a   Line,   then   it   is   rigidly   connected   to   that   object   and   so   moves   and   rotates   with  it.   All   resulting   forces   and  moments   on   the   buoy  are   transmitted  to   the   object.   In   this   case   the   buoy's  Initial  Position  and  Attitude  are  specified  relative  to  the  object  to  which  it  is  connected.  

Initial  Position  and  Attitude  

Specifies  the  initial  position  of  the  buoy  origin  and  the  buoy  initial  orientation.   If   the   buoy   is   Free   or   Fixed   then   its   initial   position   is   specified   by   giving   the   X,   Y   and   Z   coordinates   of   the   buoy   origin  B,  relative  to  the  global  axes.  And  its  initial  orientation  is  specified  by  giving   3  angles  Rotation  1,  Rotation  2,   Rotation  3,  which  are  successive  rotations  that  define  the  orientation  of  the  buoy  axes  Bxyz,  relative  to  global  axes,   as   follows.   First   align   the   buoy   with   global   axes,   so   that   Bxyz   are   in   the   same   directions   as   GXYZ.   Then   apply   Rotation   1   about   Bx   (=GX),   followed   by   Rotation   2   about   the   new  By   direction,   and   finally   Rotation   3   about   the   new   (and  final)  Bz  direction.   If  a  Free  buoy  is  not  included  in  the  static  analysis  then  this  initial  position  is  taken  to  be  the  static  position  of  the   buoy.   If   the  buoy   is  included   in   the   static   analysis,  then   this   initial   position   is  used   as   an   initial   estimate  of   the   buoy   position   and   the   static   analysis   will   move   and   rotate   the   buoy   from   this   position   until   an   equilibrium   position   is   found.  See  Degrees  of  Freedom  Included  in  Static  Analysis.   If  the  buoy  is  connected  to  a  Line,  then  the  Initial  Position  and  Attitude  specify  where  on  the  line  it  is  connected,  and   with  what  orientation,  as  follows:   x

The   Initial   Position   z-­‐coordinate   specifies   the   arc   length   at   which  the   buoy   should   be   connected   to   the   line.   The   buoy  will  be  connected  to  the  nearest  node  to  that  arc  length.  This  arc  length  may  be  measured  relative  to  either   End  A  or  End  B  as  specified  by  the  user.  

x

The  buoy  will  be  connected  to  that  node,  but  with  an  offset  relative  to  that  node's  axes  that  is  given  by  (x,  y,  0).  

x

The  buoy  orientation  relative  to  the  node  axes  is  specified  by  the  Initial  Attitude  angles.  

Degrees  of  Freedom  Included  in  Static  Analysis  

Determines  which  degrees  of  freedom  are  calculated  by  the   static  analysis.  This  data  item  only  applies  to  Free  buoys   and  it  can  be  set  to  one  of:   x

None:   the   buoy   position   and   orientation   are   not   calculated   by   the   static   analysis   Ȃ   they   are   simply   set   to   the   initial  position  and  orientation  specified  on  the  buoy  data  form.  

x

X,Y,Z:   the   buoy   position   is   calculated   by   the   static   analysis,   but   its   orientation   is   simply   set   to   the   initial   orientation  set  on  the  buoy  data  form.  

x

All:  the  buoy  position  and  orientation  are  calculated  by  the  static  analysis.  

Normally  this  data  item  should  be  set  to  All  so  that  the  static  analysis  calculates  the  true  equilibrium  position  and   orientation   of   the   buoy.   However   it   is   sometimes   useful   to   fix   the   buoy   position   or   orientation,   for   example   if   the   static  analysis  is  unable  to  find  the  equilibrium  position  or  orientation.   Mass  

Mass  or  weight  in  air.   Mass  Moments  of  Inertia  

The  solid  moments  of  inertia  of  the  buoy,  about  the  local  x,  y  and  z  buoy  axes  directions  through  its  centre  of  mass.   Note:  

These   moments   of   inertia   are   the   diagonal   terms   in   the   structural   inertia   matrix   about   the   specified  centre  of   mass.  The  off-­‐diagonal  terms  are  taken  to  be  zero,  so  the  buoy  axes  should  be   chosen  to  be  in  the  principle  directions  of  inertia  about  the  centre  of  mass  (or  at  least  be  close  to   those  directions).  

Centre  of  Mass  

The   centre   of   mass   of   the   buoy,   relative   to   the   buoy   origin.   The   weight   force,   mass   and   moments   of   inertia   are   applied  at  this  point.  

346  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

Bulk  Modulus  

Specifies  the  compressibility  of  the  buoy.  If   the  buoy  is   not  significantly  compressible,  then  the  Bulk  Modulus  can  be   set  to  Infinity,  which  means  'incompressible'.  See  Buoyancy  Variation  for  details.   Seabed  Friction  Coefficient    

”…ƒ Ž‡š ƒ’’Ž‹‡• ‘—Ž‘„ ˆ”‹…–‹‘ „‡–™‡‡ –Š‡ „—‘› ƒ† –Š‡ •‡ƒ„‡†Ǥ Š‡ ˆ”‹…–‹‘ ˆ‘”…‡ ƒ’’Ž‹‡† ‡˜‡” ‡š…‡‡†• Ɋ ™Š‡”‡‹•–Š‡•‡ƒ„‡†”‡ƒ…–‹‘ˆ‘”…‡ƒ†Ɋ‹•–Š‡ˆ”‹…–‹‘…‘‡ˆˆ‹…‹‡–Ǥ   See  Friction  Theory  for  further  details  of  the  friction  model  used.   Note:  

The  friction  coefficient  for  contact  with  elastic  solids  is  specified  on  the   Solid  Friction  Coefficients   data  form.  

Total  Contact  Area  

Used   to   determine   contact   forces   when   the   buoy   comes   in   to   contact   with   the   seabed   and   with   elastic   solids.   If   a   value  of  '~'  is  specified  then  OrcaFlex  calculates  a  default  value  for  total  contact  area  based  on  the  buoy  geometry.   See  Contact  Forces  for  details.   Specifying  a  value  of  zero  allows  you  to  disable  contact  for  the  buoy.  

6.9.3

Applied  Loads  

You   can   apply   to  the   buoy   external   Global   Loads   that   do   not   rotate   if   the   buoy   rotates.   These   are   specified   by  giving   the  components  of  Applied  Force  and  Applied  Moment  relative  to  global  axes.  These  components  can  be  constant,   vary  with  simulation  time  or  be  given  by  an  external  function.  If  the  buoy  rotates  then  the  loads  do  not  rotate  with  it.   In   addition,   you   can   specify   external   Local   Loads   that   do   rotate   with   the   buoy.   These   are   specified   by   giving   the   components  of  Applied  Force  and  Applied  Moment  relative  to  buoy  axes.  Again  these  components  can  be  constant,   vary  with  simulation  time  or  be  given  by  an   external  function.  If  the  buoy  rotates  then  the  loads   do  rotate  with  it.   These  are  suitable  for  modelling  thrusters,  for  example.   In  both  cases  the  Point  of  Application  of  the  load  is  specified  by  giving  its  x,y,z  coordinates  relative  to  buoy  axes.  

6.9.4

Wing  Data  

6D  buoys  can  have  a  number  of  wings  attached,  each  having  its  own  data  and  type.   Name  

Used  to  refer  to  the  wing.   Span  

The  length  of  the  wing,  in  the  local  Wz  direction.   Chord  

The  width  of  the  wing,  in  the  local  Wx  direction.   Centre  of  Wing  

The  position  of  the  wing  origin,  relative  to  buoy  axes.  The  wing  is  drawn  is  drawn  as  a  simple  rectangle  centred  on   this  origin,  the  rectangle's  dimensions  being  the  specified  Chord  and  Span.   Note  that  this  origin  is  the  point  at  which  the  relative  velocity  is  calculated  and  at  which  the  resulting  wing  lift  and   drag   forces   are   applied.   It   should   therefore   normally   be   at  the   centre   of   pressure,   which   may   not   be   at   the   centre   of   area  (even  though  OrcaFlex  draws  the  wing  as  being  centred  at  this  point).   Orientation  

The  orientation  of  the  wing  is  specified  by  giving  3  angles  Ȃ  azimuth,  declination  and  gamma  Ȃ  relative  to  the  buoy   axes.  The  angles  can  be  fixed,  or  can  vary  with  simulation  time,  or  can  vary  and  be  given  by  an  external  function.   The  angles  define  the  orientation  of  the  local  wing  axes  relative  to  the  buoy  axes  as  follows:   x

Start  with  the  wing  axes  Wxyz  aligned  with  the  buoy  axes  Bxyz  and  then  rotate  Wxyz  about  Bz  by  the  azimuth   angle.  This  leaves  Wz  aligned  with  Bz   but   Wx   now  points  in  the   direction  towards  which  the   declination   is  to  be   made.  

347  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

x

Now   rotate   by   the   declination   angle   about   the   new   direction   of   Wy.   This   declines   Wz   down   into   its   final   direction,  i.e.  Wz  now  points  along  the  direction  whose  azimuth  and  declination  angles  are  as  specified.  

x

Finally   rotate   by   the   gamma   angle   about   this   final   Wz   direction.   This   is   a   rotation   about   the   principal   wing   axis,   so  it  allows  you  to  adjust  the  pitch  of  the  wing.  

For   each   of   these   rotations,   positive   angles   mean   clockwise   rotation   and   negative   angles   mean   anti-­‐clockwise   rotation,  when  looked  at  along  the  axis  of  rotation.   When   setting   these   orientation   angles,   it   is   easiest   to   first   set   the   azimuth   and   declination   values   to   give   the   desired   Wz-­‐direction.   This  is  the   direction   of   the   axis  about  which  the   wing  pitch  is  set.   Then   set  gamma   to   give  the   correct   pitch   of   the   wing.   This   process   is   best   done   with   the   Draw   Local   Axes   option   set   on   (see   the   View   menu   or   the   Tools  |  Preferences  menu)  since  the  wing  axes  are  then  visible  on  the  3D  view  and  you  can  check  that  the  resulting   orientation  is  correct.   Wing  Type  

Determines  the  properties   of  the  wing.  You  can  define  a  number  of  wing  types   Ȃ  click  the  "Wing  Types"  button  to   access  the  wing  types  data  form.  

6.9.5

Wing  Type  Data  

6D  buoys  can  have  a  number  of  wings  attached,  each  having  its  own  data  and  type.   Name  

Used  to  refer  to  the  wing  type.   Wing  Type  Properties  

The   properties   of  each   wing   type   are   specified   by   giving   a   table   of  lift,   drag   and   moment   coefficients   as   a   function   of   the  incidence  angle  of  the  flow  relative  to  the  wing.   A  'Graph'  button  is  provided,  which  displays  a  graph  of  the  3  coefficients  so  that  you  can  visually  check  your  data.   Incidence  Angle  

Š‡‹…‹†‡…‡ƒ‰Ž‡‹•–Š‡ƒ‰Ž‡ǡȽǡ–Šƒ––Š‡”‡Žƒ–‹˜‡ˆŽ‘™˜‡…–‘”ƒ‡•–‘–Š‡™‹‰•—”ˆƒ…‡ǤŠ‹•‡“—ƒŽ•ͻͲ鏋—•–Š‡ angle  between  Wy  and  the  relative  flow  vector.   The   incidence   angle   is   always   in   the   range   -­‐90°   to   +90°,   where   positive   values   mean   that   the   flow   is   towards   the   positive   side   of   the   wing   (i.e.   hitting   the   negative   side)   and   negative   values   mean   that   the   flow   is   towards   the   negative  side  of  the  wing  (i.e.  hitting  the  positive  side).   The  incidence  angles  in  the  table  must  be  given  in  strictly  increasing  order  and  the  table  must  cover  the  full  range  of   incidence   angles,   so   the   first   and   last   angle   in   the   table   are   set   to   -­‐90°   and   +90°   and   cannot   be   changed.   Linear   interpolation  is  used  to  obtain  coefficients  over  the  continuous  range  of  angles.   Note:  

The   wing   lift,   drag   and   moment   are   assumed   to   depend   only   on   the   incidence   angle,   not   on   the   angle   of   attack   in   the   wing   plane.   OrcaFlex   will   therefore   use   the   same   lift,   drag   and   moment   coefficients  for  flow  (with  the  same  incidence  angle)  onto  the  front,  the  side  or  the  back  of  the  wing,   even   though   your   data   may   only   apply   over   a   limited   range   of   in-­‐plane   attack   angles.   You   can   check  that  the  angle  of  attack  in  the  wing  plane  stays  within  the  range  of  your  data  by  examining   the  Beta  angle  result  variable.  

Lift,  Drag  and  Moment  Coefficients  

These  define  the  hydrodynamic  and  aerodynamic  loads  applied  to  the  wing.  Aerodynamic  loads  are  only  applied  if   the  Include  wind  loads  on  6D  Buoy  Wings  option  is  enabled  in  the  Environment  data.   The   lift   coefficient   ClȋȽȌ †‡ˆ‹‡• –Š‡ Ž‹ˆ– ˆ‘”…‡ ƒ’’Ž‹‡† –‘ –Š‡ ™‹‰ǡ ƒ• ƒ ˆ—…–‹‘ ‘ˆ ‹…‹†‡…‡ ƒ‰Ž‡ ȽǤ Š‡ Ž‹ˆ– coefficients  can  be  positive  or  negative  and  the  lift  force  is  given  by:   Lift  Force  =  P  ½.ClȋȽȌǤɏǤǤ2   where   P  is  the  proportion  wet  or  proportion  dry  as  appropriate.   ɏ‹•–Š‡ˆŽ—‹††‡•‹–›ǡ   A  is  the  area  of  wing  that  is  inside  the  fluid  and    

348  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

V  is  the  relative  flow  velocity  at  the  wing  centre.   The   lift   force   is   applied   at   the   wing   centre,   along   the   line   that   is   at   90°   to   the   relative   flow   vector   and   in   the   plane   of   –Šƒ–˜‡…–‘”ƒ†›Ǥ ‘”ȽαάͻͲι–Š‹•Ž‹‡‹•‹ŽŽ-­‐defined   and   the   lift   coefficient   must   be   zero.   Positive   lift   coefficients   mean  lift  pushing  the  wing  towards  its  positive  side  (the   +ve  Wy  side).   The  drag  force  is  defined  by  the  drag  coefficient  CdȋȽȌ—•‹‰–Š‡ˆ‘”—Žƒǣ   Drag  Force  =  P  ½.CdȋȽȌǤɏǤǤ2   The  drag  coefficient  cannot  be  negative,  so  the  drag  force  is  always  in  the  relative  flow  direction.   The  moment  coefficient  CmȋȽȌ†‡ˆ‹‡•ƒ‘‡––Šƒ–‹•ƒ’’Ž‹‡†–‘–Š‡™‹‰ǤŠis  moment  represents  the  fact  that   –Š‡’‘•‹–‹‘‘ˆ–Š‡…‡–”‡‘ˆ’”‡••—”‡ƒ›†‡’‡†‘–Š‡‹…‹†‡…‡ƒ‰Ž‡ȽǤ   The  moment  coefficients  can  be  positive  or  negative  and  the  moment  is  given  by:   Moment  =  P  ½.CmȋȽȌǤɏǤǤ2.Chord   This   moment   is   applied   about   the  Ž‹‡–Šƒ–‹•‹–Š‡™‹‰’Žƒ‡ƒ†‹•ƒ–ͻͲι–‘–Š‡”‡Žƒ–‹˜‡ˆŽ‘™˜‡…–‘”Ǥ ‘”ȽαάͻͲι this  line  is  ill-­‐defined  and  the  moment  coefficient   must   be  zero.  Positive  moment  coefficients  mean  that  the  moment   is   trying   to   turn   the   wing   to   bring   Wy   to   point   along   the   relative   flow   direction.   Negative   moment   coefficients   mean   the  moment  tries  to  turn  the  wing  the  opposite  way.   Pen  

This  controls  which  pen  is  used  to  draw  a  Wing  which  uses  this  Wing  Type.  If  it  is  set  to  Use  Buoy's  pen  then  the  6D   Buoy  pen  will  be  used.  If  it  is  set  to  Use  this  pen  then  the  specified  pen  is  used.   Shaded  Drawing  

By  default,  for  shaded  3D  Views,  wings  are  drawn  as  plates  with  the  specified  span  and  chord.   Alternatively  the   object   can  be  represented  by  an  imported   3D   model  by   specifying  the  Shaded  Drawing   File.  This   must  be  a   Direct   X  format  file,  usually  with   the  .x   file   extension.  If  you  use  a  relative   path  then  the  path   will  be  taken   as  relative  to  the  folder  containing  the  OrcaFlex  file.   The   Browse   button   allows   you   to   specify   the   Shaded   Drawing   File   quickly   and   also   provides   quick   access   to   the   Orcina  sample  drawings  via  the  Navigate  to  Orcina  sample  drawings  button.   The  Use   Culling  option   is   normally   selected   since   it   can   provide   a   useful   performance   benefit.   However,   in   order   to   work   it  requires  that  the   triangles  defined  in  the  .x   file  have  their  outward  facing  directions  defined  correctly.  In   the   unusual   situation   where   the   outward   facing   directions   are   not   defined   correctly   then   the   .x   file   will   not   display   correctly.  If  this  happens  then  typically  sections  of  the  model  will  be  missing  when  drawn  by  OrcaFlex.  Disabling   the   Use  Culling  option  resolves  this  problem.   Draw  Size  is  provided  to  allow  you  to  scale  the  drawing.  All  directions  are  scaled  equally  to  arrange  that  the  longest   side  in  the  drawing  is   drawn  to  the  specified  Draw  Size.  This  longest  side  is  calculated  by  first  fitting  the  smallest   possible  cuboid  around  the  vertices  of  the  shaded  drawing  (these  are  defined  in  the  .x  file).  This  cuboid  is  aligned   with  the  shaded  drawing's  local  axes.  Then  the  length  of  the  longest  side  of  this  cuboid  is  found.   Specify  a  value  of  '~'  to  display  the  drawing  using  the  absolute  coordinates  as  specified  in  the  .x  file.   Note:  

If  you  use  a   value  of  '~'  for   Draw  Size  then  OrcaFlex   uses  the  coordinates  in  the  .x   file  directly.  If   these   coordinates   use   a   different   length   units   system   from   your   OrcaFlex   model   then   you   should   specify   the   units   used   in   the   .x   file   by   including   an   auxiliary   file   called   AdditionalInformation.txt.   Examples  of  this  can  be  found  in  the  sample  shaded  drawings  provided  by  Orcina.  

Shaded   Drawing   Origin   is   provided   because   the   shaded   drawing   and   the   wing   may   have   different   origins.   The   Shaded   Drawing   Origin   defines   the   origin   of   the   shaded   drawing   with   respect   to   the   wing's   local   axis   system.   Similarly  Shaded  Drawing  Orientation  allows  you  to  reorient  the  shaded  drawing  to  match  the  wing's  axis  system.  

349  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

6.9.6  

Lumped  Buoy  Properties   Vertices

z (heave)

yaw

y (sway) pitch

B (B=buoy origin)

roll

x (surge)

  Figure:  

Lumped  Buoy  

A   Lumped   Buoy   is   specified   relative   to   its   own   local   frame   of   reference   Bxyz.   The   Centre   of   Mass   is   specified   relative  to  Bxyz.   Š‡ˆ‘”…‡•ƒ†‘‡–•ƒ”‡…ƒŽ…—Žƒ–‡†ƒ•ˆ‘ŽŽ‘™•ȋɏ‹•™ƒ–‡”†‡•‹–›ǡ‰‹•  acceleration   due   to   gravity).   Each  degree   of  freedom  is  calculated  independently.   Geometry  

Volume  is  the  total  volume  of  the  buoy,  with  its  centre  at  the  Centre  of  Volume,  defined  relative  to  the  local  buoy   axes  Bxyz.   Height   is   the   buoy   vertical   dimension,   assumed   equally   spaced   about   the   centre   of   volume.   Height   is   assumed   to   be   independent  of  buoy  rotation.   The  Height  is  used  for  two  main  purposes:   x

To  calculate  the  contact  area  used  to  calculate  forces  of  contact  with  shapes  and  the  seabed.  

x

To   calculate   the   ProportionWet,   which   is   used   to   scale   the   hydrostatic   and   hydrodynamic   forces   and   to   determine  the  centre  of  wetted  volume,  which  is  the  point  at  which  they  are  applied.  

Damping  

Hydrodynamic   damping   forces   and   moments   may   be   applied   to   the   buoy.   These   are   loads   that   are   directly   proportional   to   the   relative   velocity,   or   angular   velocity,   of   the   sea   past   the   buoy.   For   each   of   the   local   buoy   axes   directions,   you   specify   the   magnitude   of   the   Unit   Force   that   is   applied   when   the   relative   velocity   is   1  length  unit/second.  OrcaFlex  then  scales  these  magnitudes  according  to  the  actual  relative  velocity  and  applies  t he   resulting   force   or   moment.   Similarly   you   can   specify   a   Unit   Moment   that   is   applied   when   the   relative   angular   velocity  is  1  radian/second.   Drag  

Hydrodynamic  drag  forces  and  moments  may  be  applied  to  the  buoy.  These  are  loads  that  are  proportional  to  the   square  of  the  relative  velocity,  or  angular  velocity,  of  the  sea  past  the  buoy.   The  drag  force  properties  are  specified  by  giving,  for  each  of  the  local  buoy  axes  directions,  the   Drag  Area  that  is   subject  to  drag  loading  in  that  direction  and  the  corresponding  Drag  Coefficient.   Drag   moment   properties   are   specified   in   a   similar   way,   except   that   instead   of   specifying   a   drag   area   you   must   specify  a  Moment  of  Area.  

350  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

Note:  

Drag  Area  Moment   is  the  3rd  absolute   moment  of   drag  area  about  the  axis.  S eparate  Cd  values  are   given  for  force  and  moment  calculations.  

Fluid  Inertia   Fluid   inertia   properties   are   those   that   are   proportional   to   the   acceleration   of   the   sea   and   the   buoy.   These   accelerations  have  two  main  effects.  Firstly,  they   result  in  forces  and  moments   being  applied   to  the   buoy  Ȃ  these   are   referred  to  as  the  fluid  acceleration  loads.  Secondly,  the  buoy  experiences  an  increase  in  inertia   Ȃ  this  is  known  as   the  added  mass.   Full  details  of  the  calculation  are  presented  in  the  Lumped  Buoy  Added  Mass,  Damping  and  Drag  topic.   The  fluid  inertia  properties  can  be  specified  either  by  providing  the   Diagonal  values  or  Full  matrices.   Fluid  Inertia  specified  by  diagonal  values  

These   properties   are   specified   separately   for   translational   and   rotational   motions   and   also   separately   for   each   local   axis  direction.   The   translational   fluid   inertia   properties   of   the   buoy   are   specified,   for   each   of   the   local   buoy   axis   directions,   by   giving   a   reference  Hydrodynamic  Mass   together   with   the   two   inertia   coefficients,   Ca   and   Cm.   The   translational   Cm   values  can  be  set  to  '~'  to  mean  1  +  Ca.   The   translational   Hydrodynamic   Mass   values   can   be   set   to   '~',   meaning   equal   to   the   fully   submerged   displaced   mass.(=  volume  x  water  density).  This  is  often  a  convenient  reference  mass  to  use.   Fluid  Inertia  specified  by  full  matrices  

The   added   mass   and   fluid  inertia   properties   are   specified   as   full   6x6   symmetric   matrices.   The   added   mass   matrix   is   simply  added  in  to  the  buoy's  virtual  mass  matrix.  The  fluid  inertia  matrix  is  multiplied  by  the  instantaneous  fluid   acceleration  vector  to  produce  the  fluid  acceleration  loads.   The  units  of  the  added  mass  and  fluid  inertia  matrices  are  as  follows,  where  M,  L  and  T  indicate  the   units  of  mass,   length  and  time:   x

The  upper-­‐left  3x3  block  of  cells  has  units  M.  

x

The  upper-­‐right  and  lower-­‐left  3x3  blocks  have  units  M.L.  

x

The  lower-­‐right  3x3  block  of  cells  has  units  M.L2.  

6.9.7

Lumped  Buoy  Drawing  Data  

Vertices  and  Edges   This  defines  a  "wire  frame"  representation  of  the  buoy.  The  wire  frame  representation  of  the  buoy  is  used  to  draw   the  buoy.  It  is  also  used  to  calculate  the  interaction  of  the  buoy  with  shapes  and  the  seabed.  See   Contact  Forces  for   details.   The  vertices  are  specified  by  giving  their  coordinates  with  respect  to  buoy  axes  Bxyz,  and   the  edges  are  specified   by   giving  the  vertex  numbers  of  the  vertices  they  connect.   Edge  diameter  (used  only  for  shaded  drawing)  

For  shaded  graphics  views,   by  default,  the  Lumped  Buoy  is  drawn  using  a  solid,   filled-­‐in  shape  based  on  the  vertices   and  edges.  As  an  alternative   you  can  use  the  vertices  and  edges  to  define  a  frame  like   structure.  If  the  edge  diameter   is   '~'   then   that   edge   will   be   used   to   build   a   filled   in   shape,   otherwise   that   edge   is   drawn   as   a   cylinder   with   the   specified  diameter.  

351  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

  Figure:  

6.9.8

Wire  frames  with  different  edge  diameter.  A   value  of  '~'  is  used  for  the   wire  frame  on  the  left   and  a  value  of  1m  is  used  for  the  wire  frame  on  the  right.  

Spar  Buoy  and  Towed  Fish  Properties  

The   figure   below   shows   the   geometry   of   a   Spar   Buoy.   The   buoy   is   specified   relative   to   its   own   local   frame   of   reference  Bxyz.  The  Centre  of  Mass  is  specified  relative  to  Bxyz.   A  Spar  Buoy  is  modelled  as  a  series  of  co-­‐axial  cylinders  mounted  end  to   end  along  the  local  z-­‐axis.  The  geometry  of   a  Towed  Fish  is  identical  except  that  the  buoy  axis  is  aligned  with  the  x-­‐axis  of  the  buoy.   The  cylinders  are  numbered  from  the  top  downwards.  So  in  the  tables  on  the  buoy  data  form  the  cylinder  at  the  base   of  the  stack  (lowest  x  or  z)  appears  at  the  bottom  of  the  table.   If  you  are  modelling  a  CALM  or  SPAR  buoy  then  see  also   Modelling  a  Surface-­‐Piercing  Buoy.  

352  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

 

Buoy Axis Cylinder 1 OD

Cylinder 1 Length

z (heave)

yaw

y (sway) pitch

B (B=buoy origin)

x (surge)

roll

Stack Base Position   Figure:  

Spar  Buoy  

Geometry   The  shape   of  a  Spar  Buoy  or  Towed  Fish  is  specified  by  the  following  data  on  the  Geometry  page  on  the   buoy  data   form.   Stack  Base  Centre  Position  

The  centre  of  the  base  of  the  stack,  relative  to  buoy  axes.   Cylinders:  Outer  Diameter,  Inner  Diameter  and  Length  

The  diameters  of  the  cylinder  and  its  length  measured  along  the  axis.  

353  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

These  parameters  define  the  buoy  geometry  from  which  buoyancy  forces  and  moments  are  determined.  When  the   buoy  pierces  the  water  surface,  OrcaFlex  allows  for  the  angle  of  intersection  between  the  sea  surface  and  the  buoy   axis   when   calculating   the   immersed   volume   and   centre   of   immersed   volume,   and   includes   the   appropriate   contributions  to  static  stability.   If   the   inner   diameter  is  greater   than   zero   then   the   cylinder   is  actually  a   hollow   cylindrical   pipe.   The   internal   region   of  the   cylinder   contains   sea   water   if  the  buoy   is   submerged.   In   this   case   then   the   contained   water   contributes   to   the   buoy's  inertia.  These  inertial  effects  are  only  included  in  directions  normal  to  the  buoy.   The  remaining  parameters  determine  the  hydrodynamic  loads  acting  on  each  cylinder.  Loads  are  calculated  for  each   cylinder  individually,  then  summed  to  obtain  the  total  load  on  the  buoy.   Hydrodynamic  Loads   See  the  Added  Mass  and  Damping  and  Drag  pages  on  the  buoy  data  form.  

6.9.9

Spar  Buoy  and  Towed  Fish  Added  Mass  and  Damping  

There  is  a  choice  of  how  to  model  these  first  order  effects,  on  the  Added  Mass  and  Damping  page  of  the   spar  buoy   data  form.  They  can  either  be  specified  by  giving  added  mass  and  damping  values  for  each  cylinder,  or  else  by  giving   RAOs  and  matrices  for  the  whole  buoy.   Values  for  Each  Cylinder   With   this   option,   the   added   mass   and   damping   effects   are   calculated   separately   for   each   cylinder   using   Morison's   Equation.   Added  Mass  

Translational   added   mass   effects   are   calculated   using   the   displaced   mass   as   the   reference   mass   for   each   cylinder.   Separate   added   mass   coefficients   are   given   for   flow   normal   (x   and   y   directions)   and   axial   (z   direction)   to   the   cylinder.   Rotational  added  inertia  is  specified  directly  (so  no  reference  inertia  is  involved).  Separate  values  can  be  given  for   rotation  about  the  cylinder  axis  and  normal  to  that  axis.   See  Spar  Buoy  Theory.   Damping  Forces  and  Moments  

Damping  forces  and  moments  are  the  hydrodynamic  loads   that  are  proportional  to  fluid  velocity  (angular  velocity   for   moments)   relative   to   the   cylinder.   They   are   specified   by   giving   the   Unit   Damping   Force   and   Unit   Damping   Moment  for  the  normal  and  axial  directions.  These  specify  the  force  and  moment  that  the  cylinder  will  experience,  in   that   direction,   when   the   relative   fluid   velocity   (angular   velocity   for   moments)   in   that   direction   is   1   unit.   See   Damping  Forces  and  Moments  for  details.   RAOs  and  Matrices  for  Buoy   In  this  option  the  linear  hydrodynamic  effects  are  specified  by  giving  wave  force  and  moment  RAOs,  and  added  mass   and  damping  matrices.  Normally  this  data  would  come  from  a  separate  program,  such  as  a  d iffraction  program.   Note:  

This   option   is   only   appropriate   for   circularly-­‐symmetric   spar   buoys   whose   axis   undergoes   only   small  oscillations  about  vertical.  It  was  developed  primarily  for  CALM  buoys  with  diameter  in  the   range  5m  to  15m.  It  is  therefore  not  available  for  towed  fish.  

Warning:  

The  RAOs   and   added   mass   and  damping  matrices   specified   must   apply   to   the   mean   position   of   the   buoy.  They  are   not   modified  to  account  for  any  variations  in  buoy  attitude  or  immersion,  so  this   option  is  only  suitable  for  buoys  that  undergo  small  oscillations  about  their  mean  position.  

RAO,  Added  Mass  and  Damping  Origin  

This   specifies   the   coordinates,   with   respect   to   buoy   axes,   of   the   point   on   the   buoy   at   which   the   RAOs   and   added   mass  and  damping  matrices  are  applied.  This  means  that:   x

The   RAOs   are   applied   to   the   wave   conditions   at   this   point   to   give   the   wave   loads,   which   are   then   applied   at   this   point.  

x

The  damping  matrix  is  multiplied  by  the  current  velocity  relative  to  the  buoy  at  this  point  to  give  the  damping   load,  which  is  applied  at  this  point.  

354  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

x

The  added  mass  matrix  is  applied  at  this  point.  

Wave  Force  and  Moment  RAOs  

The  RAOs  are  specified  in  a  table  giving  the  amplitudes  and  phases  of  the  surge  force,  heave  force  and  pitch  moment   RAOs,  as  a  function  of  wave  angular  frequency.  RAOs  must  be  specified  for  both  of  the  limiting  cases,  i.e.  for  long  and   short  waves,  and  for  at  least  one  other  frequency.   The   RAOs   must   be   specified   using   the   following   Orcina   conventions.   The   force   and   moment   amplitudes   must   be   the   forces   and   moments   that   are   applied   at   the   buoy   origin   when   a   unit   amplitude   wave   is   applied.   The   phases   m ust   be   lags,  in  degrees,  from  the  time  the  wave  crest  passes  the  buoy  origin  until  the  maximum  positive  force  or  moment   occurs.   Added  Mass  and  Damping  Matrices  

In  reality  the  added  mass  and  damping  matrices  are  frequency-­‐dependent,  but  in  OrcaFlex  you  can  only  enter  the   matrices  for  one  frequency.  You  should  give  the  values  for  the  principal  frequency  of  excitation  expected.   The   added   mass   and   damping   matrices   are   6   x   6   symmetric   matrices,   so   only   the   cells   on   and   above   the   leading   diagonal  are  editable.  The  cells  below  the  leading  diagonal  are  non-­‐editable  and  always  equal  the  corresponding  cell   values  above  the  diagonal.   The   units   of   the   added   mass   and   damping   matrices   are   as   follows,   where   M,   L   and   T   indicate   the   units   of   mass,   length  and  time:   x

Added   mass   matrix.   The   upper-­‐left   3x3   block   of   cells   has   units   M.   The   upper-­‐right   and   lower-­‐left   3x3   blocks   have  units  M.L.  The  lower-­‐right  3x3  block  of  cells  has  units  M.L2.  

x

Damping   matrix.   The   upper-­‐left   3x3   block   of   cells   has   units   F/(L/T).   The   upper-­‐right   3x3   block   has   units   F/(rad/T).  The  lower-­‐left  3x3  block  of  cells  has  units  (F.L)/(L/T).  The  lower-­‐right   3x3  block  of  cells  has  units   (F.L)/(rad/T).  

Because   the   buoy   is   assumed   to   be   axi-­‐symmetric   these   matrices   should   both   have   the   following   axi-­‐symmetric   form:  

§ ¨ ¨ ¨ ¨ ¨ ¨ ¨ ¨ ©

a 0 0 0 b 0

0 a 0 b 0 0

0 0 c 0 0 0

0 b 0 d 0 0

b 0 0 0 d 0

0 0 0 0 0 e

· ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸ ¹  

OrcaFlex  warns  at  simulation  time  if  the  matrices  are  not  of  this  form.  

6.9.10

Spar  Buoy  and  Towed  Fish  Drag  

Munk  Moment  Coefficient   Slender   bodies   in   near-­‐axial   flow   experience   a   destabilising   moment   called   the   Munk   moment.   This   effect   can   be   modelled  by  specifying  a  non-­‐zero  Munk  moment  coefficient.   Drag  Forces  and  Moments   Drag  loads  are  the  hydrodynamic  loads  that  are  proportional  to  the  square  of  fluid  velocity  relative  to  the  cylinder.   For   details   of   the   drag   load   formulae   see   Spar   Buoy   and   Towed   Fish   Theory.   For   information   when   modelling   a   SPAR  or  CALM  buoy  see  Modelling  a  Surface-­‐Piercing  Buoy.   The  drag  forces  are  calculated  on  each  cylinder  using  the  "cross  flow"  assumption.  That  is,  the  relative  velocity  of  the   sea  past  the  cylinder  is  split  into  its  normal  and  axial  components  and  these  components  are  used,  together  with  the   specified  drag  areas  and  coefficients,  to  calculate  the  normal  and  axial  components  of  the  drag  force.   The   drag   forces   are   specified   by   giving   separate   Drag   Area   and   Drag   Coefficient   values   for   flow   in   the   normal   direction  (local  x  and  y  directions)  and  in  the  axial  direction  (local  z  direction).  The  Drag  Area  is  a  reference  area   that  is  multiplied  by  the  Drag  Coefficient  in  the  drag  force  formula.  You  can  therefore  use  any  positive  Drag  Area  that   suits  your  need,  but  you  then  need  to  give  a  Drag  Coefficient  that  corresponds  to  that  specified  reference  area.  

355  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

The  Drag  moments  are  specified  and  calculated  in  a  similar  way  to  the  drag  forces,  except  that  the  reference  drag   area   is   replaced   by   a   reference   Area   Moment.   This   and   the   Drag   Coefficient   are   multiplied   together   in   the   drag   moment   formula,   so   again   you   can   use   any  positive   Area   Moment   that   suits   your   need,   providing   you   then   specify   a   Drag  Coefficient  that  corresponds  to  the  specified  Area  Moment.   There   are   two   alternative   methods   that   you   can   adopt   when   specifying   the  drag  data.   The   first   method   is  to   set   the   OrcaFlex   data   to   get  best  possible   match  with  real   measured  results   for   the  buoy   (e.g.  from  model   tests   or   full   scale   measurements).  This  is  the  most  accurate  method,  and  we  recommend  it  for  CALM  and  discus  buoys  Ȃ  see  Modelling   a   Surface-­‐Piercing   Buoy   for   details.   Because   the   Drag   Area   and   Drag   Coefficient   data   are   simply   multiplied   together,   you  can  calibrate  the  model  to  the  real  results  by  fixing  one  of  these  two  data  items  and  then  adjusting  the  other.  For   example,   you   could   set   the   axial   Drag   Coefficient   to   1   and   adjust   the   axial   Drag   Area   until   the   heave   response   decay   rate   in   the  OrcaFlex  model   best  matches  the  model  test  results.  Or,  you  could  set  the   axial  Drag  Area  to  a  fixed  value   (e.g.  1  or  some  appropriate  reference  area)  and  then  adjust  the  axial  Drag  Coefficient  until  the  heave  response  decay   rate  in  OrcaFlex  best  matches  the  model  test  results.   The  second  method  is  to  set  the  drag  data  using  theoretical  values  or  given  in  the  literature.  It  is  less  accurate  but   can  be  used  if  you  cannot  get  any  real  buoy  results  against  which  to  calibrate.   To   use  this  method,  set  the  data  as   follows.   Set  the  Drag  Areas  to  the  projected  surface  area  that  is  exposed  to  drag  in  that  direction  and  then  set  the  Drag  Force   Coefficients   based   on   values   given   in   the   literature   (see  Barltrop  &   Adams,   1991,   Hoerner,1965   and  DNV-­‐RP-­‐C205).   Note   that   the   drag   area   specified   should   be   the   total   projected   area   exposed   to   drag   when   the   buoy   is   fully   submerged,   since   OrcaFlex   allows   for   the   proportion   wet   in   the   drag   force   formula.   For   a   simple   cylinder   of   †‹ƒ‡–‡”  ƒ† Ž‡‰–Š  –Š‡ –‘–ƒŽ ’”‘Œ‡…–‡† †”ƒ‰ ƒ”‡ƒ ‹• Ǥ ˆ‘” –Š‡ ‘”ƒŽ †‹”‡…–‹‘ ƒ† ȋɎǤ 2)/4   for   the   axial   direction,  but  if  the  buoy  has  attachments  that  will  experience  drag  then  their  areas  must  also  be  included.   Set   the   Drag   Area   Moments   to   the   3rd   absolute   moments   of   projected   area   exposed   to   drag   in   the   direction   concerned;   see  Drag   Area   Moments  for   details.   And   then   set   the   Drag   Moment   Coefficients   based   on   values   given   in   the  literature.  

6.9.11

Spar  Buoy  and  Towed  Fish  Drawing  

There  are  two  alternative  methods  of  drawing  the  Spar  Buoy  and  Towed  Fish  cylinders:   1.

Draw  square   cylinders.   If   this   option   is   selected   then   each   cylinder   is   drawn   as   a   square-­‐based   cuboid.  OrcaFlex   automatically  generates   the  vertices   and   edges.   There   are  eight   vertices   per   cylinder   and   the   square   base   has   a   side  length  equal  to  the  diameter  of  the  actual  circular  cylinder.  

2.

Draw  circular  cylinders.   With  this  option  selected  the  circular  cylinders  are  drawn  as  such.  You  can  specify  pens   for   the  outside  and  inside  of  the  cylinders.  In  addition   you  specify  how   many  lines  are  drawn.  Increasing  this   number  makes  the  buoy  drawing  more  realistic  but  increases  the  time  taken  to  draw  the  buoy.   Notes:  

The   selected   drawing   method   only   affects   the   way   the   object   is   drawn.   It   does   not   affect   calculations  in  any  way.  

 

The  vertices   of  the  square-­‐based  cuboids  are  used   for  calculation  of  any   contact  with  the  seabed  or   shapes.  This  applies  even  if  you  choose  for  presentation  purposes  to  draw  circular  cylinders.  

6.9.12

Shaded  Drawing  

By   default,   for   shaded   3D   Views,   Spar   Buoys   and   Towed   Fish   are   drawn   as   solid   objects   using   the   specified   cylinder   geometry.  Lumped  Buoys  are  drawn  using  the  wire  frame  data.   The   wire   frame   drawing   data   comprises   vertices   and   edges,   but   OrcaFlex   needs   a   solid   surface   for   the   shaded   graphics   representation.   OrcaFlex   uses   the   following   procedure   to   generate   this   surface   from   the   wire   frame   vertices  and  edges.   First  any  edges  with  specified  diameters  (i.e.  diameters  not  equal  to  '~')  are  drawn  as   cylinders.  This  allows  you  to   use   such   edges   to   visualise   parts   of   the   structure   that   are   not   solid,   e.g.   crane   boom   latticework.   These   edges   are   now  handled  and  are  excluded  from  the  remainder  of  the  procedure.   The  remaining  edges  are  used  to  partition  the  vertices  into  sets  of  connected  vertices.  Two  vertices  are  deemed  to   be  connected  if  there  exists  a  path  of  edges  between  the  two  vertices.   Finally,  for  each  set  of  connected  vertices,  the  smallest  convex  hull  enclosing  the  set  is  drawn.   This   algorithm   does   not   always   generate   the   shaded   drawings   that   you   might   expect.   Consider   the   following   two   wire  frame  vessels.  When  drawn  in  wire  frame  mode  they  look  the  same,  but  in  shaded  mode  they  differ.  

356  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

  Figure:  

Wire  Frame  and  Shaded  Drawing  

For  the  green  vessel  the  superstructure  and  the  hull  share  vertices  and  so  all  vertices  are  connected.  This  results  in  a   single  convex  hull  for  all  vertices  being  drawn.  In  the  red  vessel,  the  superstructure  and   hull  do  not  share  vertices   and   so   there   are   two   distinct   sets   of   connected   vertices.   This   results   in   two   separate   convex   hulls   and   a   better   representation.   Alternatively  the   object   can  be  represented  by  an  imported   3D   model  by   specifying  the  Shaded  Drawing   File.  This   must  be  a   Direct   X  format  file,  usually  with   the  .x   file   extension.  If  you  use  a  relative   path  then  the  path   will  be  taken   as  relative  to  the  folder  containing  the  OrcaFlex  file.   The   Browse   button   allows   you   to   specify   the   Shaded   Drawing   File   quickly   and   also   provides   quick   access   to   the   Orcina  sample  drawings  via  the  Navigate  to  Orcina  sample  drawings  button.   The  Use   Culling  option   is   normally   selected   since   it   can   provide   a   useful   performance   benefit.   However,   in   order   to   work   it  requires  that  the  triangles  defined  in  the  .x   file  have  their  outward  facing  directions  defined  correctly.  In   the  

357  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

unusual   situation   where   the   outward   facing   directions   are   not   defined   correctly   then   the   .x   file   will   not   display   correctly.  If  this  happens  then  typically  sections  of  the  model  will  be  missing  when  drawn  by  OrcaFlex.  Disabling   the   Use  Culling  option  resolves  this  problem.   Draw  Size  is  provided  to  allow  you  to  scale  the  drawing.  All  directions  are  scaled  equally  to  arrange  that  the  longest   side  in  the  drawing  is  drawn  to  the  specified  Draw  Size.  This  longest  side  is  calculated  by  first  fitting  the  smallest   possible  cuboid  around  the  vertices  of  the  shaded  drawing  (these  are  defined  in  the  .x  file).  This  cuboid  is  aligned   with  the  shaded  drawing's  local  axes.  Then  the  length  of  the  longest  side  of  this  cuboid  is  found.   Specify  a  value  of  '~'  to  display  the  drawing  using  the  absolute  coordinates  as  specified  in  the  .x  file.   Note:  

If  you  use  a   value  of  '~'  for   Draw  Size  then  OrcaFlex   uses  the  coordinates  in  the  .x   file  directly.  If   these   coordinates   use   a   different   length   units   system   from   your   OrcaFlex   model   then   you   should   specify   the   units   used   in   the   .x   file   by   including   an   auxiliary   file   called   AdditionalInformation.txt.   Examples  of  this  can  be  found  in  the  sample  shaded  drawings  provided  by  Orcina.  

Shaded   Drawing   Origin   is   provided   because   the   shaded   drawing   and   the   buoy   may   have   different   origins.   The   Shaded   Drawing   Origin   defines   the   origin   of   the   shaded   drawing   with   respect   to   the   buoy's   local   axis   system.   Similarly  Shaded  Drawing  Orientation  allows  you  to  reorient  the  shaded  drawing  to  match  the  buoy's  axis  system.  

6.9.13

Other  uses  

6D  Buoys  are  frequently  used  for  non-­‐standard  purposes.  For  example  they  can  be  used  to  connect  Lines  together.   Sometimes  they  are  used  as  frame  of  reference  objects  where  the  properties  are  provided  by  various  other  objects   (usually  single  segment  Lines)  connected  to  the  6D  Buoy.   For  such  uses  the  6D  Buoy  needs  negligible  structural  and  hydrodynamic  properties.   Because  this  is  such  a  common   thing  to  do  we  have  provided  a  facility  on  the  6D  Buoy   data  form  to  set  up  appropriate  properties.  Clicking  on  the   Give  Buoy  negligible  properties  button  has  the  following  effects:   x

Buoy  Type  will  be  set  as  Lumped  Buoy.  

x

Mass  will  be  set  to  a  negligible  value.  

x

Height  will  not  be  modified  since  it  cannot  be  set  to  zero.  Because  other  properties  are  set  to  zero  its  value  is  not   significant.  

x

All  other  Lumped  Buoy  properties  are  set  to  zero.  

x

Any  existing  Wing,  Applied  Load  and  Drawing  data  is  not  modified.  

6.9.14

External  Functions  

Parameters  

This  data  item  specifies  the   External  Function  Parameters,  a  free   form  multi-­‐line  text   field   which  is  passed  to   any   external  function  used  by  the  Environment.  

6.9.15

Properties  Report  

The  6D  Buoy  properties  report  is  available  from  the   popup-­‐menu  on  the  data  form.  It  reports  the  following:   Weight  in  air  

The  force  due  to  gravity  acting  on  the  buoy's  mass.   Displacement  

The  weight  of  water  displaced  by  the  buoy's  volume.  The  reported  value  uses  the  water  density  at  the  sea  surface.   Weight  in  water  

Equals  Weight  in  air  -­‐  Displacement.   Volume  

The  volume  of  the  buoy.   Centre  of  volume  

Reported  with  respect  to  the  buoy  frame  of  reference.  

358  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

x-­‐mass  radius  of  gyration,  y-­‐mass  radius  of  gyration,  z-­‐mass  radius  of  gyration  

“—ƒŽ•ξȋƒ••‘‡–‘ˆ ‡”–‹ƒȀƒ••ȌǤ   Total  contact  area  

Reports  the  Total  contact  area  data.  If  a  value  of  '~'  is  set  on  the  data  form  then  the  value  reported  here  is  the   default   value  calculated  by  OrcaFlex.   Number  of  vertices  

The  number  of  vertices  in  the  wire  frame  drawing.  Contact  for  6D  Buoys  is  calculated  at  each  of  these  vertices.   Contact  area  per  vertex  

Equals  Total  contact  area  /  Number  of  vertices.   Statics  force  accuracy,  Statics  moment  accuracy  

These  are  only  reported  if  the  buoy  is  included  in  statics.   The  static  analysis  searches  for  an  equilibrium  position  for  the  buoy  Ȃ  that  is  a  position  for  which  the  resultant  force   and  moment   on  the  buoy  is  zero.   We   refer  to  the   resultant  force  and   moment  as  the  out  of  balance  load.  Because   computers  have  limited  numerical  precision  the  static  analysis  cannot  always  find  a  configuration  where  the  out  of   balance   load   is  exactly  zero.  OrcaFlex   accepts  a   position  as   a   static   equilibrium  position   if  the   largest  out   of  balance   load  component  is  less  than  the  statics  accuracy.   The  Statics  force  accuracy  equals   Tolerance  *  buoy  typical  force  and  the  Statics  moment  accuracy  equals   Tolerance  *   buoy   typical   moment.   The   buoy   typical   force   and   moment   are   based   on   the   weight   in   air,   the   height   of  the   buoy   and   the  forces  and  moments  applied  by  connected  objects.   Reducing   the   Tolerance   value   will   give   a   more   accurate   static   equilibrium   position,   but   will   take   more   iterations.   OrcaFlex   may   not   be   able   to   achieve   the   Tolerance   specified   if   it   is   too   small,   since   the   computer   has   limited   numerical  precision.   Note:  

6.9.16

The  statics  accuracies  change  during  the   static  analysis   because  the  forces  and   moments  applied   by   connected   objects   vary   with   the   position   of   the   buoy.   The   statics   accuracies   reported   in   Reset   state   may   be   quite   different   from   those   used   for   the   final   equilibrium   position   and   should   be   treated  as  rough  approximations  to  the  true  statics  accuracies.  

Results  

For  details  on  how  to  select  results  variables  see  Selecting  Variables.   6D  Buoy  Results   For  6D  Buoys  the  available  variables  are:   X,  Y  and  Z  

The   position   of   a   user   specified   point   P   on   the   buoy,   relative   to   global   axes.   The   point   P   is   specified   in   buoy   local   coordinates.  If  P=(0,0,0)  then  the  global  X,  Y  and  Z  coordinates  of  the  buoy  origin  are  reported.   Rotation  1,  Rotation  2  and  Rotation  3  

Define  the  orientation  of  the  buoy  relative  to  global  axes.  They  are  3  successive  rotations  that  take  the  global  axes   directions  to  the  buoy  axes  directions.  See  Initial  Position  and  Attitude  for  the  definition  of  these  angles.   Rotation  2  is  in  the  range  -­‐90°  to  +90°.  Range  jump  suppression  is  applied  to  Rotation  1  and  Rotation  3  (so  values   outside  the  range  -­‐360°  to  +360°  might  be  reported).   Azimuth,  Declination  

The  azimuth  and  declination  of  the  local  z  axis.   Velocity,  GX-­‐Velocity,  GY-­‐Velocity,  GZ-­‐Velocity  

The  magnitude  and  components  of  the  velocity  of  the  buoy,  relative  to  global  axes.  These  results  are  reported  at  a   user  specified  point  P  on  the  buoy.  The  point  P  is  specified  in  buoy  local  coordinates.   Angular  Velocity,  x-­‐Angular  Velocity,  y-­‐Angular  Velocity,  z-­‐Angular  Velocity  

The  magnitude  and  components  of  the  angular  velocity  of  the  buoy,  relative  to  buoy  axes.  

359  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

Acceleration,  GX-­‐Acceleration,  GY-­‐Acceleration,  GZ-­‐Acceleration  

The  magnitude  and  components  of  the  acceleration  of  the  buoy,  relative  to  global  axes.  These  results  are  r eported  at   a  user  specified  point  P  on  the  buoy.  The  point  P  is  specified  in  buoy   local  coordinates.   Acceleration  (incl.  g),  x-­‐Acceleration  (incl.  g),  y-­‐Acceleration  (incl.  g),  z-­‐Acceleration  (incl.  g)  

The   magnitude   and   components   (with   respect   to   buoy   axes)   of   the   vector   a   -­‐   g   where   a   is   the   acceleration   of   the   buoy  and  g  is  the  acceleration  due  to  gravity,  a  vector  pointing  vertically  downwards.  These  results  are  reported  at  a   user  specified  point  P  on  the  buoy.  The  point  P  is  specified  in  buoy  local  coordinates.   These  results  can  be  used  to  compare  against  accelerometer  readings.   Angular  Acceleration,  x-­‐Angular  Acceleration,  y-­‐Angular  Acceleration,  z-­‐Angular  Acceleration  

The  magnitude  and  components  of  the  angular  acceleration  of  the  buoy,  relative  to  buoy  axes.   Sea  Surface  Z  

The  global  Z  coordinate  of  the  sea  surface  directly  above  the  instantaneous  position  of  the  buoy  origin.   Dry  Length  

The  length  of  buoy  above  the  water  surface,  measured  along  the  buoy  z  axis,  calculated  as  follows:   x

For   a   Lumped   Buoy,   this   is   calculated   by   assuming   that   the   z-­‐extent   of   a   Lumped   Buoy   is   ½Height  either   side   of   its  centre  of  volume.  

x

For  a  Spar  Buoy  it  is  the  sum  of  the  dry  lengths  of  each  of  its  cylinders,  where  the  dry  length  of  an  individual   cylinder  is  calculated  as:   (cylinder  length)  ×  (cylinder  volume  above  surface)  /  (cylinder  total  volume).  

Force,  Moment   Lx-­‐Force,  Ly-­‐Force,  Lz-­‐Force,  Lx-­‐Moment,  Ly-­‐Moment,  Lz-­‐Moment,   Gx-­‐Force,  Gy-­‐Force,  Gz-­‐Force,  Gx-­‐Moment,  Gy-­‐Moment,  Gz-­‐Moment  

These  results  are  not  available  for  buoys  that  are  connected  to  other  objects  Ȃ  you  can  instead  use  the  Connection   Force  and  Connection  Moment  results.   These   results   are   the   total   force   and   moment   applied   to   the   buoy.  Force  and  Moment  report   the   magnitudes   of  the   loads.   The   Lx,   Ly   and   Lz   results   report   the   components   of  the   force   and   moment   in   the   local   axes   directions.   The   Gx,   Gy  and  Gz  results  report  the  components  of  the  force  and  moment  in  the  global  axes  directions.   Connection  Force,  Connection  Moment   Connection  x-­‐Force,  Connection  y-­‐Force,  Connection  z-­‐Force   Connection  x-­‐Moment,  Connection  y-­‐Moment,  Connection  z-­‐Moment  

These   connection   load   results   are   only   available   for   buoys   that   are  connected   to   other   objects.   They   report   the   total   force  and  moment  applied  to  the  buoy  by  the  object  to  which  it  is  connected.   Connection   Force   and   Connection   Moment   report   the   magnitudes   of   the   connection   loads.   The   other   results   report  the  components  of  the  connection  force  and  moment  in  the  buoy  axes  directions.   These   connection   force   and   moment   results   include   the   inertial   load   on   the   buoy   due   to   any   acceleration   of   the   object   to  which   it  is   attached.   This   means   that   these   results   can   be   used   for   sea   fastening   calculations,   by   using   a   6D   buoy   to   model   the   object   to  be   fastened   and   then   attaching   it   to   a   vessel.   The   connection   force   and   moment   include   the  weight  of  the  buoy  and   the  inertial  loads  due  to  the  vessel  acceleration.  Note  that  if  the  vessel  motion  is  specified   by   a   time   history   then   the   time   history   interpolation   method   used   is   important   since   it   affects   the   calculation   of   vessel  acceleration  and  hence  affects  the  inertial  loads.   Wing  Results   If  the  6D  buoy  has  wings  attached  then  for  each  wing  the  following  results  are  available.   Wing  X,  Wing  Y,  Wing  Z  

The  position  of  the  wing  origin,  relative  to  global  axes.   Wing  Azimuth,  Declination  and  Gamma  

The  orientation  angles  of  the  wing,  relative  to  the  buoy.  

360  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

Lift,  Drag,  Moment  

The  lift  force,  drag  force  and  drag  moment  applied  to  the  wing.   The   lift   force   is   applied   at   90°   to   the   relative   flow   direction.   Positive   values   mean   a   force   trying   to   push   the   wing   towards  its  positive  side,  negative  values  towards  its  negative  side.   The  drag  force  is  applied  in  the  relative  flow  direction  and  is  always  positive.   The  drag  moment  is  applied   about  the  line  that  is  in  the   wing  plane  and  at  90°  to  the  relative  flow  direction.  Positive   values  are  moments  trying  to  turn  the  wing  to  bring  the   wing  y-­‐axis  Wy  to  point  along  the  relative  flow  direction;   negative  values  are  moments  trying  to  turn  the  wing  the  opposite  way.   Note:  

When  the  wing  is  less  than  half  submerged,  and  you  have   included  wind  loads  on  wings,  then  the   lift  force,  drag  force  and  moment  reported  are  the  air  loads.  Otherwise  they  are  the  water  loads.  

Incidence  Angle  

Š‡ƒ‰Ž‡ǡȽǡ–Šƒ––Š‡”‡Žƒ–‹˜‡ˆŽ‘™˜‡…–‘”ƒ‡•™‹–Š–Š‡’Žƒ‡‘ˆ–Š‡™‹‰ǡ‹–Š‡”ƒ‰‡-­‐90°  to  +90°.  Positive  values   mean   that  the  flow   is  towards  the  positive   side   of   the   wing   (i.e.   hitting  the  negative  side)   and  negative   values   mean   that  the  flow  is  towards  the  negative  side  of  the  wing  (i.e.  hitting  the  positive  side).   The  value  reported  is  with  respect  to  the  principal  fluid  affecting  the  wing.   Beta  Angle  

The   angle   of  the   relative   flow   direction,   measured   in  the  wing   plane.   More   specifically,  it   is  the   angle  between  wing   Wx  axis  and  the  projection  of  the  relative  flow  vector  onto  the  wing  plane,  measured  positive  towards  Wz.  Zero  beta   angle   means   that   this   projection   is   in   the   Wx   direction,   90°   means   it   is   along   Wz   and   -­‐90°   means   it   is   along   the   negative  Wz  direction.   The  value  reported  is  with  respect  to  the  principal  fluid  affecting  the  wing.   Range  jump  suppression  is  applied  to  the  Beta  Angle  (so  values  outside  the  range  -­‐360°  to  +360°  might  be  reported).  

6.9.17

Buoy  Hydrodynamics  

3D   and   Lumped   6D   buoys   are   generalised   objects  for   which  no  geometry  is  defined  in   the   data  other   than   a   height:   This  is  used  for  proportioning  hydrodynamic  properties  when  the  object  is  partially  immersed,  and  for  drawing  a   3D  buoy.   Since   the   geometry   of  the   object   is   undefined,   it   is   necessary   to   define   properties   such   as   inertias,   drag   areas,   added   masses,  etc.  explicitly  as  data  items.  This  can  be  a  difficult  task,  especially  where  a  6D  buoy  is  used  to  represent  a   complex  shape  such  as  a  midwater  arch  of  the  sort  used  to  support  a  flexible  riser  system.   We  cannot  give  a  simple  step-­‐by-­‐step  procedure  for  this  task  since  the  geometry  of  different  objects  can  be  widely   different.  As  an   example,   the  hydrodynamic  properties   in  6  degrees  of  freedom  are  derived  for  a  rectangular   box.   This   gives   a   general   indication   of   the   way   in   which   the   problem   should   be   approached.   If   a   3D  buoy   is   used,   the   rotational  properties  are  not  used.  

361  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

6.9.18

Hydrodynamic  Properties  of  a  Rectangular  Box  

O is the centre of the box

Z

Y

X

z

O

y

x  

Figure:  

Box  Geometry  

Drag  areas   In  X  direction:  Ax  =  y  .  z   In  Y  direction:  Ay  =  x  .  z   In  Z  direction:  Az  =  x  .  y   Drag  Coefficients  for  Translational  Motions   These   are   obtained   from   ESDU   71016,   Figure   1   which   gives   data   for   drag   of   isolated   rectangular   blocks   with   one   face  normal  to  the  flow.  The  dimensions  of  the  block  are   a  in  the  flow  direction   b  and  c  normal  to  the  flow  direction  (c>b).   The   figure   plots   drag   coefficient,   Cx   against   (a/b)   for   (c/b)   from   1   to   infinity   (2D   flow).   Cx   is   in   the   range   0.9   to   2.75   for  blocks  with  square  corners.   Note:  

ESDU   71016   uses   Cd   for   the   force   in   the   flow   direction;   Cx   for   the   force   normal   to   the   face.   For   present  purposes  the  two  are  identical.  

Drag  Properties  for  Rotational  Motions   There   is   no   standard   data   source.   As   an   approximation,   we   assume   that   the   drag   force   contribution   from   an   elementary  area  dA  is  given  by   † αΦǤɏǤ2.Cd.dA   where  Cd  is  assumed  to  be  the  same  for  all  points  on  the  surface.   Note:  

This  is   not  strictly  correct.  ESDU  71016  gives  pressure   distributions   for  sample   blocks  in   uniform   flow  which  show  that  the  pressure  is  greatest  at  the  centre  and  least  at  the  edges.  However  we  do   not  allow  for  this  here.  

362  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

Z

dz z

O

X

  Figure:  

Integration  for  rotational  drag  properties  

Consider  the   box  rotating  about  OX.  The  areas  Ay  and  Az   will  attract  drag  forces  which  will  result  in   moments  about   OX.  For  the  area  Ay,  consider  an  elementary  strip  as  shown:   ‘”ƒƒ‰—Žƒ”˜‡Ž‘…‹–›ɘƒ„‘—–ǡ–Š‡†”ƒ‰ˆ‘”…‡‘–Š‡•–”‹’‹•   † αΦǤɏǤȋɘœȌǤȁɘœȁǤ†ǤšǤ†œ   and  the  moment  of  this  force  about  OX  is   †αΦǤɏǤȋɘœȌǤȁɘœȁǤ†ǤšǤ†œǤœαȋΦǤɏǤɘǤȁɘȁǤ†ȌǤšǤœ3.dz   Total  moment  is  obtained  by  integration.  Because  of  the  V.|V|  form  of  the  drag  force,  simple  integration  from  -­‐Z/2  to   +Z/2  gives  M  =  0.  We  therefore  integrate  from  0  to  Z/2  and  multiply  the  answer  by  2.  The  result  is   M  αȋΦǤɏǤɘǤȁɘȁǤ†ȌǤȋšǤœ4/32)   OrcaFlex  calculates  the  drag  moment  by   αȋΦǤɏǤɘǤȁɘȁǤ†ȌǤȋȌ   so  we  set   Cdm  =  Cd,  AM  =  x.z4/32.   This  is  the  drag  moment  contribution  about  OX  from  the  Ay  area.  There  is  a  similar  contribution  from  the  Az  area.   Since  Cd  is  generally  different  for  the  2  areas,   it  is  convenient   to  calculate  the  sum  of  (Cd.AM)  for   both,  set  AM   equal   to  this  value  and  set  Cd  equal  to  1.   Added  Mass   OrcaFlex   requires   the   added   mass   and   inertia   contributions   to   the   mass   matrix,  plus   the   hydrodynamic   masses   and   inertias  to  be  used  for  computation  of  wave  forces.  For  each  degree  of  freedom  (3  translations,  3  rotations),  3  data   items   are   required.   These   are   Hydrodynamic   Mass   in   tonnes   (or   Inertia   in   tonne.m 2);   and   coefficients   Ca   and   Cm.   Added   mass   is   then   defined   as   Hydrodynamic   Mass  .  Ca;   and   wave   force   is   defined   as   (Hydrodynamic   mass  .  Cm)   multiplied  by  the  water  particle  acceleration,  aw.   On   the   usual   assumptions   intrinsic  in  the   use  of  Morison's   Equation  (that   the   body   is   small   by  comparison  with  the   ™ƒ˜‡Ž‡‰–ŠȌǡ–Š‡™ƒ˜‡ˆ‘”…‡‹•‰‹˜‡„›ȋȟΪ  AM)  .  awǡ™Š‡”‡ȟ‹•„‘†›†‹•’Žƒ…‡‡–ƒ†‹•ƒ††‡†ƒ••Ǥ”…ƒ Ž‡š calculates  the  wave  force  as  Cm  .  HM  .  aw  where  HM  is  the  Hydrodynamic  Mass  given  in  the  data.   ‘” –”ƒ•Žƒ–‹‘ƒŽ ‘–‹‘•ǡ •‡–  α ȟ ˆ‘” ƒŽŽ †‡‰”‡‡• ‘ˆ ˆ”‡‡†‘Ǥ Š‡ ƒ  =  Ȁȟǡ   =  1  +  Ca.   For   rotational   ‘–‹‘•ǡ•‡– αȟ ǡ–Š‡‘‡–‘ˆ‹‡”–‹ƒ‘ˆ–Š‡†‹•’Žƒ…‡†ƒ••ǤŠ‡ƒ  =   Ȁȟ ǡ  =  1  +  Ca  where  AI  is  the  added   inertia  (i.e.  the  rotational  analogue  of  added  mass).   Translational  Motion   DNV-­‐RP-­‐C205,  Table  6.2,   gives  added  mass  data  for  a  square  section   prism  accelerating  along  its  axis.  The  square   section   is   of  side   a,   prism   length   is   b,   and   data   are   given   for   b/a   =   1.0   and   over.   The   reference   volume   is   the   volume  

363  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

of  the  body  which  is  the  same  definition  we  have  adopted.  We  can  therefore  use  the  calculated  Ca  without  further   adjustment.   Consider  the  X  direction:  Area  normal  to  flow  =  Ax.   ‘”ƒ•“—ƒ”‡‘ˆ–Š‡•ƒ‡ƒ”‡ƒǡƒαξȋšȌǤ   Length  in  flow  direction  =  x.   ‡…‡„ȀƒαšȀξȋšȌǤ   Hence  Ca  can  be  obtained  from  DNV-­‐RP-­‐C205  by  interpolation,  and  then    Cm  =  1  +  Ca.   If   b/a   <   1.0   this   approach  fails   and   we   use   the   data   given   in  DNV-­‐RP-­‐C205  for   rectangular   flat   plates.   If   y  >   z,  aspect   ratio  of  the   plate   =   y/z.   Hence   CA   from  DNV-­‐RP-­‐C205  by   interpolation.   The   reference   volume   in   this   case   is   that   of  a   cylinder  of  diameter  z,  length  y.  Hence:   ††‡†ƒ••αǤɏǤȋɎȀͶȌǤ›Ǥœ2  =  AMx,  say   ƒ†–Š‡ƒαšȀȟƒ†αͳΪƒǤ   Note:  

If  y  <  z,  then  aspect  ratio  =  z/y  and  reference  volume  =  CA  .  ɏǤ  ȋɎȀͺȌǤœǤ›2.  

Rotational  Motion   DNV-­‐RP-­‐C205  gives  no  data  for  hydrodynamic  inertia  of  rotating  bodies.  The  only  data  for  3D  solids  we  know  of  is   for   spheroids   (Newman   1977).   Fig   4.8   of   Newman   1977   gives   the   added   inertia   for   coefficient   for   spheroids   of   varying  aspect  ratio  referred  to  the  moment  of  inertia  of  the   displaced  mass.  We  assume  that  the  same  coefficient   applies  to  the  moment  of  inertia  of  the  displaced  mass  of  the  rectangular  block.   Rotation  about  X   ȟ αȟȋ2  +  Z2)/12   Added  inertia:  

Using  data  for  spheroids  from  Newman  1977  :   Length  in  flow  direction  =  2a  =  x,  so  a  =  x/2.   “—‹˜ƒŽ‡–”ƒ†‹—•‘”ƒŽ–‘ˆŽ‘™ǡ„ǡ‹•‰‹˜‡„›Ɏ„2  α›œǡ•‘„αξȋ›  .  œȀɎȌǤ   Hence  Ca  from  Newman  1977.   For  b/a  <  1.6  

Ca   can   be  read   from   the   upper  figure  where   the   value  is  referred   to   the   moment   of  inertia   of  the  displaced   mass.   In   this  case  no  further  adjustment  is  required.   For  b/a  >  1.6  

The  coefficient  CA  is  read  from  the  lower  graph  in  which  the  reference  volume  is  the  sphere  of  radius  b.  In  this  case:   Ca  =  CA  .  (2  .  b3)/(a  .  (a2+b2))   In  either  case,  Cm  =  1  +  Ca.  

6.9.19

Modelling  a  Surface-­‐Piercing  Buoy  

Surface-­‐piercing   buoys,   such   as   CALM   buoys,   SPAR   buoys   or   meteorological   discus   buoys,   can   be   modelled   in   OrcaFlex   using   the  Spar   Buoy  version  of   a  6D  Buoy.   Despite   its  name,  the  OrcaFlex  Spar   Buoy   can  be   used   to  model   any  axi-­‐symmetric  body.   Spar   Buoys   have   many   data   items   available.   This   enables   you   to   model   a   wide   range   of   effects,   but   it   also   makes   setting  up  a  Spar  Buoy  model  more  complicated.  To  help  in  this  task  we  describe,  in  this  section,  the  approach  we   adopt  for  setting  up  an  OrcaFlex  model  of  a  surface-­‐piercing  buoy.   1.  Create  a  simple  model  containing  just  a  Spar  Buoy   Start  by  modelling  the  free-­‐floating  behaviour  of  the  buoy,  without  any  lines  attached.  This  allows  us  to  get  the  basic   behaviour  of  the   buoy  correct,  before  complications  such  as  moorings  etc.  are   introduced.   We  therefore  set  up   an   OrcaFlex  model  containing  just  a  Spar  Buoy  and  with  no  waves  or  current.  

364  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys  

 

Set  the  buoy's  Applied  Load  to  zero.  This  data  allows  you  to  apply  extra  forces  and  moments  to  the   buoy,  in  addition   to   those  from   any  lines   that  you   attach  to   it.   You   can   use   this   later   to   model   the  wind   force  on   the   upper   part  of   the   buoy.   To  do  this  you  will  need  to  know  the  projected  area  (i.e.  the  area  exposed  to   wind)   of  the   pipe   work   etc.  in  the   upper  part  of  the  buoy.   Set  the  buoy's  Munk  Moment  Coefficient  to  zero.  This  data  item  is  only  used  for  slender  bodies  in  near  axial  fully-­‐ submerged  flow  only.   Set  the  number  of  wings  to  zero.  Wings  are  normally  only  relevant  for  towed  fish.   Finally,  we  start  by  setting  all  the  buoy's  drag  and  added  mass  data  to  zero.  We  will  set  up  the  actual  values  later.   2.  Set  up  the  geometry  data   The  Spar  Buoy  has  its  own  local  buoy  axes  coordinate  system.  You  can  choose  the  origin  for  these   buoy  axes,  but   the   buoy   z-­‐axis   direction   must   be   chosen   to   be   parallel   to   the   axis   of   the   buoy,   and   the   buoy   x-­‐   and   y-­‐axes   should   be   parallel   to  the   other   two   principle   directions   of  structural   inertia.   You   also   need   to   specify   the   position   of  the   centre   of  gravity,  relative  to  the  local  coordinate  system.  The  buoy  manufacturer  should  supply  this  information.   Set  the  Stack  Base  Position.  This  is  the  position  of  the  centre  of  the  bottom  of  the  buoy,  relative  to  the  buoy  axes.  The   Stack  Base  Position  therefore  has  z-­‐coordinate  =  -­‐h,  where  h  is  the  distance  axially  from  the  bottom  of  the  buoy  to   the  buoy  origin.   Now  set  up  a  number  of  cylinders,  and  their  lengths  and  diameters,  in  order  to  model  the  shape  of  the  buoy.  To  do   this   you   need   the   dimensions   of   the   various   parts   of   the   buoy.   The   buoy   manufacturer   should   supply   this   information.  Set  the  cylinder  lengths  and  diameters  so  that  you  get  the  correct  length  and  volume  for  each  section.   You  can  represent  tapered  sections  by  a  series  of  short  cylinders  with  diameters  changing  progressively  from  one  to   the  next.   We   recommend   using   a   number   of   short   cylinders,   even   where   the   buoy   diameter   is   constant   over   a   long   length.   Using  more  cylinders  gives  more  accurate  results,  though  at  the  cost  of  reduced  computation  speed.   You  can  check  your  geometry  data  by  zooming  in  on  the  buoy  in  a  3D  View  window.  Turn  on  the  local  axes  so  that   you  can  check  that  the  buoy  origin  is  in  the  correct  place.   The  Bulk  Modulus  data  item  is  not  relevant  to  a  surface-­‐piercing  buoy,  so  it  can  be  left  at  the  default  value  of  Infinity.   3.  Set  up  the  mass  and  inertia  data   Now  set  the  Mass  and  Moments  of  Inertia  of  the  buoy.  The  buoy  manufacturer  should  supply  this  information.   The   mass   equals   the   weight  of  the   buoy   in   air.   The   moments   of   inertia   are   those   of  the   buoy   (in   air)   about   its   centre   of  gravity,  as  follows:   x

Iz  =  the  moment  of  inertia  about  the  buoy  axis.  

x

Ix   and   Iy   =   the   moments   of   inertia   about   axes   perpendicular   to   the   buoy   axis,   through   the   centre   of   gravity.   Usually  it  is  sufficient  to  assume  that  Ix  =  Iy.  

If  you  cannot  obtain  data  for  the  moments  of  inertia,  then  they  can  be  approximately  calculated  from  a  knowledge  of   the  masses  of  the  various  parts  of  the  buoy,  and  approximately  how  that  mass  is  distributed.   4.  Check  that  the  buoy  floats  at  the  correct  draught   Set  the  Initial  Position  and  Initial  Attitude  of  the  buoy  so  that  the  buoy  is  in  its  expected  equilibrium  position.  The   initial   position   is   the   position   of   the   buoy   local   origin,   and   therefore   of   the   CG,   and   you   can   calculate   this   point's   expected  equilibrium  position  from  the  buoy  draught,  which  should  be  available  from  the  buoy  manufacturer.   The  Initial  Attitude  defines  the  initial  orientation  of  the  buoy.  Set  it  to  (0,0,0),  which  orients  the  buoy  with  its  axis   vertical  and  the  buoy  local  x,y  axes  aligned  with  the  global  X,Y  axes.   Set  the  Degrees  of  freedom  included  in  statics  to  None  and  then  run  the  simulation  and  look  at  the  time  history  of   buoy   Z.   If   the   data   has   been   set   up   correctly   then   the   buoy   should   have   stayed   basically   in   its   initial   position   and   attitude,  with  perhaps  just  small  oscillations  about  that  position.   If   the   buoy   Z   has   oscillated   significantly   then   the   model's   equilibrium   position   does   not   match   the   expected   equilibrium  position.  This  means  that  something  is  wrong  in  the  data  and  this  needs  tracing  and  correcting  before   you   proceed.   You   can   estimate   the   model's   equilibrium   position   by   looking   at   the   mean   Z   position   in   the   time   history.  

365  

w

 

System  Modelling:  Data  and  Results,  6D  Buoys    

5.  Check  that  the  buoy  is  stable   Now  check  that  the  buoy  is  stable   -­‐  i.e.  that  if  it  is  pitched  over  to  one  side  and  released  then  it  rights  itself.  In  the   Initial  Attitude  data,  set  the  Rotation  2  value  to  say  10°  and  run  the  simulation.  If  the  buoy  falls  over  then  there  is   something  wrong  with  the  CG  position  or  the  volume  distribution,  and  this  must  be  corrected.   Note:  

The   buoy   on   its   own   may   not   be   intended   to   be   stable,   e.g.   stability   may   only   be   achieved   when   the   moorings  are  attached.  In  this  case  you  will  need  to  model  the  moorings  in  order  to  check  stability.  

6.  Set  the  Added  Mass  data   The  x  and  y  added  mass  coefficients  can  be  set  to  1.0,   which  is  the  standard  value   for   a  cylinder  in  flow   normal  to  its   axis.   Added  mass  in  the  z  direction  should  be  estimated  for  the  buoy  from  the  published  literature  (DNV  rules,  Barltrop  &   Adams,   1991)   and   distributed   between   the   immersed   cylinders   (remember   that   hydrodynamic   loads   are   only   applied  to  the  immersed  parts  of  the  model).   Ideally,   this   data   should   then   be   checked   by   comparing   the   heave   and   pitch   natural   periods   of   the   model   against   values  obtained  from  model  tests  or  full  scale  measurements,  and  adjustments  made  as  necessary.   7.  Set  the  drag  and  damping  data   The  best  approach  depends  on  whether  the  buoy  is  a  SPAR  whose  length  is  great  by  comparison  with  its  diameter,   or   a   surface-­‐following   Discus   shape   such   as   an   oceanographic   buoy.   CALM   buoys   are   usually   closer   to   the   Discus   configuration,  often  with  a  damping  skirt  which  is  submerged  at  normal  draft.   Spar  Buoys  

Set   the   Drag   Areas   for   each   cylinder   to   the   areas,   of   the   part   of   the   buoy   which   that   cylinder   represents,   that   are   exposed   to   fluid   drag   in   the   direction   concerned.   Note   that   you   should   specify   the   areas   that   are   exposed   to   drag   when   the   buoy   is   fully   submerged.   OrcaFlex   automatically   calculates   the   proportion   of   the   cylinder   that   is   submerged  and  scales  all  the  fluid  loads  on  the  cylinder   using  that  'proportion  wet'  as  a  factor.   So  if  a  cylinder  is   not   submerged,  or  is  partially  submerged,  then  the  drag  loads  will  be  scaled  accordingly  for  you.   For   a   simple   cylinder,   of   diameter   D   and   length   L,   the   normal   drag   area   is   D.L   since   that   is   the   area   of   a   cylinder   ™Š‡˜‹‡™‡†‘”ƒŽ–‘‹–•ƒš‹•Ǥ†–Š‡ƒš‹ƒŽ†”ƒ‰ƒ”‡ƒ‹•ȋɎǤ 2)/4  since  that  is  the  area  of  the  cylinder  when  viewed   along   its   axis.   However,   where   a   cylinder   is   representing   part   of  the   buoy   that   is   not   in   reality   a   simple   cylinder   (for   example,  we  may  represent  the  pipework  and  turntable  on  the  deck  of  a  SPAR  buoy  as  an  equivalent  cylinder)  or   where  the  cylinder  is  shielded  from  drag  by  adjacent  structure,  then  the  drag  areas  should  be  set  accordingly.  For   example,  if  the  cylinder  is  shielded  below  by  another  cylinder  of  diameter  d  (less  than  D)  then  the  axial  drag  area   •Š‘—Ž†„‡”‡†—…‡†„›ȋɎǤ†2)/4  to  model  that  shielding.   Set   the   Drag   Force   Coefficient   based   on   values   given   in   the   literature.   For   short   simple   cylinders   fully   immersed   there   are   standard   values   given   in   the   literature   (see   Barltrop   &   Adams,   1991,   Hoerner,1965   and   DNV-­‐RP-­‐C205).   However,  the  standard  book  values  do  not  include  energy  absorption  by  wave-­‐making  at  the  free  surface.  Strictly,   this   is   a   linear   term   (forces   directly   proportional   to   velocity),   but   in   OrcaFlex   this   must   be   done   by   adjusting   the   drag  coefficients  of  one  or  more  cylinders.   The   Unit   Damping   Force   data   can   be   set   to   zero.   If   you   later   find   that   the   buoy   shows   persistent   small   amplitude   oscillations  then  you  may  wish  to  set  a  non-­‐zero  value  to  damp  this  out.   Set   the   Drag   Area   Moments,   Drag   Moment   Coefficients   and   Unit   Damping   Moment   data.   For   the   normal   direction   these   data   items   can   usually   all   be   left   as   zero,   providing   you   have   subdivided   the   buoy   into   short  enough   cylinders   (since  these  terms  involve  a   high  power   of  L,   the  cylinder  length).  For  the  axial  direction   these  data  items  model   the   yaw  drag  and  damping  effects,  so  if  this  is  important  to  you  then  set  them  to  model  the  two  main  sources,  namely   skin  friction  on  the  cylinder  surface  and  form  drag  on  any  protuberances  on  the  buoy.   Having  set  up  this  drag  and  damping  data,  it  is  well  worth  now  running  simulations  of  heave  and  pitch  oscillations   and  checking  that  their  rate  of  decay  is  reasonable  and  consistent  with  any  real  data  you  have  available.   Discus  and  CALM  Buoys  

These   types   of   buoy   require   different   treatment   since   they   have   little   axial   extension.   Instead   it   is   their   radial   extension   that   most   affects   the   buoy's   pitch   properties.   As   a   result   the   axial   discretisation   of  the   buoy   into   cylinders   does   not  capture   the   important   effects.   For   example   the   pitch   damping   is   often   mostly   due   to   radiation   damping,   i.e.   surface  wave  generation;  this  is  especially  important  for  a  CALM  buoy  with  a  skirt.  

366  

w

 

System  Modelling:  Data  and  Results,  3D  Buoys  

 

To   deal   with   this   OrcaFlex   offers   the   rotational   drag   and   damping   data,   but   there   is   little   information   in   the   literature  to  help  in  setting  up  this  data.  We  therefore  strongly  recommend  that  you  set  the  data  up  by  calibration   against   real   test   results   from   model   or   full   scale   tests.   The   easiest   information   to   work   with   are   time   history   graphs   of   the   buoy   heave   and   pitch   in   still   water,   starting   from   a   displaced   position.   This   will   give   the   heave   and   pitch   natural  periods  and  the  rates  of  decay  and  you  can  adjust  the  buoy's   drag  and  damping  data  until  you  get  a   good   match  with  this  measured  behaviour.   Here  is  the  approach  we  use:   x

For   the   normal   direction,   set   the   Drag   Area,   Drag   Force   Coefficient   and   Unit   Damping   Force   as   described   for   Spar  buoys  above.  

x

Then  set  the  axial  Unit  Damping  Force  to  zero  and  run  a  simulation  that  matches  the  conditions  that  existed  in   the  real  heave  time  history  results,  i.e.  with  the  same  initial  Z  displacement.  

x

Then  adjust  the  axial  Drag  Area  and  Drag  Force  Coefficients  until  the  OrcaFlex  buoy's  Z  time  history  matches  the   real   time   history.  These   two   data   items   are   simply   multiplied   together   when   they   are   used   to   calculate   the   drag   force,  so  you  can  give  one  of  the  two  data  items  a  fixed  positive  value  (e.g.  1)  and  then  adjust  the  other.  

x

The  match  will  probably  be  poor  in  the  later  parts  of  the  time  history,  where  the  heave  amplitude  has  decayed   to   small   values.   This   is   because   the   square   law   drag   term   is   insignificant   at   small   amplitude   and   instead   the   damping   force   takes   over.   Therefore   we   now   adjust   the   axial   Unit   Damping   Force   to   further   improve   the   match   where   the  amplitude   is  small.  You  may  find  that  this  disturbs  the  match  in   the  large  amplitude  part,  in   which   case  you  might  need  to  readjust  the  drag  data.  

x

For   the   axial   direction,   set   the   Drag   Area   Moment,   Drag   Moment   Coefficient   and   Unit   Damping   Moment   as   described  for  Spar  buoys  above.  

x

Then   set   the   normal   Drag   Area   Moment,   Drag   Moment   Coefficient   and   Unit   Damping   Moment   to   best   match   the   real  pitch  time  history,  in  a  similar  way  to  that  used  above  to  match  the  heave  time  history.  

6.10

3D  BUOYS  

OrcaFlex  3D  Buoys  are  simplified  point   elements  with   only  3  degrees  of  freedom:  X,  Y  and  Z.  They  do  not  rotate,  but   remain  aligned   with  the  global  axes.  They  therefore  do   not  have  rotational  properties  and   moments  on  the   buoy  are   ignored.  They  should  therefore  be  used  only  where  these  limitations  are  unimportant.   3D   Buoys   are   able   to   float   part-­‐submerged   at   the   surface,   and   may   also   be   used   independently,   with   no   lines   attached.  Although  they  are  much  less  sophisticated  than  6D  Buoys,  3D  Buoys  are  easier  to  use  and  are  convenient   for  modelling  buoys  at  line  junctions  etc.  

367  

w

 

System  Modelling:  Data  and  Results,  3D  Buoys    

height/2

z B

Buoy Axes always aligned with Global Axes

y x

height/2

  Figure:  

6.10.1

3D  Buoy  

Data  

Name  

Used  to  refer  to  the  3D  Buoy.   Included  in  Static  Analysis  

Determines  whether  the   equilibrium  position  of  the  buoy  is  calculated  by  the  static  analysis.  See   Buoy  Degrees  of   Freedom  Included  in  Static  Analysis.   Initial  Position  

Specifies   the   initial   position   for   the   buoy   origin   as   coordinates   relative   to   the   global   axes.   If   the   buoy   is   not  included   in   the   static   analysis  then   this   initial   position   is  taken   to   be   the   static   position   of   t he  buoy.   If   the  buoy   is  included   in   the   static   analysis,   then   this   initial   position   is   used   as   an   initial   estimate   of   the   buoy   position   and   the   statics   calculation  will  move  the  buoy  from  this  position  iteratively  until  an  equilibrium  position  is  found.  See  Buoy  Degrees   of  Freedom  Included  in  Static  Analysis.   Mass  

Mass  or  weight  in  air.   Volume  

Used  to  calculate  buoyancy  and  added  mass.   Bulk  Modulus  

Specifies  the  compressibility  of  the  buoy.  If   the  buoy  is   not  significantly  compressible,  then  the  Bulk  Modulus  can  be   set  to  Infinity,  which  means  "incompressible".  See  Buoyancy  Variation.   Height  

Used  to  model  floating  buoys  correctly,  where  the  buoyancy,  drag  etc.  vary  according  to  the  depth  of  immersion.  It   also  determines  the  height  used  to  draw  the  buoy.  The  Height  is  the  vertical  distance  over  which  the  fluid-­‐related   forces  change  from  zero  to  full  force  as  the  buoy  pierces  the  surface.  It  is  taken  to  be  symmetrical  about  the  buoy's   origin.  

368  

w

 

System  Modelling:  Data  and  Results,  3D  Buoys  

 

Seabed  Friction  Coefficient  

OrcaFlex   applies   Coulomb   friction   between   the   buoy   and   the   seabed.   The   friction   fo”…‡ ƒ’’Ž‹‡† ‡˜‡” ‡š…‡‡†• Ɋ ™Š‡”‡‹•–Š‡•‡ƒ„‡†”‡ƒ…–‹‘ˆ‘”…‡ƒ†Ɋ‹•–Š‡ˆ”‹…–‹‘…‘‡ˆˆ‹…‹‡–Ǥ   See  Friction  Theory  for  further  details  of  the  friction  model  used.   Note:  

The  friction  coefficient  for  contact  with  elastic  solids  is  specified  on  the  Solid  Friction  Coefficients   data  form.  

Contact  Area  

Used   to   determine   contact   forces   when   the   buoy   comes   in   to   contact   with   the   seabed   and   with   elastic   solids.   If   a   value  of  '~'  is  specified  then  OrcaFlex  uses  Volume  /  Height.   Specifying  a  value  of  zero  allows  you  to  disable  contact  for  the  buoy.   Drag  

Drag  forces  are  applied  in  each  of  the  global  axes  directions  GX,  GY  and  GZ.  For  each  direction  you  must  specify  a   Drag  Coefficient  and  Drag  Area.   Added  Mass  

You   must   specify   the   added   mass   coefficient   Ca   for   each   global   axis   direction.   The   added   mass   is   set   to   be   Ca   multiplied  by  the  mass  of  water  currently  displaced.  The  inertia  coefficient,  Cm,  is  set  automatically  to  equal  1+Ca.  

6.10.2

Properties  Report  

The  3D  Buoy  properties  report  is  available  from  the   popup-­‐menu  on  the  data  form.  It  reports  the  following:   Weight  in  air  

The  force  due  to  gravity  acting  on  the  buoy's  mass.   Displacement  

The  weight  of  water  displaced  by  the  buoy's  volume.  The  reported  value  uses  the  water  density  at  the  sea  surface.   Weight  in  water  

Equals  Weight  in  air  -­‐  Displacement.   Statics  force  accuracy,  Statics  moment  accuracy  

These  are  only  reported  if  the  buoy  is  included  in  statics.   The  static  analysis  searches  for  an  equilibrium  position  for  the  buoy  Ȃ  that  is  a  position  for  which  the  resultant  force   on   the   buoy   is   zero.   We   refer   to   the   resultant   force   as   the   out   of   balance   load.   Because   computers   have   limited   numerical  precision  the  static  analysis  cannot  always  find  a  configuration  where  the  out  of  balance  load  is  exactly   zero.   OrcaFlex   accepts   a   position   as   a   static   equilibrium   position   if   the   largest   out   of   balance   load   component   is   less   than  the  statics  accuracy.   The  Statics  force  accuracy  equals   Tolerance   *  buoy  typical  force.  The  buoy  typical  force  is  based  on  the  weight  in   air,   the  displacement  and  the  forces  applied  by  connected  objects.   Reducing   the   Tolerance   value   will   give   a   more   accurate   static   equilibrium   position,   but   will   take   more   iterations.   OrcaFlex   may   not   be   able   to   achieve   the   Tolerance   specified   if   it   is   too   small,   since   the   computer   has   limited   numerical  precision.   Note:  

6.10.3

The   statics   accuracies   change   during   the   static   analysis   because   the   forces   applied   by   connected   objects   vary   with   the   position   of   the   buoy.   The   statics   accuracies   reported   in   Reset   state   may   be   quite   different   from   those   used   for   the   final   equilibrium   position   and   should   be   treated   as   rough   approximations  to  the  true  statics  accuracies.  

Results  

For  details  on  how  to  select  results  variables  see  Selecting  Variables.   For  3D  Buoys  the  available  variables  are:  

369  

w

 

System  Modelling:  Data  and  Results,  Winches    

X,Y  and  Z  

Positions  of  the  buoy  origin,  relative  to  global  axes.   Velocity,  X  Velocity,  Y  Velocity,  Z  Velocity   Acceleration,  X  Acceleration,  Y  Acceleration,  Z  Acceleration  

The  magnitude  and  components  (with  respect  to  global  axes)  of  the  velocity  and  acceleration  of  the  buoy.   Surface  Z  

The  global  Z  coordinate  of  the  sea  surface  directly  above  the  instantaneous  position  of  the  buoy  origin.   Dry  Length  

Length  of  buoy  above  the  water  surface,  measured  along  the  buoy  z  axis.  For  this  purpose,  the  z -­‐extent  of  a  3D   buoy   is  assumed  to  be  Height/2  either  side  of  its  volume  centre.  

6.11

WINCHES  

Winches   provide   a   way   of   modelling   constant   tension   or   constant   speed   winches.   They   connect   two   (or   more)   points  in  the  model  by  a  winch   wire,  fed  from  a  winch   inertia  (typically  representing  a  winch  drum)  that  is  then   driven  by  a  winch  drive  (typically  representing  the  winch  hydraulics  that  d rive  the  drum).   As  well  as  connecting   its  two  end  points,  the   winch   wire  may,  optionally,  pass  via  intermediate   points,  in   which  case   it  does   so   as   if   passing   over   a   small   frictionless   pulley   at   that   point.   The   wire   tension   either   side   of  the   intermediate   point   is  then   applied   to   that   point;   if   the   point   is  offset   on   the   object   involved   then   this   also   gives  rise   to   an   applied   moment.  

Winch may pull via intermediate objects Drive Force f Winch Drive

t

t

Winch Inertia

t

Wire Tension t

Winch wire   Figure:  

Winch  Model  

Two  types  of  winch  are  available  in  OrcaFlex:   Simple  Winches  

Simple   Winches  model  perfect  constant  tension   or  constant  speed  performance  and   are   easiest  to  use.  It   is  assumed   that  the  winch  inertia  is  negligible  and  the   winch  drive  is  perfect,  so  that  it  always  exactly  achieves  the  requested   constant  tension  or  constant  speed.  Because  of  these  assumptions,  no  data  needs  to  be  given  for  the  winch  inertia  or   winch  drive.   Detailed  Winches  

Detailed  Winches  include  modelling  of  the  performance  of  the  winch  drive  system  Ȃ  its  deadband,  stiffness,  inertia,   damping  and  drag  Ȃ  but  therefore  require  more  data  and  are  harder  to  set  up.   We   recommend  using   Simple  winches  unless  you  know  the  characteristics  of  the   winch  drive  system  and  believe   that  its  performance  significantly  differs   from  the  constant  tension   or  speed  ideal.  In   particular,  Simple  winches   are   appropriate:   x

At  the  early  design  stage,  when  the  type  of  winch  to  be  used  has  not  yet  been  d ecided.  

370  

w

 

System  Modelling:  Data  and  Results,  Winches  

 

x

If  the  duty  is  such  that  the  winch  drive  will  give  near  to  perfect  constant  tension  or  constant  speed  performance.  

x

If  the  winch  drive  data  are  not  available.  

Winch  Control   OrcaFlex   winches   allow   quite   complex   offshore   operations   to   be   modelled.   The   winch   drive   can   be   operated   in   either  of  two  modes:   Length  Control  Mode  

For  modelling  constant  speed  winches.  The  winch  wire  is  paid  out  or  hauled   in  at  a  velocity  specified  in  the  data.   Force  Control  Mode  

For   modelling   tension   controlled   winches.   Since   such   winches   are   usually   hydraulic   devices   whose   performance   deviates   quite   seriously   from   the   target   tension   ideal,   OrcaFlex   Winches   provides   facilities   for   modelling   winch   deadband,   damping   and   drag   forces   (force   decrements   proportional   to   velocity   and   velocity 2   respectively)   and   winch  stiffness  effects  such  as  those  caused  by  hydraulic  accumulators.   The   winch   can   be   switched   between   these   two   modes   at   predetermined   times   during   the   simulation   and   the   constant  velocity  or  target  tension  can  also  be  varied.  

6.11.1

Data  

Name  

Used  to  refer  to  the  Winch.   Type  

May  be  either  Simple  or  Detailed.  See  Winches.   Connect  to  Object  and  Object  Relative  Position  

The  (mass-­‐less)  winch  wire  connects  at  least  two  objects,  one  at  each  end  of  the  winch  wire.   If   more   than   2   are   specified   then   the   winch   wire   passes   from   the   first   connection   point   to   the   last   via   the   intermediate   points   specified.   When   intermediate   connections   are   specified,   the   winch   wire   slides   freely   through   these   intermediate   points   as   if   passing   via   small   friction-­‐less   pulleys   mounted   there.   The   winch   wire   tension   on   either   side   then   pulls   on   the   intermediate   points,   so   applying   forces   and   moments   (if   the   points   are   offset)   to   the   objects  concerned.   Each   connection   is   defined   by   specifying   the   object   connected   and   the   object-­‐relative   position   of   the   connection   point.   For   connecting   to   a   Line,   the   object-­‐relative   z   coordinate   specifies   the   arc   length   to   the   connection   point.   The   z   coordinate  specifies  the  arc  length  along  the  Line  and  this  arc  length  may  be  measured  relative  to  either  End  A  or   End   B   as   specified   by  the   user.   The   connection  point   is  attached   to   the   nearest  node.   If  torsion  is  not   modelled   then   the  x,y  coordinates  are  ignored  and  the  connection  point  is  at  the  centreline  of  the  Line.  If   torsion  is  modelled  then   the  x,y  coordinates  allow  you  to  offset  the  connection  from  the  centreline.   For  Fixed  connections  the  object-­‐relative  coordinates  given  are  the  global  coordinates  of  the  point.   For  connecting  to  an  Anchor,  the  object-­‐relative  x,y  coordinates  given  are  the  global  X,Y  coordinates  of  the  anchor   point,   and   the   z-­‐coordinate   is   the   distance   of   the   anchor  above   (positive)   or   below   (negative)   the   seabed   at   that   X,Y   position.   For   connecting   to  other  objects,   the   coordinates   of   the   connection   point   are   given   relative   to  the   object   local  frame   of  reference.   Release  at  Start  of  Stage  

The  winch  wire  can  be   released  at  the  start  of  a  given  stage  of  the  simulation,   by  setting  this  number  to  the  stage   number   required.   Once   released   the   winch   no   longer   applies   any   forces   to   the   objects   it   connects.   If   no   release   is   required,  then  set  this  item  to  '~'.  

6.11.2

Wire  Properties  

Wire  Stiffness  

Elastic  stiffness  of  the  winch  wire.  

371  

w

 

System  Modelling:  Data  and  Results,  Winches    

Wire  Damping  

Material  damping  value  for  the  winch  wire.   Note:  

The  mass  of  the  winch  wire  is  not  modelled.  

Winch  Inertia  (Detailed  Winches  only)  

The  inertia  of  the  winch  drive,  which  resists  changes  in  the  rate  of  pay  out  of  haul  in  of  the  winch  wire  if  the  winch  is   in  Force  Control  mode.  The  Winch  Inertia  has  no  effect  if  the  winch  is  in  Length  Control  mode.   This   is   a   linear,   rather   than   rotational,   inertia.   To   represent   the   rotational   inertia   of   a   winch   drum,   set   the   winch   inertia  to   l  /  r2   where   I  =  drum  rotational  inertia,   r  =  radius  at  which  the  wire  is  fed.   Notes:  

The   winch   inertia   does   not   contribute   to   the   mass   of   any   objects   to   which   the   winch   is   attached   and  so  does  not  directly  resist  acceleration  of  any  of  the  connection  points.  (Such  accelerations  are   resisted   indirectly,   of   course,   through   the   changes   they   cause   to   the   winch   wire   path   length   and   hence  to  the   winch   wire  tension.)  To   include  the  true  translational  inertia   of  the  winch   drive,  drum   and  wire  it  is  necessary  to  suitably  increase  the  masses  of  the  objects  to  which  it  is  attached.  

 

Setting   the   winch   inertia   to   a   small   value   to   model   a   low   inertia   winch   can   lead   to   very   short   natural  periods  for  the  winch  system.  These  then  require  very  short  time  steps  for  the  simulation,   slowing  the  simulation.  To  avoid  this,  the  winch  inertia  can  be  set  to  zero,   rather  than  to  a  small   value;  the  winch  system  inertia  is  then  not  modelled  at  all,  but  the  short  natural  periods  are  then   avoided.  See  Winch  Theory  for  full  details  of  the  algorithm  used  when  the  winch  inertia  is  zero.  

6.11.3

Control  

Control  Type  

Can  be  either  By  Stage  or  Whole  Simulation.   When   By   Stage   is   selected   the   winch   is   controlled   on   a   stage   by   stage   basis.   For   each   stage   of   the   simulation   you   choose  from  the  winch  control  modes.  These  modes  allow  you  to  control  the  winch  payout  rate,  control  the  rate  of   change  of  target  tension  or  specify  a  constant  target  tension.   Note:  

The   control   mode   remains   fixed   for   the   duration   of   each   stage.   Because   there   is   a   limit   on   the   number  of  stages  in  an  OrcaFlex  simulation  this  can  be  restrictive.  

When   Whole   Simulation   is   selected   the   winch   is   either   tension   controlled   or   length   controlled   for   the   whole   simulation.  For  the  tension  controlled  mode  the  target  tension  can  be  fixed,  vary  with  simulation  time  or  be  given  by   an   external   function.   Likewise   for   the   length   controlled   mode   the   payout   rate   of   unstretched   winch   wire   can   be   fixed,  vary  with  simulation  time  or  be  given  by  an  external  function.  

6.11.4

Control  by  Stage  

Winch  Control  for  Statics   For  the  static  analysis,  the  Mode  of  the  winch  drive  can  be  set  to  one  of  Specified  Length  or  Specified  Tension.   Specified  Length  

The  winch  drive  is  locked   with  the  unstretched  length  of  winch  wire  out,  L 0,  being  set  to  the  Value  specified.  The   winch  wire  tension  t  then  depends  on  the  stretched  length  L  of  the  winch  wire  path.   Specified  Tension  

The  winch  drive  operates  in   perfect  constant  tension  mode,  the  tension  t  being  the  Value  specified.  The  unstretched   length  out  L0  is  then  set  to  correspond  to  this  tension.  

372  

w

 

System  Modelling:  Data  and  Results,  Winches  

 

Winch  Control  for  Dynamics   During   the   simulation   the   winch   is   controlled   on   a   stage   by   stage   basis.   For   each   stage   the   winch   control   mode   can   be   set   to   one   of   Specified   Payout,   Specified   Payout   Rate,   Specified   Tension,   Specified   Tension   Change   or   Specified  Tension  Rate  of  Change.   Specified  Payout  

The   Value   specifies   the   unstretched   length   of   winch   wire   to   be   paid   out   (positive)   or   hauled   in   (negative)   at   a   constant  rate   during   this   stage.  That   is,  the  Value  specifies   the   total  change  in   unstretched   length  during   the   stage,   so  to  keep  a  constant  length  set  the  Value  to  zero.   Specified  Payout  Rate  

The  Value  specifies   the   rate   at   which  the  winch  wire   is   to   be   paid   out   (positive)   or   hauled   in   (negative)  during   this   stage.   Specified  Tension  

The  Value  specifies  the  target  constant  tension  for  this  stage.   For  Simple  winches   the   winch  drive   is  assumed   to   always   achieve   this   nominal   tension,   so   the  Value  is  used   as   the   actual  winch  wire  tension.   For   Detailed   winches   this   nominal   tension   is   used   as   the   target   tension   for   the   winch   drive,   which   then   applies   drive   force   to   the   winchinertia   to   try   to   achieve   this   target   tension.   The   algorithm   for   the   winch   drive   force   is   designed   to   model   the   characteristics   of   real-­‐world   winches   that   are   nominally   "constant   tension".   See   Winch   Theory.   Note:  

Changes  of  nominal  tension  are  applied  instantly  at  the  start  of  each  stage,  and  this  can  therefore   apply  a  shock  load  which,  if  large  enough,  may  affect  the  stability  of  the  simulation.  

Specified  Tension  Change  

The   Value   specifies   the   change   in   target   tension   for   this   stage.   That   is,   the   Value   specifies   the   total   change   in   nominal  tension  during  the  stage,  so  to  keep  at  a  constant  nominal  tension  set  the   Value  to  zero.   The  above  comments  for  the  Specified  Tension  control  method  also  apply  to  this  method.   Specified  Tension  Rate  of  Change  

The  Value  specifies  the  rate  of  change  in  target  tension  for  this  stage.   The  above  comments  for  the  Specified  Tension  control  method  also  apply  to  this  method.   Note:  

6.11.5

The   Specified   Tension   Change   and   Specified   Tension   Rate   of   Change   modes   allow   you,   f or   example,   to   model   a   constant   tension  winch  where   the   tension   is  determined  by   an   earlier   simulation   stage.   By   specifying   a   tension   change   of   0  following   a   payout   stage   you   can   lock   the   winch   at   the   tension   used  at  the  end  of  the  payout  operation.  

Control  by  Whole  Simulation  

Statics  Mode  

The  winch  control  data  for  statics  is  the  same  as  the  data  specified  in  the   By  Stage  control  type.   Dynamics  Mode  

The  winch  is  controlled  either  by  Specified  Tension  or  by  Specified  Payout  Rate.   If   Specified   Tension   is   set   then   you   must   also   specify   the   target   tension.   This   value   can   be   fixed,   vary   with   simulation  time  or  be  given  by  an  external  function.   If  Specified  Payout  Rate  is  set  then  you  must  also  specify  the  payout  rate  of  unstretched  winch  wire.  This  value  can   be  fixed,  vary  with  simulation  time  or  be  given  by  an  external  function.  

6.11.6 Note:  

Drive  Unit   The  drive  unit  data  applies  to  Detailed  Winches  only  

373  

w

 

System  Modelling:  Data  and  Results,  Winches    

Winch  Drive  

The   winch   drive   controls   the   winch   wire   in   one   of   two   winch   control   modes:   Length   Control   mode   ("Specified   Length",   "Specified   Payout"   or   "Specified   Payout   Rate")   or   Force   Control   mode   ("Specified   Tension",   "Specified   Tension  Change"  or  "Specified  Tension  Rate  of  Change").   x

Length  Control  mode  is  for  modelling  a  constant  speed  winch.  The  winch  tension  then  depends  simply  on  the   unstretched  length  of  winch  wire  out,  and  the  wire  properties  (Stiffness  and  Damping).  

x

Force  Control  mode  is  for  modelling  a  (nominally)  constant  tension   winch.  Because  such  winches  often  deviate   quite   seriously   from   the   constant   tension   ideal,   facilities   are   provided   for   modelling   winch   Deadband,   Damping,  Drag  and  Stiffness.  

Deadband  

A  deadband  of  +/-­‐  this  value  is  applied  to  the  winch  drive  force  between  hauling  in  and  paying  out  the  winch.  See   Winch  Theory  for  full  details.   Stiffness  

This  can  be  used  to  model,  for  example,  winch  hydraulic  accumulators.  It  is  the  rate  at  which  the  zero-­‐velocity  winch   force  (the  drive  force  applied  when  the  winch  is   neither   hauling  in   nor  paying  out)  varies   with   the  total  unstretched   length  of  winch  wire  paid  out.  See  Winch  Theory.   Damping  Terms  A  and  B  

These  terms  can  be  used  to  model  damping  in  a  winch's  hydraulic  drive  system.  The  winch  drive  force  is  taken  to   vary  with  haul-­‐in/payout  velocity  at  rates  A  and  B,  respectively.  See  Winch  Theory.   Drag  Terms  C  and  D  

These  terms  can  be  used  to  model  drag  in  a  winch's  hydraulic  drive  system.  The  winch  drive  force  is  taken  to  vary   with  haul-­‐in/payout  velocity2  at  rates  C  and  D,  respectively.  See  Winch  Theory.  

6.11.7

External  Functions  

Parameters  

This  data  item  specifies  the   External  Function  Parameters,  a  free   form  multi-­‐line  text   field   which  is  passed  to   any   external  function  used  by  the  Environment.  

6.11.8

Results  

For  details  on  how  to  select  results  variables  see   Selecting  Variables.   For  winches  the  available  variables  are:   X,  Y  and  Z  

The  global  coordinates  of  the  specified  winch  connection  point.   Tension  

The  tension  in  the  winch  wire.   Length  

The  unstretched  length  of  winch  wire  paid  out.   Velocity  

The  rate  of  pay  out  of  winch  wire.  Positive  value  means  paying  out,  negative  value  means  hauling  in.   Azimuth  and  Declination  

The  azimuth  and  declination  angles  of  the  direction  of  the  winch  wire  between  the  final  2  connection  points,  relative   to  the  global  axes.  This  direction  is  measured  from  the  last  connection  point  towards  the  previous  connection  point.   Declination  is  in  the  range  0°  to  180°.  Range  jump  suppression  is  applied  to  Azimuth  (so  values  outside  the  range  -­‐ 360°  to  +360°  might  be  reported).   Sea  Surface  Z  

The  global  Z  coordinate  of  the  sea  surface  directly  above  the  instantaneous  position  of  the  winch  mount.  

374  

w

 

System  Modelling:  Data  and  Results,  Links  

 

Connection  Force,  Connection  GX-­‐Force,  Connection  GY-­‐Force,  Connection  GZ-­‐Force  

The   magnitude   and   components   relative   to   global   axes   of   the   connection   force   at   the   specified   winch   connection   point.   We   adopt   the   convention   that   the   force   reported   is   that   applied   by   the   winch   to   the   object   to   which   it   is   connected.  

6.12

LINKS  

Links  are  simple  spring  or  spring/damper  connections  linking  two  points  in  the  model,  for  example  a  node  on  a  line   to  a   vessel,   or   a   buoy   to  an   anchor.   They   pull   the   two   points   together,   or   hold   them   apart,   with   a   force   that   depends   on  their  relative  positions  and  velocities.   Links   have   no   mass   or   hydrodynamic   loading   and   simply  apply  an   equal   and   opposite  force   to   the   two  points.   They   are   useful   for   modelling   items   such   as   wires   where   the   mass   and   hydrodynamic   effects   are   small   and   can   be   neglected;  for  example  buoy  ties  can  sometimes  be  modelled  using  links.   Two  types  of  Link  are  available:   Tethers  

Simple   elastic   ties   that   can   take   tension   but   not   compression.   The   unstretched   length   and   stiffness   of   the   tether   are   specified.  The  tether  remains   slack   and   does  not   apply  a   force   if   the   distance  between   the  ends  is   less  than  the  unstretched  length.   Spring/Dampers  

Combined  spring  and  independent  damper  units.   The  spring  can  take   both  compression  and   tension  and  can   have  either  a  linear   or  a  piecewise-­‐linear  length-­‐force  relationship.   The  damper  velocity-­‐force   relationship   can  also  be  either  linear  or  piecewise-­‐linear.  

Tether:

Spring-Damper:

  Figure:  

6.12.1

Types  of  Link  

Data  

Name  

Used  to  refer  to  the  Link.   Type  

may  be  either:   x

Tether:  a  simple  elastic  tie  having  linear  stiffness  and  no  damping.  

x

Spring/Damper:   a   combined   spring   and   independent   damper,   each   of   which   can   be   either   linear   or   piecewise-­‐ linear.  

Connect  to  Object  and  Object  Relative  Position  

Specifies  the  objects  to  be  linked.   For   connecting   to   a   Line,   the   object-­‐relative   z   coordinate   specifies   the   arc   length   to   the   connection   point.   The   z   coordinate  specifies  the  arc  length  along  the  Line  and  this  arc  length  may  be  measured  relative  to  either  End  A  or   End   B   as   specified   by  the   user.   The   connection  point   is  attached   to   the   nearest  node.   If  torsion  is  not   modelled   then  

375  

w

 

System  Modelling:  Data  and  Results,  Links    

the  x,y  coordinates  are  ignored  and  the  connection  point  is  at  the  centreline  of  the  Line.  If   torsion  is  modelled  then   the  x,y  coordinates  allow  you  to  offset  the  connection  from  the  centreline.   For  Fixed  connections  the  object-­‐relative  coordinates  given  are  the  global  coordinates  of  the  point.   For  connecting  to  an  Anchor,  the  object-­‐relative  x,y  coordinates  given  are  the  global  X,Y  coordinates  of  the  anchor   point,   and   the   z-­‐coordinate   is   the   distance   of   the   anchor  above   (positive)   or   below   (negative)   the   seabed   at   that   X,Y   position.   For   connecting   to  other  objects,   the   coordinates   of   the   connection   point   are   given   relative   to  the   object   local  frame   of  reference.   Release  at  Start  of  Stage  

The  link  can  be  released  at  the  start  of  a  given  stage  of  the  simulation,  by  setting  this  number  to  the  stage  number   required.  Once  released  a  link  no  longer  applies  any  forces  to  the  objects  it  connects.  If  no  release  is  required,  then   set  this  item  to  '~'.   Unstretched  Length  

Is  the  unstretched  length  of  the  Tether  or  Spring.   Linear  

Both   the   spring   and   damper   in   a  Spring/Damper  can   have   either   simple   linear   force   characteristics   or   else   a   user-­‐ specified  piecewise-­‐linear  force  table.   Stiffness  

For  a  tether  the  tension  t  depends  on  its  strain  and  stiffness  as  follows:   t  =  k.(L-­‐L0)/L0   where   k  is  the  specified  Stiffness,   L  is  the  current  stretched  length  between  the  two  ends,   L0  is  the  specified  Unstretched  Length.   Tethers  remain  slack  and  exert  no  force  if  L  is  less  than  L 0.   For  a  linear  spring  in  a  Spring/Damper  the  tension  (positive)  or  compression  (negative)  is  given  by:   t  =  k.(L-­‐L0)   where   k  is  the  specified  Stiffness,   L  is  the  current  stretched  length  between  the  two  ends,   L0  is  the  specified  Unstretched  Length.   The  linear  spring  does  not  go  slack  if  L  is  less  than  L0,  but  instead  goes  into  compression.   Warning:  

Please  note  that  this  is  not  the  same  formula  as  for  tethers.  

Damping  

A  linear  damper  in  a  Spring/Damper  exerts  an  extra  tension  of   t  =  c.(rate  of  increase  of  L)   where   c  is  the  specified  Damping,   L  is  the  current  stretched  length  between  the  two  ends.   Non-­‐linear  force  tables  

For   a   non-­‐linear   spring   (or   damper)   the   force   characteristic   is   specified   as   a   table   of   tension   against   length   (or   velocity).   The   table   must   be   arranged   in   increasing   order   of   length   (velocity)   and   a   negative   tension   indicates   compression.   For   a   passive   damper   the   tensions   specified   should   therefore   normally   have   the   same   sign   as   the  

376  

w

 

System  Modelling:  Data  and  Results,  Shapes  

 

velocities,   since   otherwise   the   damper   will   apply   negative   damping.   For   lengths   (velocities)   between,   or   outside,   those  specified  in  the  table  the  program  will  use  linear  interpolation,  or  extrapolation,  to  calculate  the  tension.  

6.12.2

Results  

For  details  on  how  to  select  results  variables  see   Selecting  Variables.   For  links  the  following  variables  are  available:   Tension  

The  total  tension  in  the  link.   Length  

The  current  stretched  length  of  the  link.   Velocity  

The  rate  of  increase  of  the  stretched  length.   Azimuth  and  Declination  

The  azimuth  and  declination  angles,  relative  to  global  axes,  of  the  End  A  to  End  B  direction  of  the  link.   End  A  X,  End  A  Y,  End  A  Z,  End  B  X,  End  B  Y  and  End  B  Z  

The  global  coordinates  of  the  link.  

6.13

SHAPES  

  Figure:  

Examples  of  Block,  Cylinder  and  Curved  plate  shapes  

Shapes  are  simple  3  dimensional  geometric  objects  that  can  be  configured  in  a  variety  of  ways:   1.

Elastic  solids  are  used  to  model  physical  obstacles,  

2.

Trapped  water  are  used  to  model  moonpools  or  other  areas  where  fluid  motion  is  suppressed.  

3.

Drawing  shapes  have  no  physical  effect  on  the  model  and  are  just  intended  for  drawing  purposes.  

You   may   choose   between   a   number   of   different   basic   geometric   shapes   and   several   shapes   can   then   be   placed   together  to  defined  more  complex  shapes.  The  basic  shapes  available  are   planes,  blocks,  cylinders  and  curved  plates.  

377  

w

 

System  Modelling:  Data  and  Results,  Shapes    

Elastic  solids   An   elastic   solid   represents   a   physical   barrier   to   the   motion   of   lines   and   buoys.   It   is  made   of  a   material   of  a   specified   stiffness  and  resists  penetration  by  applying  a  reaction  force  normal  to  the  nearest  surface  of  the  elastic  solid  and   proportional  to  the  depth  of  penetration  of  the  object  into  the  elastic  solid.   Note:  

Elastic  solids  do  not  resist  penetration  by  Vessels,  Links,  Winches  or  other  Shapes.  

Each   elastic   solid   has   an   associated   stiffness,   which   determines   the   rate   at   which   the   force   applied   to   an   object   increases  with  the  area  of  contact  and  depth  of  penetration  into  the  elastic  solid.  The  stiffness  is  the  force  per  unit   area  of  contact  per  unit  depth  of  penetration.   Contact   with   elastic   solids   can   model   friction.   Friction   coefficients   are   specified   on   the   Solid   Friction   Coefficients   data  form.  Note  that  friction  for  contact  with  elastic  solids  is  only  included  during  dynamics.   Where  an  object  interacts  with  more  than  one  elastic  solid  simultaneously,  the  force  acting  on  it  is  the  sum  of  the   individual  forces  from  each  elastic  solid.   Elastic  solids  are  intended  only  for  modelling  the  overall  limitation  on  movement  that  a  physical  barrier  presents;   they  are  not  intended  to  model  an  object's  interaction  with  the  barrier  in  detail.  For  example  the  calculation  of  the   contact   area   and   penetration   depth   are   very   simplistic   and   do   not   allow   for   the   detailed   geometric   shape   of   the   object.   The   value   given   for   Stiffness   is   therefore   not   normally   important,   providing   it   is   high   enough   to   keep   penetration   small.   On   the   other   hand,   although   the   actual   stiffness   of   real   barriers   is   usually  very  high,   the   Stiffness   should   not   be   set   too   high   since   this   can   introduce   very   short   natural   periods   which   in   turn   require   very   short   simulation  time  steps.   Lines  only  interact  with  elastic  solids  by  their  nodes  coming  into  contact,  so  elastic  solids  that  are  smaller  than  the   segment   length   can   "slip"   between   adjacent   nodes.   The   segment   length   in   a   line   should   be   therefore   be   small   compared  with  the  dimensions  of  any  elastic  solid  with  which  the  line  may  make  contact.   Trapped  water   Trapped  water  can  be  used  to  model  hydrodynamic  shielding  Ȃ  i.e.  areas  such  as  moonpools,  the  inside  of  spars  or   behind  breakwaters,  where  wave  and  current  effects  are  suppressed.   Inside   a   trapped   water   shape   the   fluid   motion   is   calculated   as   if   the   fluid   was   moving   with   the   shape.   So   if   the   trapped  water  shape  is  fixed  then  no  fluid  motion  occurs  in  the  shape  Ȃ  this  could  be  used  to  model  a  breakwater.   But  if  the  shape  is  connected  to  a  moving  vessel,  for  example,  then  the  trapped  water  is  assumed  to  move  with  the   vessel  Ȃ  this  could  be  used  to  model  a  moonpool.   Note:  

Objects   ignore   any   trapped   water   shapes   which   are   connected   to   that   particular   object.   If   this   wasn't   done  then  if  you  connected  a  trapped  water  shape  to  a  buoy  and  part  of  the  buoy  was  in  the   trapped  water  shape  then  a  feedback  would  occur  (the  buoy  motion  determines  the  motion  of  the   shape,  which  in  turn  would  affect  the   fluid  forces  on  the   buoy  and  hence   its  motion).  Such  feedback   is  undesirable  so  the  buoy  ignores  any  trapped  water  shapes  that  are  connected  to  it.  

Drawing  shapes   Drawing   shapes   have   no   physical   effect   on   the   model.   They   can   be   used   to   draw   objects   of   interest   and   do   not   interact  with  other  objects.  

6.13.1

Data  

Name  

Used  to  refer  to  the  shape.   Type  

Either  Elastic  Solid  or  Trapped  Water.   Shape  

Can  be  one  of  Block,  Cylinder,  Curved  Plate  or  Plane.   Connection  

Can  be  Fixed,  Anchored  or  connected  to  another  object  ( Vessels,  3D  Buoys  or  6D  Buoys).  

378  

w

 

System  Modelling:  Data  and  Results,  Shapes  

 

Position  

Each   shape   has   position   data.   For   blocks   it   is   named   Origin,   for   cylinders   and   curved   plates   it   is   named   End   Position  and  for  planes  it  is  named  Point  on  Plane.   This  point  is  taken  as  the  origin  of  the  shape's  local  x,y,z  axes.   For  Fixed  connections  this  is  the  global  position  of  the  point.   For   Anchored   connections   the   object-­‐relative   x,y   coordinates   given   are   the   global   X,Y   coordinates   of   the   anchor   point,   and   the   z-­‐coordinate   is   the   distance   of   the   anchor  above   (positive)   or   below   (negative)   t he   seabed   at   that   X,Y   position.   For  connections  to  other  objects,  the  coordinates  of  the  connection  point  are  given   relative  to  the  object  local  frame   of  reference.   Pens  and  Number  of  Lines  

Each  surface  of  the  solid  is  drawn  as  a  wire  frame  using  one  the  specified  pens.  To  aid  visualisation,  the  Outside  pen   is  used  if  the  surface  is  being  viewed  from  the  outside  of  the  solid,  and  the   Inside  pen  is  used  if  it  is  being  viewed   from  the  inside.   The  Number  of  Lines  determines  how  many  lines  are  used  in  the  wire  frames  Ȃ  a  larger  value  gives  a  more  realistic   picture,  but  takes  a  little  longer  to  draw.   Data  for  Elastic  Solids   Normal  Stiffness  

This  is  the  reaction  force  that  the  solid  applies  per  unit  depth  of  penetration  per  unit  area  of  contact.  Stiffness  may   be  set  to  zero,  giving  a  solid  that  is  drawn  but  which  has  no  effect  on  the  other  objects  in  the  system.   Shear  Stiffness  

The  Shear  Stiffness  is  used  by  the  friction  calculation.  A  value  of  '~'  results  in  the  Normal  Stiffness  being  used.   Damping  

The  percentage   of   critical   damping   for   the  elastic   solid.  Damping   is  always   zero   when   using   the  implicit   integration   scheme.   See  Shape  Theory  for  technical  details.  

6.13.2

Blocks  

z

z-size

y Block Position B x-size

x y-size  

A  Block  shape  is  a  rectangular  cuboid,  defined  by  giving:  

379  

w

 

System  Modelling:  Data  and  Results,  Shapes    

Size  

This  defines  the  block's  dimensions  in  its  local  x,  y  and  z  directions.   With  respect  to  its  local  axes,  the  block  occupies   the  volume  x=0  to  Size(x),  y=0  to  Size(y),  z=0  to  Size(z).   Orientation  

This  is  defined  by  giving  three  rotation  angles,  Rotation  1,  2  and  3,  that  define  its  orientation  relative  to  the  object  to   which  the  block  is  attached,  or  else  relative  to  global  axes  if  it  is  not  attached  to  another  object.  For  example,  if  the   block   is   attached   to  an   object   with   local   axes   L xyz,   then   the   3   rotations   define   the   orientation   of   the   block   axes   Bxyz   as   follows.   First   align   the   block   with  the   local   axes   of   the   object   to   which  it   is  attached,   so   that   Bxyz   are   in   the   same   directions  as  Lxyz.  Then  apply  Rotation  1  about  Bx  (=Lx),  followed  by  Rotation  2  about  the   new  By  direction,   and   finally  Rotation  2  about  the  new  (and  final)  Bz  direction.  

6.13.3

Cylinders  

r = Inner Radius R = Outer Radius

R End 2 Position r End 1 Position   A  cylinder  shape  is  a  thick  walled  hollow  pipe  defined  by  giving:   x

Inner  and  Outer  Diameter.  

x

Length.  

x

Azimuth  and  Declination  of  the  axis.  

The  azimuth  and  declination  define  the  direction  of  the  axis  relative  to  the  local  axes  of  the  object  to  which  the  end  is   connected.  For  objects  that  rotate,  such  as  vessels  and  6D  buoys,  the  axis  direction   therefore  rotates  with  the  object.   For  Fixed  or  Anchored  ends  it  is  defined  relative  to  global  axes.   Cylinders  are  drawn  using  circles  to  represent  the   end  faces  and  a  number  of  rectangular  facets  to  represent  around   the  curved  surfaces.  The  number  of  facets  used  is  the  Number  of  Lines  specified.  Two  gives  a  very  simple  wire  frame   profile  of  the  cylinder,  whilst  a  very  large  number  gives  a  pseudo-­‐opaque  cylinder  at  the  expense  of  drawing  speed.   If  the  Inner  Diameter  is  zero  then  a  solid  disc  is  formed.   If  the  cylinder  is  an  elastic  solid  then  reaction  forces  are  applied:   x

Radially  inwards  if  an  object  comes  into  contact  with  the  inner  curved  surface.  

x

Radially  outwards  if  an  object  comes  into  contact  with  the  outer  curved  surface.  

x

Normally  outwards  if  an  object  comes  into  contact  with  one  of  the  end  faces.  

380  

w

 

System  Modelling:  Data  and  Results,  Shapes  

 

6.13.4

Curved  Plates  

  Figure:  

Some  example  curved  plate  shapes  

Curved   plate   shapes   are   particularly   suited   to   modelling   bellmouths   although   they   are   not   restricted   to   this   application.   The   curved   plate   shape   is   similar   to   the   cylinder.   It   differs   in   that   the   radius   of   the   shape   can   vary   smoothly  between  the  ends.  Curved  plates  can  be  either  filled  in  or  hollow.   Shape  is  hollow  

If   ticked   then   the   shape   is   hollow   and   has   both   inner   and   outer   surfaces.   The   middle   picture   above   is   not   hollow   whereas  the  other  two  are.   Orientation  

This   is   defined   by   giving   three   rotation   angles,   Azimuth,   Declination   and   Gamma,   that   define   its   orientation   relative   to   the   object   to   which   the   shape   is   attached,   or   else   relative   to   global   axes  if   it   is   not   attached   to   another   object.   The   Azimuth   and   Declination   values   define   the   direction   of   the   principal   axis.   The   Gamma   value   specifies   rotation  of  the  shape  about  its  own  axis  and  so  is  only  relevant  when  the  Included  Angle  of  Revolution  is  not  equal  to   360°.   Included  Angle  of  Revolution  

The   curved   plate   is   a   solid  of  revolution.   A   value   of  360°   gives   a   complete   revolution   as   shown   in   the   first   2   pictures   above.   Other   values   can   be   used   to   model   partial   or   cut-­‐away   curved   plates   Ȃ   for   example   the   right-­‐most   picture   above  has  an  included  angle  of  90°.   Thickness  

If  the  shape  is  hollow  then  this  data  item  specifies  the  wall  thickness.  This  thickness  specifies  the  thickness  normal   to  the  shape's  axis  or  centreline.   Note:  

You  may  need  to  specify  an   artificially  large  value  for  thickness  in  order  to  avoid  objects  passing   through  the  shape's  wall  during  the  static  calculation.  

Profile  

A   table   specifying   the   variation   of   diameter   with   distance   along   the   shape's   axis   or   centreline.   The   radius   is   the   radial  distance  (i.e.  in  direction  normal  to  the  shape  axis)  from  the  axis  to  the  surface.  If  the  shape  is  hollow  then  the   profile  defines  the  radius  to  the  inner  surface.  If  the  shape  is  not  hollow  then  the  profile  defines  the  radius  to  the   outer  surface.  Cubic  Bessel  interpolation  is  used  to  generate  a  smooth  profile.  

381  

w

 

System  Modelling:  Data  and  Results,  Shapes    

6.13.5

Planes   Direction of Maximum Slope Slope

Point on Plane   A  plane  shape  is  an  infinite  plane  surface  Ȃ  one  side  of  the  plane  is  outside  and  the  other  is  inside.  The  position  of   the  plane  is  defined  by  specifying  a  Point  on  Plane  through  which  it  passes.   The   angle   of   the   plane   is  specified   by   giving  its   (maximum)   Slope   Angle   and   Slope   Direction,   relative   to   the   object   to  which  it  is  connected,  as  follows.   x

For   a   fixed   or   anchored   shape,   the   Slope   Angle   is   specified   by   giving   the   angle   of   elevation   of   the   line   of   maximum  slope,  relative  to  the  global  XY  plane  (i.e.  relative  to  the  horizontal).  A  Slope  Angle  of  90°  is  therefore  a   vertical  plane.  The  Slope  Direction  is  specified  as  the   direction  of  the  line  of  maximum  upwards  slope,  relative  to   global   axes.   For   example   a   plane   having  a   Slope   Angle   of   30°   and   a   Slope   Direction   of   90°   slopes   upwards   in   the   positive  Y  direction  at  30°  to  the  horizontal.  

x

For  a  shape  connected  to  another  object,  the  Slope  Angle  and  Slope   Direction  are   relative  to  the   object's  local  xy   plane.   For   example   with   a   Slope   Angle   of   30°   and   a   Slope   Direction   of   90°,   the   plane   slopes   upwards   in   the   positive  y  direction  at  30°  to  the  object's  local  xy  plane.  

A  plane  with  zero  slope  angle  is  therefore  parallel  to  the  xy  plane  of  the  object  to  which  it  is  connected,  or  parallel  to   the  global  XY  plane  (i.e.  horizontal)  in  the  case  of  a  Fixed  or  Anchored  plane.   The   'inside'   of   a   plane   is  on   the   negative   z  side   (i.e.   below   for   a   Fixed   or   Anchored   plane)   if   the   Slope   Angle   is  in   the   range  -­‐90°  to  +90°,  and  on  the  positive  z  side  (i.e.  above  for  a  Fixed  or  Anchored  plane)  otherwise.   Planes  are  drawn  as  a  rectangular  grid,  with  the  specified   Number  of  Lines,  using  a  spacing  determined  by  the  view   size.   Planes  extend   to  infinity  in   all   directions,  but  only  a   part  of   the   infinite   plane   local   to   the   view   centre  is  shown   on  the  3D  view.  

6.13.6

Drawing  

Wire  frame  drawing   Representation   of  shapes  in  the  wire  frame  drawing   mode   can  be   confusing.   OrcaFlex   does   not   provide   hidden-­‐line   removal  so  shape  objects  are  displayed  by  simple  wire-­‐frame  drawings.  You  may  exercise  control  over  the  display   by  selecting  the  number  of  lines  drawn  for  each  object,  and  the  sequence  in  which  they  are  drawn.  For  pen  details,   see  How  Objects  Are  Drawn.   Where  it  is   necessary  to  keep  the  display  simple  you  should  set  Number  of  Lines  to  2  for   blocks  and  cylinders.  If  the   number   of   lines   is   set   large  for  blocks   or   cylinders   they  appear   as   solid   objects,   although   they  may  take   a   long   time   to  draw.   For   planes   you   can   control   how   they   are   drawn   with   the   Grid   Density   data   item.   This   is   specified   in   terms   of   the   length  of   the  scale  bar  on  the  3D  view.  A  density  of  d   means   that  there  are   d  lines  per  scale   bar  length,  so  higher   density  values  give  a  finer  grid  (but  takes  longer  to  draw).   Please   note   also   that   the   Number   of   Lines   only   affects   the   drawing,   and   not   the   calculations   (which   are   correctly   performed  with  curved  geometry).  Planes  and  Blocks  are  drawn  first,  and  then  Cylinders,  but  otherwise  the  solids  in   the   model   are   drawn   in   the   sequence   that   they  were   created.   You   can   sometimes   take   advantage   of  this,   by  defining   background   shapes   before   foreground   ones,   to   obtain   a   pseudo-­‐hidden   line   effect.   You   are   encouraged   to   experiment,  but  simplicity  is  best.   Hint:  

Although  the  program  provides  depth  clues  to  the  eye  by  drawing  rear  faces  in  a  different  colour,   the  eye  can  sometimes  be  fooled  by  the  picture  Ȃ  try  rotating  the  view  back  and  forth  a  few  times.  

382  

w

 

System  Modelling:  Data  and  Results,  All  Objects  Data  Form  

 

Shaded  Drawing   By  default,  for  shaded  3D  Views,  shapes  are  drawn  as  solid  objects  using  the  specified  geometry.   Alternatively  the   object   can  be  represented  by  an  imported   3D   model  by   specifying  the  Shaded  Drawing   File.  This   must  be  a   Direct   X  format  file,  usually  with   the  .x   file   extension.  If  you  use  a  relative   path  then  the  path   will  be  taken   as  relative  to  the  folder  containing  the  OrcaFlex  file.   The   Browse   button   allows   you   to   specify   the   Shaded   Drawing   File   quickly   and   also   provides   quick   access   to   the   Orcina  sample  drawings  via  the  Navigate  to  Orcina  sample  drawings  button.   The  Use   Culling  option   is   normally   selected   since   it   can   provide   a   useful   performance   benefit.   However,   in   order   to   work   it  requires  that  the  triangles  defined  in  the  .x   file  have  their  outward  facing  directions  defined  correctly.  In   the   unusual   situation   where   the   outward   facing   directions   are   not   defined   correctly   then   the   .x   file   will   not   display   correctly.  If  this  happens  then  typically  sections  of  the  model  will  be  missing  when  drawn  by  OrcaFlex.  Disabling   the   Use  Culling  option  resolves  this  problem.   Draw  Size  is  provided  to  allow  you  to  scale  the  drawing.  All  directions  are  scaled  equally  to  arrange  that  the  longest   side  in  the  drawing  is  drawn  to  the  specified  Draw  Size.  This  longest  side  is  calculated  by  first  fitting  the  smallest   possible  cuboid  around  the  vertices  of  the  shaded  drawing  (these  are  defined  in  the  .x  file).  This  cuboid  is  aligned   with  the  shaded  drawing's  local  axes.  Then  the  length  of  the  longest  side  of  this  cuboid  is  found.   Specify  a  value  of  '~'  to  display  the  drawing  using  the  absolute  coordinates  as  specified  in  the  .x  file.   Note:  

If  you  use  a   value  of  '~'  for   Draw  Size  then  OrcaFlex   uses  the  coordinates  in  the  .x   file  directly.  If   these   coordinates   use   a   different   length   units   system   from   your   OrcaFlex   model   then   you   should   specify   the   units   used   in   the   .x   file   by   including   an   auxiliary   file   called   AdditionalInformation.txt.   Examples  of  this  can  be  found  in  the  sample  shaded  drawings  provided  by  Orcina.  

Shaded   Drawing   Origin   is   provided   because   the   shaded   drawing   and   the   shape   may   have   different   origins.   The   Shaded   Drawing   Origin   defines   the   origin   of   the   shaded   drawing   with   respect   to   the   shape's   local   axis   system.   Similarly   Shaded   Drawing   Orientation   allows   you   to   reorient   the   shaded   drawing   to   match   the   shape's   axis   system.   Shaded  Drawing  Plane  Translucency  (only  available  for  planes)  

Controls  how  translucent  the  plane  appears  in  the   Shaded  Graphics  mode.  A  value  of  0%  gives  a  solid  surface  and  all   objects   behind   the   surface   will   not  be   visible.   A   value   of  100%   specifies   transparency   and   leads   to   a   completely   see-­‐ through  surface.   Note:  

6.13.7

This  data  item  is  not  used  if  an  imported  3D  model  is  used  to  draw  the  shape.  

Results  

For  details  on  how  to  select  results  variables  see  Selecting  Variables.   Contact  Force,   Contact  GX-­‐Force,  Contact  GY-­‐Force,  Contact  GZ-­‐Force,   Contact  Lx-­‐Force,  Contact  Ly-­‐Force  and  Contact  Lz-­‐Force  

The   magnitude   and   components   of   the   total   force   applied   by   an   elastic   solid   to   other   objects   in   the   model.   The   components  are  reported  relative  to  either  global  axes  (GX,  GY,  GZ)  or  local  axes  (Lx,  Ly,  Lz).   These  variables  are  only  available  for  elastic  solids.   X,  Y,  Z  

The  global  coordinates  of  the  shape  origin.   Velocity,  GX-­‐Velocity,  GY-­‐Velocity,  GZ-­‐Velocity,   Acceleration,  GX-­‐Acceleration,  GY-­‐Acceleration,  GZ-­‐Acceleration  

The  magnitude  and  components  (with  respect  to  global  axes)  of  the  velocity  and  acceleration  of  the  shape  origin.  

6.14

ALL  OBJECTS  DATA  FORM  

The  All  Objects  Data  Form  allows  you  to  view  or  edit  data  for  all  the  objects  in  the  model  on  a  single  form.  This  is   particularly   useful   for   simultaneous   viewing   or   editing   of   properties   of   multiple   objects.   The   form   can   be   opened   using  the  model  browser.  

383  

System  Modelling:  Data  and  Results,  All  Objects  Data  Form    

w

 

There  are  two  modes  of  operation:  Connections  or  Other  data.   Connections   The   Connections   mode   allows   you   to   edit   connection   specific   data   for   all   Lines,   Links   and   Winches   in   the   model.   You   can   choose   to   filter   which   object   types   appear   in   the   list   Ȃ   each   object   type   (Lines,   Links   or   Winches)   can   be   included  or  excluded  from  the  list.   Positions  and  Connections  pages  

The  Positions  and  Connections  pages  allow  you  to  view  or  edit  all  the   connection  data.  This  is  the  same  data  as  on   the  individual  data  forms  and  includes  the  following  data  items:   x

The   positions  (as  Cartesian   coordinates)   relative  to   the   frame   of  reference  of  the   object  to   which  the  connection   is  attached.  

x

For  Links  and  Winches  connected  to  a  line,  the  reference  line  end  (either  End  A  or  End  B)  for  the  z  coordinate.  

x

Height  above  seabed  for  anchored  line  ends.  

x

Connection  orientation  and  stiffnesses  for  line  ends.  

x

Release  stage  for  line  ends.  

Polar  Coordinates  page  

The   Polar   Coordinates   page   provides   a   way   of   viewing   or   setting   the   positions   of   the   connections   using   polar   coordinates,  relative  to  a  choice  of  frames  of  reference.  This  facility  is  useful  for  cases,  for  example  mooring  arrays,   where  a  series  of  connections  need  to  be  laid  out  around  a  circle.   Š‡’‘Žƒ”…‘‘”†‹ƒ–‡•ȋǡɅǡȌƒ”‡–Š‘•‡‘ˆ–Š‡…‘‡…–‹‘’‘•‹–‹‘”‡Žƒ–‹˜‡–‘–Š‡•‡Ž‡…–‡†’‘Žƒ” coordinates  frame   of   reference   (see   below).   The   Cartesian   coordinates   of   the   connection,   relative   to   the   same   reference   frame,   are   ȋ…‘•ȋɅȌǡ•‹ȋɅȌǡȌǤ   On  the  other  hand,  the  Object  Relative  Position  data  are  the  Cartesian  coordinates  of  the  connection  relative  to  the   frame  of  reference  of  the  object  to  which  it  is  connected.   OrcaFlex   keeps   the   two   sets   of   coordinates   synchronised,   so   if   you   change   one   then   the   other   is   automatically   updated   to   match.   If   you   change  any   other   data   then   the   Cartesian   Object   Relative   Position   coordinates   are   taken   to   be  the  master  data  and  so  left  unchanged,  and  the  polar  coordinates  are  updated  to  match.   You  have  a  quite  a  lot  of  flexibility  to  choose  what  reference  frame  you  want  for  the  polar  coordinates.  The  reference   frame   has   its   origin   at   your   chosen   Reference   Origin   and   has   its   axes   are   parallel   to   those   of   your   chosen   Reference  Axes.   For  the  reference  origin  you  can  choose  between:   x

The  global  origin.  

x

The   point   on   the   seabed   that   is   directly   below   the   global   origin.   This   is   only   available   if   a   connection   point   is   anchored.  

x

The  origin  of  the  frame  of  reference  of  any  connected  object.  

x

The  position  of  the  other  end  of  the  object  Ȃ  available  for  Lines  and  Links  only.  

And  for  the  reference  axes  directions  you  can  choose  between:   x

The  global  axes  directions.  

x

The  axes  directions  of  the  frame  of  reference  of  any  connected  object.  

Example  of  Using  Polar  Coordinates  

The   choices   of   reference   frame   for   the   polar   coordinates   may   seem   complex   at   first   sight,   but   they   allow   various   useful  coordinate  transformations  to  be  done  easily  and  accurately.  Here  is  an  example.   Consider   mooring   a   spar   with   an   array   of   4   lines,   each   of   which   has   End   A   connected   to   the   spar   and   End   B   anchored.  Suppose  you  want  to  place  the  A  ends  of  the  lines  so  that  they  are  evenly  spaced  circumferentially  around   the   spar,   all   at   radius   5m   from   the   spar   axis   and   all   3m   below   the   spar   origin.   To   do   this   easily,   first   sort   into   Connection  order  so  that  all  the  End  A  connections  are  grouped  together.  Then,  for  the  first  line,  set  the  reference   frame   origin   and   axes   to   be   the   spar   origin   and   spar   axes   and   set   its  polar   coordinates   to   be   R=5,   and   Z=-­‐3.   You  can  

384  

w

 

 

System  Modelling:  Data  and  Results,  All  Objects  Data  Form  

now   use   copy/paste   or   fill   down   to   set   all   the   other   A   ends   to   the   same   reference   origin,   axes   and   R   and   Z   …‘‘”†‹ƒ–‡•Ǥ ‹ƒŽŽ››‘—…ƒ•‡––Š‡Ʌ…‘‘”†‹ƒ–‡•ˆ‘”–Š‡‡†•–‘Ͳιǡ90°,  180°  and  270°.   Similarly,   suppose   you   want   the   End   B   connections   to   be   anchored   to   the   seabed,   with   the   anchors   again   evenly   spaced   circumferentially,   and   with   each   line   spanning   200m   horizontally.   The   easiest   reference   frame   for   this   is   with  the   refe”‡…‡‘”‹‰‹„‡‹‰†ƒ†–Š‡”‡ˆ‡”‡…‡ƒš‡•„‡‹‰–Š‡•’ƒ”ƒš‡•ǤŠ‡Ʌ…‘‘”†‹ƒ–‡••Š‘—Ž†ƒ‰ƒ‹„‡ set  to  0°,  90°,  180°  and  270°  and  the  R  coordinates  set  to  200m.  But  this  time,  to  set  the  vertical  positions  of  the  B   ends,   it  is   easier   (especially   if  the   seabed   is   sloping)   to   go   to   the   Connections   page   and   set  Connect   To  Object  to   be   Anchored  and  then  go  to  the  Positions  page  and  set  the  Object  Relative  Position  z  coordinate  to  zero.   Other  data   The  Other  data  mode  presents  tables  of  data  for   Vessels,  Lines,  6D  Buoys,  3D  Buoys,  Links,   Winches  and   Shapes.   The  tables  are  laid  out  with  each  row  containing  related  data  for  a  single  object.  The  notable  exception  is  the  Line   sections  page  in  which  each  row  contains  data  for  a  single  section  of  a  Line.   This  tabulation   allows   data   for   multiple   objects   to  be   set   in   an   efficient   and   coordinated  way.  The   copy/paste  or  fill   down  keyboard  shortcuts  are  particularly  useful  here.  Another  useful  technique  is  to  build  a  table  of  data  in  Excel   with   identical   layout   of   columns   and   rows.   This   allows   you   to   make   use   of   Excel's   formulae   and   data   handling   facilities  to  prepare  data,  and  then  to  paste  it  into  OrcaFlex  in  a  single  operation.  

385  

w

 

Modal  Analysis,  All  Objects  Data  Form  

 

7

MODAL  ANALYSIS  

The   modal   analysis   form   enables   you   to  calculate   and   view   the   undamped   natural   modes   of   the   whole   system,   or   of   a  single  line.   To  open  this  form,  see  the  Modal  Analysis  command   on  the  Results  menu.  Note  that  the  analysis  is  only   available  when  the  static  position  of  the  model  has  been  calculated.   For  full  details  of  the  calculation,  and  a  discussion  of  its  limitations,  see  the   Modal  Analysis  theory  section.   Performing  modal  analysis   To  perform  a  modal  analysis  you  need  to  specify  the  following:   x

What  you  want  to  analyse:  the  whole  system  or  a  single  line.  

x

Which   modes   you   want   to   calculate.   You   can   ask   for   All  modes   or   a   specified   range   of  modes.   For   large   systems   it  is  much  quicker  to  calculate  only  a  small  number  of  modes  Ȃ  see  Modal  Analysis  Theory  for  more  details.  

x

Whether  you  want  to  calculate  the  mode  shapes  or  just  the  natural  periods.  If  you  exclude  the  mode  shapes  then   the  analysis  only  calculates  the   natural  periods,   not  the  shapes   of  the  natural  modes.  If  you  include  the  mode   shapes  then  the  analysis  takes  longer.  

When   you   have   made   your   selections   click   the   Calculate   button.   The   modal   analysis   will   then   calculate   the   undamped   natural   periods   and,   if   requested,   the   mode   shapes.   Each   mode   is   normalised   to   have   largest   offset   magnitude  equal  to  1,  i.e.  the  offsets  vectors  are  scaled  so  that  largest  offset  vector  is  a  unit  vector.  The  modes  are   numbered  in  order  of  increasing  frequency.   Mode  Table   The  Table  page  then  displays  a  spreadsheet  giving  the  results  in  numerical  form.  If  you  do  not  calculate  the  mode   shape  then  the  table   reports  only  the  periods  of  the   requested   natural  modes.  If  you   calculate  the  mode  shapes   then   the   table   also   gives   the   shape   in   the   form   of   the   displacements   of   each   degree   of   freedom.   Mode   shapes   can   be   reported  with  respect  to  either  global  axes  directions  or  local  axes  directions.   If  mode  shapes  are  calculated  then  the  table  includes  the  following  information:   x

Offset   distribution   displays   a   measure   of   how   inline,   transverse   and   axial   the   mode   is.   For   details   see   Offset   Distribution  below.  

x

Mode   type   classifies   each   mode   according   to   the   offset   distribution.   Transverse   means   that   the   transverse   component   is   more   than   90%   of   the   total,   Mostly   transverse   means   that   it   is   between   50%   and   90%,   and   similarly  for  inline,  mostly  inline,  axial  and  mostly  axial.  Mixed  means  that  none  of  the  components  are  more   than  50%  of  the  total.   Note:  

The  Offset   Distribution   and  Mode   Type  information   is  only   available  for  single   line   analyses   where   there   is   relative   flow   normal   to   the   line.   So,   if   the   line   is   entirely   above   the   water,   or   there   is   no   current  defined  then  this  information  will  not  be  available.  

Mode  View   If   you   requested   the   mode   shapes   then   the  View  page   displays   a   3D   view   of   the   system   showing   one   selected   mode   shape  superimposed  on  the  static  position  of  the  system.  The  current  direction  is  also  shown  on  the  view,  and  you   can   control   the   view  angle,   zoom   etc.,   as   on   any  3D   view.   You   may   need   to   zoom   out   in   order   to   see   the  system,   and   you  may  need  to  adjust  the  view  angle  to  suit  the  mode  that  you  are  viewing.  For  example  an  out  of  plane  mode  for  a   catenary  is  best  viewed  by  looking  along  the  plane  of  the  catenary.   You   can   use   the   mode   drop-­‐down   list   to  control   which  mode   is   shown   on   the   view.   Note   that   when   that   drop-­‐down   list  has  the  focus  (click  it  to  give  it  the  focus)  then  you  can  use  the  arrow  keys  to  quickly  increment  or  decrement  the   mode  shape  number  that  is  displayed.   The  drawing  exaggeration  value  allows  you  to  vary  the  amplitude  of  the  drawn  mode  shape.  The   animate  mode   shape  and  draw  node  axes  options  allow  further  control  of  the  mode  shape  drawing.   If   the   mode   shape   is  being   animated   then   there   is  a   further   choice   to  make,   the  animation   period.   If   you   select   the   mode   period   option   then   the   animation   has   a   cycle   period   equal   to   the   mode   period.   However,   for   modes   with   either   very   long   or   very   short   periods,   this   option   can   make   visualisation   of   the   mode   shape   quite   difficult.   The   alternative  option,  fixed,  animates  the  mode  with  a  5s  cycle  period.  

387  

w

 

Modal  Analysis,  Modal  Analysis  Theory    

For   single   line   analyses,   the  offset   distribution  and  mode   type  for   the   selected   mode   is   also   shown,   provided   that   there  is  relative  flow  normal  to  the  line.   VIV  Page   The   VIV   facilities   related   to  modal   analysis  are   not   yet   available   for   lines   that   have  torsion   included.   These   facilities   are  also  currently  only  available  for  single  line  analyses.   If   you   requested   that  mode  shapes   be   calculated   then  the  View  page   displays   a   table   of   information   relating   to   VIV.   Each  row  of  the  table  refers  to  a  single  mode  and  contains  the  following  information:   x

The  mode  number.  

x

The  mode's  period  and  frequency.  

x

The  mode's  offset  distribution  and  mode  type.  

x

Export  to  SHEAR7  Mds  file  determines  whether  or  not   the  mode   will  be  included  in  the  exported  SHEAR7  Mds   file.  

The  filter  allows  you  to  restrict  the  table  to  show  only  certain  types  of  mode.  For  example,  you  may  wish  to   view   only  the  transverse  modes  when  considering  transverse  VIV.  The  table  reports  modes  in  order  of  increasing  mode   number.   Offset  Distribution  

For  a  given  mode,  let   V  be   the  mode  offset  vector  at  a  given  node,  let   V i,   Vt  and  Va   be  V's  components  in  the  local   inline,   transverse   and   axial   directions   and   let   L   be   the   length   of   line   represented   by   that   node.   Then   OrcaFlex   calculates  the  inline,  transverse  and  axial  components  of  the  overall  mode  shape  as  follows:   Mode  shape  inline  component  =  Mi  αȋσi2)½   Mode  shape  transverse  component  =  Mt  αȋσt2)½   Mode  shape  axial  component  =  Ma  αȋσa2)½   where  the  summations  are  over  all  nodes  in  the  line.   OrcaFlex  then  reports  these  3  mode  shape  components  as  percentages  of  their  sum,  i.e.  the  offset  distribution  values   reported  by  OrcaFlex  are  Mi/M,  Mt/M  and  Ma/M  where  M  =  Mi  +  Mt  +  Ma.  These  values  are  only  used  to  provide  some   measure  of  how  inline,  transverse  and  axial  the  mode  is.   The  root   sum   of   squares  formulation   used   above   is  the  multi-­‐dimensional   equivalent   of   the   standard   3D   formula   |V|   =  (Vi2  +  Vt2  +  Va2)½.  The  scaling  by  L  is  there  so  that  the  values  are  independent  of  the  level  of  discretisation.   Export  SHEAR7  Mds  File  

This  button  exports  a  SHEAR7  .Mds  file.  

7.1

MODAL  ANALYSIS  THEORY  

A   modal   analysis   calculates   the   undamped   natural   modes   of  a   system.   These   modes   are   given   in   decreasing   order   of  period  and  are  numbered  starting  from  1.   The   analysis   calculates   the   natural   modes   of   the   discretised   model,   not   those   of   the   real   continuous   system.   However  the  discretised  modes  are  close  to  the  continuous  ones  and  for  a  mode  number  the  accuracy  improves  as   more  and  more  elements  are  used  to  model  the  system.  For  any  given  level  of  discretisation  the  accuracy  is  better   for  the  lower  modes  and  progressively  worsens  as  you  go  to  higher  and  higher  modes.  The  highest  numbered  modes   are   unlikely   to   be   realistic   since   they   are   oscillations   whose   wavelengths   are   of   the   same   order   as   the   segment   length.   If  the  system  hangs  in  one  of  the  global  axis  planes  then  you  can  often  distinguish  whether  a  mode  is  in-­‐plane  or  out-­‐ of-­‐plane  by  looking  at  the  pattern  of  zeros  in  the  table  of  displacements.  For  example  if  the  system  hangs  in  the  XZ   plane   then   the   out-­‐of-­‐plane  modes   have   non-­‐zero   Y-­‐displacements   but   zero   (or   very  small)   X-­‐  and   Z-­‐displacements,   and  the  in-­‐plane  modes  have  the  opposite  pattern  of  zeros.   Outline  Theory  

Modal   analysis   is   a   standard   technique   that   is   well-­‐documented   in   the   literature,   but   here   is   a   brief   outline.   First   consider  a  single  degree  of  freedom  system  consisting  of  a  mass  attached  to  a  linear  spring.  The  undamped  equation   of  motion  is:  

388  

w

 

Modal  Analysis,  Modal  Analysis  Theory  

 

Mx''(t)  =  -­‐Kx(t)   where   x(t)  is  the  offset  (at  time  t)  from  mean  position,  x''(t)  is  the  acceleration,  M  is  its  mass  and  K  is  the  stiffness  of   the  spring.  Since  this  analysis  neglects  any  damping  the  results  are  referred  to  as  the  undamped  modes.   Š‡ •‘Ž—–‹‘ ‘ˆ –Š‡ ‡“—ƒ–‹‘ ‹• ‘™ –‘ „‡ •‹’Ž‡ Šƒ”‘‹…ǡ ‹Ǥ‡Ǥ ‘ˆ –Š‡ ˆ‘” šȋ–Ȍ α ƒǤ•‹ȋɘ–Ȍǡ ™Š‡”‡ ƒ ƒ† ɘ ƒ”‡ unknowns  to  be  found  by  solving  the  equation.  Differentiating  x(t)  gives:   x''(t)  =  -­‐ɘ2ǤƒǤ•‹ȋɘ–Ȍ   so  when  we  substitute  into  the  equation  of  motion  we  obtain:   -­‐Ǥɘ2ǤƒǤ•‹ȋɘ–Ȍα-­‐ǤƒǤ•‹ȋɘ–ȌȋͳȌ   which  can  be  rearranged  to  give:   ɘαȋȀȌ½.   This  is  the  angular  frequency  of  the  oscillation  and  so  the  natural  period  T  is  given  by:   T  αʹɎȋȀȌ½   For  this  simple  harmonic  oscillator  there  is  just  a  single  undamped  natural  mode,  corresponding  to  the  single  degree   of  freedom.  For  a  continuous  riser  there  are  an  infinite   number  of  degrees  of  freedom,  and  hence  an  infinite  number   of   undamped   natural   modes,   but   computers   work   with   discretised   models   with   finite   numbers   of   degrees   of   freedom.   Consider  a  discretised  line  in  OrcaFlex  with  N  degrees  of  freedom.  In  this  situation  the  above  equations  still  apply,   but  they  now  have  to  be  interpreted  ƒ•ƒ–”‹šȀ˜‡…–‘”‡“—ƒ–‹‘•™Š‡”‡ɘƒ†”‡ƒ‹•…ƒŽƒ”•ǡƒǡšƒ†š̵̵„‡…‘‡ vectors  with  N  elements,  and  M  and  K  become  N×N  matrices.   Equation  (1)  is  an  eigen-­‐problem  with  N  solutions,  the  ith  •‘Ž—–‹‘„‡‹‰ɘi  and  aiǡ•ƒ›ǡ™Š‡”‡ɘi  is  a  scalar  and  ai  is  a   vector  with  N  elements.  This  ith  solution  is  called  the  ith  natural  mode.  It  is  an  oscillation  of  the  line  in  which  all  the   †‡‰”‡‡• ‘ˆ ˆ”‡‡†‘ ‘•…‹ŽŽƒ–‡ ƒ– –Š‡ •ƒ‡ ƒ‰—Žƒ” ˆ”‡“—‡…› ɘi.   But   different   degrees   of   freedom   have   different   amplitudes,  given  by  the  components  of  ai.  This  amplitude  variation  is  called  the  mode's  shape.   Eigen-­‐solvers  

Two  eigen-­‐solvers  are  used  to  perform  modal  analysis.  The  choice  of  which  to  use  is  made  based  on  the   number  of   modes  extracted,  n,  and  the  number  of  degrees  of  freedom,  N.   ˆ  ζ Ȁ͵ ƒ†  ζ ͳͲͲͲ –Š‡ ƒ ‹–‡”ƒ–‹˜‡ ƒ…œ‘• ƒŽ‰‘”‹–Š ™‹ŽŽ „‡ —•‡†Ǥ –Š‡”™‹•‡ ƒ †‹”‡…– ‡–Š‘† „ƒ•‡† ‘ tridiagonal  MATRIX  diagonalisation  is  used.  For  large  problems  the  iterative  Lanczos  algorithm  is  much  faster  and   requires  much  less  memory  and  so  should  be  used  if  at  all  possible.   One   final   subtlety   concerns   the   precise   definition   of   n   in   the   above   inequalities.   The   Lanczos   algorithm   works   by   finding  the  largest  (or  smallest)  eigenvalue  first,  then  the  next  largest  (or  smallest)  and  so  on.  Consequently  if  you   ask  for  modes  5  to  10  then  the  solver  has  to  find  modes  1  to  4  first  and  so  the  number  of  modes  extracted,  n,  is  10.   Seabed  friction  

The  theory  outlined  above   requires  that  the  mass  and  stiffness  matrices  are  symmetric  which  is   not  always  the  case   in   an   OrcaFlex   model.   The   most   important   example   of  this   is   the  friction   model.   Friction   is   a   non-­‐conservative   effect   and  non-­‐conservatism  equates  to  non-­‐symmetric  terms  in  the  stiffness  matrix.  Clearly  this  presents  a  problem.   The   non-­‐conservatism   of   the   standard   OrcaFlex   friction   model   arises   when   a   node   is   slipping,   that   is   when   the   deflection   from   its   friction   target   position   exceeds   Dcrit.   When   performing   modal   analysis   OrcaFlex   assumes   that   nodes   on   the   seabed   are   restrained   by   a   linear   stiffness   effect   determined   by   the  seabed's   shear   stiffness,   Ks   and   the   node's  contact  area,  A.  This   stiffness  term  corresponds  to  the  stiffness  of  a  linear  spring  acting  in  the  plane  of  the   seabed,  connecting  the  node  and  its  target  position,  and  with  a  stiffness  of  K sA.   This  has  the  effect  of  restraining  movement  of  the  nodes  on  the  seabed,  in  the  plane  of  the  seabed  which  is  desirable   for   a   modal   analysis   of   a   system   with   seabed   contact.   This   modification   to   the   seabed   friction   model   results   in   a   symmetric,  conservative  system  and  hence  enables  modal  analysis  t o  be  performed  successfully.   Stiffness  terms  due  to  fluid  loading  

As   mentioned   above,   the   modal   analysis   is   an   undamped   analysis   which   means   that   damping   terms   (i.e.   those   dependent  on  x')  are  neglected.  So,  for  example,  the  effects  of  drag  loading  are  neglected  in  the  modal  analysis.   However,   fluid  loads  do  contribute  stiffness   terms   because  perturbations  of   position  and  orientation  can   result  in   changes  of  direction  of  the  fluid  load  vector.  These  are  stiffness  terms,  as  opposed  to  damping  terms,  because  they  

389  

w

 

Modal  Analysis,  Modal  Analysis  Theory    

arise   from   displacements   even   though   the   magnitude   of   the   loads  themselves   are   dependent   on   velocities.   OrcaFlex   does   include   these   stiffness   terms   to   improve   convergence   for   whole   system   statics   and   implicit   dynamics   calculations.  However,  these  fluid  load  terms  are  neglected  in  the  modal  analysis.   Non-­‐linear  bend  stiffness  

Modal  analysis  inherently  assumes  linearity  of  the  system  under  consideration.  For  non-­‐linear,  elastic  bend  stiffness   the  local  tangent  stiffness  is  used.  For  small  oscillations  about  the  static  configuration,  such  a  system  is  linear  and   modal  analysis  may  be  adequately  accurate.   For   hysteretic   bend   stiffness   the   situation   is   more   complex.   The   local   tangent   stiffness   is,   in   general,   ill-­‐defined   because   it   can   be   many-­‐values   depending   on   whether   the   perturbation   increases   or   decreases   curvature.   In   this   situation  OrcaFlex  uses  an  average  of  the  possible  stiffness  values.   Yet   another   case   to   consider   is   externally   calculated   bend   stiffness.   The   external   function   interface   provides   no   mechanism   for   specifying   the   local   tangent   bend   stiffness.   Hence   OrcaFlex   uses   the   nominal   bend   stiffness   as   provided  by  the  external  function.   Clearly  the  modal  analysis  will  be  less  accurate  for  hysteretic  and  externally  calculated  bend  stiffness  than  it  is  for   the   other   bend   stiffness   options.   However,   it   is   often   the   case   that   tension,   as   opposed   to   bend   stiffness,   is   the   dominant  contribution  to  lateral  stiffness  and  so  the  limitations  described  above  are  often  not  significant.   Vessel  added  mass  

When   performing   a   whole   system   modal   analysis   with   vessel   degrees   of   freedom   included,   the   added   mass   of   a   vessel   is   usually  a   significant   factor  in  the  analysis.  However,   if  the  vessel  added   mass  and  damping  data  is  specified   as  frequency  dependent  there  is  no  easy  way  for  the  modal  analysis  to  account  for  this  dependency  on  frequency.   Therefore,  for  modal  analyses,  the  program  neglects  added  mass  when  the  data  is  frequency  dependent.   In  order  to  perform  a  more  accurate  modal  analysis  you  can  use  the  constant  option  for  added  mass  specification.  In   this  situation  you  provide  a  single  added  mass  matrix   which  is  included   in  the  system   wide  mass  matrix.  Because   you   can   only   specify   a   single   added   mass   matrix   you   must   first   assess   what   modes   are   of   interest   and   choose   appropriate  added  mass  values.  If  the  added  mass  varies  significantly  with  mode  period,  over  the  range  of  periods   under  consideration,  then  you  may  need  to  perform  multiple  analyses  with  different  added  mass  matrices.  

390  

w

 

Fatigue  Analysis,  Modal  Analysis  Theory  

 

8

FATIGUE  ANALYSIS  

The   OrcaFlex   fatigue   analysis   is   a   post-­‐processor   which   calculates   fatigue   damage   using   a   variety   of   methods.   Damage   is   then   collated   and   summed   for   specified   load   cases   and   then   presented,   either   as   plots   or   in   tabular   fashion.   Damage  can  be  calculated  in  a  variety  of  ways:   x

The  S-­‐N  curve  approaches  recover  stress  using  either  homogeneous  pipe  stress  (suitable  for  metal  risers)  or   stress   factors   (suitable   for   umbilicals   and   flexibles).   Damage   is   then   calculated   based   on   the   specified   S-­‐N   curve.  

x

Mooring  line  fatigue  calculates  damage  from  effective  tension  ranges  using  T -­‐N  curves.  

Likewise,  damage  is  summed  in  a  variety  of  ways:   x

Deterministic  regular  wave  fatigue  analysis.  

x

Deterministic  irregular  wave  fatigue  analysis  using  the  rainflow  cycle  counting  method.  

x

Stochastic  irregular  wave  fatigue  analysis  using  spectral  methods.  

The  SHEAR7  option  is  rather  different  because  damage  is  calculated  external  to  OrcaFlex  by  SHEAR7.  This  option   allows  you  to  collate  fatigue  damage  from  a  number  of  SHEAR7  load  cases.   The   fatigue   analysis  tool   is  accessed  by  selecting   the   Fatigue   Analysis  command   from   the   OrcaFlex   Results  menu.  It   is  essentially  a  self-­‐contained  sub-­‐program  within  OrcaFlex,  with  its  own  menus,  data  and  results.   The  steps  involved  in  performing  a  fatigue  analysis  are:   1.

Use  the  normal  OrcaFlex  facilities  to  set  up  and  run  simulations  that  model  the  various   load  cases  that  the  line   will  experience.  Alternatively,  for  a  SHEAR7  analysis,  create  a  set  of  SHEAR7  .plt  output  files  to  represent  your   VIV  load  cases.  

2.

Open  the  fatigue  analysis  tool  and  set  up  the  fatigue  analysis  data.  This  fatigue  analysis  data  is  held  separately   from  the  other  OrcaFlex  data  and  can  be  saved  in  a  separate  file  with  the  file  extension  .ftg.  

3.

Check  the  data  for  errors.  

4.

Calculate  and  collate  the  damage.   Notes:  

The  calculation  stage  of  a  fatigue  analysis  can  take  a  long  time  ,  especially  a  rainflow  analysis  with   a   lot   of   load   cases.   To   help   with   this   there   is   an   Estimate   Calculation   Time   facility   and   fatigue   analyses  can   be  run   in   batch  mode.  The  calculation  makes  use  of  all  available  processor  cores  to   process  load  cases  concurrently.  

 

The  fatigue  calculation  performance  is  often  limited  by  disk  access   Ȃ  consequently  it  is  important   that  the  disk  access  speed  is  as  fast  as  possible.  Usually  this  means  that  the  simulation  files  should   be  stored  on  a  local  disk  of  the  machine  performing  the  fatigue  calculation.  

Load  Cases   Before  the  fatigue  analysis  can  be   performed  you   must  first  prepare  a  set   of  OrcaFlex  simulation  files  that  model  the   same  system  but  under  the  various  load  conditions  that  the  system  will  experience  in  its  lifetime.   The   approach   is   to   divide   the   range   of   sea   states   that   the   system   will   experience   into   a   number   of   wave   classes;   typically  this  is  done  with  a  wave  scatter  table.   For  both  regular  and  rainflow  analysis  you  typically  represent  each  wave  class  with  a  distinct  OrcaFlex  simulation   file.   For   regular   analysis   the   simulation   should   use   a   regular   wave   representative   of   the   wave   class   and   for   rainflow   analysis  the  simulation  should  use  an  irregular  wave  representative  of  the  wave  class.   For   spectral   analysis   multiple   wave   classes   with   similar   Hs   values   may   be   represented   by   a   single   response   calculation  simulation.   The   reason   this  is  possible   is  that   the   spectral   response   analysis  provides   information   about   how  the  system  responds  to  a  range  of  wave  frequencies.  Typically  you  will  have  a  range  of  simulations  which  cover   the  range  of  Hs  values  in  your  wave  scatter  table.   For  SHEAR7  fatigue  analysis  the  load  cases  are  specified  by  a  set  of  SHEAR7  .plt  output  files.  These  are  most  easily   generated   using   the   direct   SHEAR7   interface,   together   with   the   standard   OrcaFlex   automation   facilities   Ȃ   the   .plt   files  are  automatically  exported  if  you  run  the  direct  SHEAR7  interface  in  batch  mode.  

391  

w

 

Fatigue  Analysis,  Commands    

Each  load  case  is  assigned  an   exposure  level.  For  regular  load  cases  this  is   the  total   number  of  occurrences  of  waves   within   the   wave   class.   For   the   other   methods   the   exposure   level   is   specified   as   the   total   time   exposed   to   waves   within  the  wave  class.   Choice  of  fatigue  analysis  method   As  described  above  OrcaFlex  can  perform  three  different  types  of  fatigue  analysis:  regular,  rainflow  or  spectral.   Rainflow   fatigue   is   the   most   accurate   of   the   methods,   but   also   the   most   time   consuming   and   demanding   of   disk   storage.  The  time  and  storage  requirements  can  be  somewhat  alleviated  by  careful  selection  of  load  cases.  The  other   factor  which  can  be  adjusted  is  the  duration  of  the  irregular  wave  load  case  simulations.  In  our   experience  it  is  often   possible  to  achieve  accurate  damage  predictions  with  simulations  of  20  minutes  duration.   Regular  wave  fatigue  analysis  is  much  faster  and  requires  much  less  disk  storage  than  rainflow  fatigue.  The   wave   scatter  conversion  facility  provides  an  efficient  and  productive  way  to  generate  a  regular  wave  scatter  table  from  a   random   sea   scatter   table.   Provided   that   the   regular   wave   bin   discretisation   is   performed   well,   the   results   from   a   regular  wave  fatigue  analysis  will  generally  agree  well  with  an  equivalent  rainflow  analysis.   The   spectral   fatigue   analysis   method   was   originally   included   to   provide   a   very   quick   alternative   to   the   other   methods.  The  spectral  fatigue  method  in  OrcaFlex  is  much  more  difficult  to  use  effectively  than  the  other  methods.   This  is  largely  due  to   weaknesses  and  limitations  in  the  response  calculation  approach  used  to  generate   response   RAOs.   If   you   do   perform   a   spectral   fatigue   analysis   in   OrcaFlex   then   it   is   very   important   that   you   check   that   the   spectral  response  RAOs  are  smooth.  The  response  calculation  method  often  results  in  very  noisy  RAOs  which  in  turn   result   in   gross   over-­‐predictions   of   damage.   It   is   our   experience   that   use   of   the   spectral   fatigue   method   usually   results  in  poor  and  inaccurate  results.   Recommendations  

The   advent   of   multi-­‐core   processors   and   the   wave   scatter   conversion   facility   mean   that   regular   wave   fatigue   analysis  is  often   just   as   fast  as   spectral   fatigue   analysis,  as   well   as   giving  much  more   reliable   and   accurate   answers.   Because  of  this  we  no  longer  recommend  the  use  of  spectral  fatigue  analysis  in  OrcaFlex.   This  then  reduces  the  choice  of  methods  to  regular  and  rainflow.  Because  of  the  calculation  time  and  disk  storage   advantages  it  is  clearly  desirable  to  use  regular  wave  fatigue.  Certainly  during  system  design  these  advantages  are   significant  because  they  allow  for  greater  coverage  and  exploration  of  the  design  space.   Another  effective  strategy  is  to  use  regular  fatigue  analysis  for  the  bulk  of  the  time  and  switch  to  rainflow  analysis   for  a  final,  more  detailed  check.  If  the  regular  wave  fatigue  analysis  predicts  a  system  life  significantly  in  excess  of   the  design  life  then  this  final  detailed  check  could  be  omitted.  

8.1

COMMANDS  

File  Menu    New  

Clears  previously  entered  Fatigue  Analysis  data  and  resets  data  to  default  values.    Open  

Opens   a   Fatigue   Analysis   file   (.ftg).   If  the   file   contains   results   then   these   will   be   available   without   having   to   perform   the  time-­‐consuming  calculation  again.    Save  

Saves   the   data   to  the   currently   selected   file   name   (shown   in   title   bar   of  the   window).   If   results   have   been   calculated   then  these  are  also  saved  to  the  file.  This  allows  you  to  view  results  at  a  later  date   without  having  to  perform  the   calculation  again.    Save  As  

This  is  the  same  as  Save,  but  allows  you  to  specify  the  file  name  to  save  to.  

392  

w

 

Fatigue  Analysis,  Data  

 

Open  Data  

If  the  file  contains  results  and  there  are  a  large  number   of  load  cases  then  the  file  can   take  a  long  time  to  load.  If  you   want  to  work  with  just  the  input  data  then  this  command  loads  just  the  input  data  which  is  a  much  quicker  process.   Most  Recent  Files  List  

A   list   of  the   most   recently   used   files.   Selecting   an   item   on   the   list   causes   the   file   to   be   loaded.   The   size   of   the   list  can   be  adjusted  from  the  Preferences  form.   Analysis  Menu   Estimate  Calculation  Time  

Gives  an  estimate  of  how  long  it  will  take  to  do  the  fatigue  analysis  and  present  the  results.  This  is  useful  for  long   analyses,  e.g.  rainflow  analyses  involving  a  lot  of  cases  or  long  simulations.    Check  

The   Check   command   performs   a   preliminary   check   of   the   fatigue   analysis   data.   For   example   it   checks   that   all   the   specified   load   case   simulation   files  exist   and   that   the  named   line   and   the   specified   arc   length  intervals   exist   in  each   load  case.   The   Check   command   is   generally   much   quicker   that   the   fatigue   analysis   itself,   so   we   recommend   that   the   Check   command   is   used   before   the   Fatigue   Analysis   is   run,   since   the   check   can   often   detect   data   errors   that   would   otherwise  only  be  found  part  way  through  what  may  be  quite  a  long  fatigue  analysis.  It  is  particularly  important  to   use  the  Check  command  when  a  new  fatigue  analysis  has  been  first  set  up  or  when  significant  changes  have   been   made  to  the  data.    Calculate  

The   Calculate   command   starts   the   Fatigue   Analysis.  The  fatigue   analysis  can   take   a   long   time   if   there   are   many  load   cases,  or  if  there  are  many  log  samples  in  the  load  case  simulations,  or  finally  if  there  are  a  lot  of  segments  in  the  arc   length  intervals  specified.  A  progress  window  is  displayed  and  you  can  cancel  the  analysis  if  desired.   When  the  calculation  is  complete  the  results  are  displayed  in  a  spreadsheet  window.  

8.2

DATA  

Title  

Used  to  label  all  output  of  the  fatigue  analysis.   Damage  calculation  

A  variety  of  methods  are  available  for  calculating  damage:   x

Homogeneous   pipe   stress   which   assumes   a   pipe   made   of   a   straight,   uniform,   homogeneous,   linear   material.   The  damage  calculation  is  based  on  ZZ  Stress.  This  option  is  appropriate  for  metal  risers.  

x

An   approach   based   on   stress   factors.   Here   the   stress   is   assumed   to   comprise   a   tensile   contribution   (proportional   to   either   wall   tension   or   effective   tension)   and   a   bending   contribution   (proportional   to   curvature).  This  approach  is  commonly  used  for  umbilicals  but  could  also  be  used  for  unbonded  flexibles.  

x

The  mooring  fatigue  option  calculates  damage  from  effective  tension  ranges  using  T -­‐N  curves.  

x

The   SHEAR7   option   is   rather   different   because   damage   is   calculated   external   to   OrcaFlex   by   SHEAR7.   This   option  allows  you  to  collate  fatigue  damage  from  a  number  of  SHEAR7  load  cases.  

x

The  final  option,  externally  calculated  stress,  is  not  available  for  general  use.  

Analysis  Type  

Three  types  of  fatigue  analysis  are  available:   x

Regular  analysis  must  be  based  on  a  series  of   regular  wave  simulations  that  represent  the  various  load  cases   that   will   occur.   For   each   of   these   load   cases   a   single-­‐occurrence   damage   value   is   calculated   based   on   the   last   wave   cycle   in   the   simulation.   This   damage   value   is   then   scaled   up  by   the   specified   number   of  cycles   expected   to   occur   during   the   structure's   life,   and   this   gives   the   total   load   case   damage   value.   Finally   these   total   load   case   damage  values  are  then  summed  for  each  load  case  to  give  the  overall  total  damage.  

393  

w

 

Fatigue  Analysis,  Load  Cases  Data  for  Regular  Analysis    

x

Rainflow   analysis   is   normally   based   on   a   series   of  random   wave   simulations.   It   uses   a   cycle   counting   technique   to   break   down   each   random   wave   case   into   a   series   of   half   cycles,   and   then   sums   the   damage   from   each   half   cycle  according  to  the  Palmgren-­‐Miner  law.  For  details  see  the  book  by  Maddox  and  the  paper  by  Rychlik.  This   gives  the  damage  value  for  that  load  case,  which  is  then   scaled  to  the  specified  total  exposure  time.  Finally  these   total  load  case  damage  values  are  then  summed  for  each  load  case  to  give  the  overall  total  damage.  

x

Spectral  analysis  calculates   damage  in  the  frequency  domain  using  statistical  methods.  The  method  requires  a   power   spectral   density   function   (PSD)   for   a   particular   load   variable   (stress   or   tension).   The   PSD   is   obtained   from  a  response  calculation  simulation.  This  calculates  RAOs  for  the  load  variables  of  interest  and  these  are  then   combined   with   the   load   case   wave   spectrum   to   give   PSDs   for   load.   These   PSDs   are   then   used   to   calculate   damage   using   either   Dirlik's   formula   or   the   Rayleigh   distribution.   The   damage   is   scaled   to   the   specified   total   exposure  time  for  the  load  case.  Finally  these  total  load  case  damage  values  are  then   summed   for   each  load  case   to  give  the  overall  total  damage.  

Units  

The   units   to   be   used   for   the   fatigue   analysis,   for   both   the   fatigue   analysis   data   and   for   its   results.   The   units   are   specified  in  the  same  way  as  elsewhere  in  OrcaFlex.   Note  that  the  units  specified  for  the  fatigue  analysis  need  not  match  the  units  that  were  used  in  the  various  load  case   simulation  files.  If  they  do  not  match,  then  the  results  from  that  simulation  file  will  automatically  be  converted  to  the   units   specified   for   the   fatigue   analysis.  This  is  useful,   since   it   allows   the   fatigue   analysis  to   be   done   using   m   and  MN   as   the   length   and   force   units   (giving   stresses   in   MN/m2   =   MPa),   for   example,   even   if  the   simulation   load   cases   use   m   and  kN  (which  corresponds  to  stresses  in  kN/m2  =  kPa).  Similarly,  in  US  units,  the  fatigue  analysis  can  use  inches   (giving  stresses  in  ksi)  even  if  the  simulation  files  use  feet  as  the  length  unit.   If  you  change  units,  then  all  existing  fatigue  analysis  data  is  automatically  changed  to  match  the  new  units.  This  is   useful   if  you   want   to   enter   data   in   some   other   set   of  units,   since   you   can   simply   change   to   the   units   of  the   new  data,   then  enter  the  new  data,  and  then  change  back  to  the  o riginal  units  again.   Cycle  Range  Distribution  (spectral  analysis  only)  

The  spectral  fatigue  calculation  assumes  that  cycle  ranges  follow  a  statistical  distribution  which  is  specified  by  this   data  item.  The  distribution  can  be  either  Dirlik's  formula  or  the  Rayleigh  distribution.  The  Rayleigh  distribution   is   appropriate   if   the   variation   of   the   response   is   a   narrow   banded   random   Gaussian   process.   Dirlik's   formula   is   applicable   even   if   the   variation   of   the   response   is   not   a   narrow   banded   process.   For   this   reason   we   recommend   using  Dirlik's  formula.  

8.3

LOAD  CASES  DATA  FOR  REGULAR  ANALYSIS  

Simulation  File  Name  

The  name  of  the  simulation  file  which  represents  the  load  case.  You  can  either  specify  the  full  path  or  a   relative  path.   Line  Name  

The  name,  in  this  load  case  simulation  file,  of  the  line  to  be  analysed.   Note:  

Normally   the   line   name   will   be   the   same   in   all   of   the   load   cases   (though   this   is   not   necessary).   However  the  named   lines   in   the   various   load   cases   must,   of   course,   all   represent   the   same   physical   line  and  use  the  same  discretisation  in  the  areas  being  analysed.  

Number  of  Cycles  

The  number  of  wave  cycles,  of  this  particular  set  of  load  conditions,  that  the  line  will  experience.  

8.4

LOAD  CASES  DATA  FOR  RAINFLOW  ANALYSIS  

Simulation  File  Name  

The  name  of  the  simulation  file  which  represents  the  load  case.  You  can  either  specify  the  full  path  or  a  relative  path.   Line  Name  

The  name,  in  this  load  case  simulation  file,  of  the  line  to  be  analysed.  

394  

w

 

Fatigue  Analysis,  Load  Cases  Data  for  Spectral  Analysis  

 

Note:  

Normally   the   line   name   will   be   the   same   in   all   of   the   load   cases   (though   this   is   not   necessary).   However  the  named   lines   in   the   various   load   cases   must,   of   course,   all   represent   the   same   physical   line  and  use  the  same  discretisation  in  the  areas  being  analysed.  

Simulation  Period  

The  period  of  the  pre-­‐run  simulation  file  that  defines  the  load  case.   Exposure  Time  

The  total  time  the  system  is  exposed  to  this  load  case.  

8.5

LOAD  CASES  DATA  FOR  SPECTRAL  ANALYSIS  

Simulation  File  Name  

The  name  of  the  simulation  file  which  represents  the  load  case.  You  can  either  specify  the  full  path  or  a   relative  path.   Line  Name  

The  name,  in  this  load  case  simulation  file,  of  the  line  to  be  analysed.   Note:  

Normally   the   line   name   will   be   the   same   in   all   of   the   load   cases   (though   this   is   not   necessary).   However  the  named   lines   in   the   various   load   cases   must,   of   course,   all   represent   the   same   physical   line  and  use  the  same  discretisation  in  the  areas  being  analysed.  

Exposure  Time  

The  total  time  the  system  is  exposed  to  this  load  case.   Spectral  Form,  Spectral  Parameters  

For  spectral  analysis  the  simulation  file  specifies  a  response  calculation  simulation  file  from  which  response  RAOs   are   derived.   The   spectral   fatigue   calculation   then   proceeds   by   combining   these   response   RAOs   with   a   wave   spectrum  to  produce  power  spectral  density  (PSD)  functions.   Spectral  Form  can  be  one  of  JONSWAP,  ISSC,  Ochi-­‐Hubble  or  Torsethaugen.  This  specifies  the  general  form  of  the   wave  spectrum.   The  parameters  for  the  chosen  spectral  form  are  specified  as  follows:   x

If   JONSWAP   is   selected   then   Spectral   Parameters   can   be   either   Automatic,   Partially   Specified   or   Fully   Specified.  This   determines   exactly  how  the  spectral  parameters  for   each  load  case  are  specified.  For  example,   if   Automatic   is   selected   then   you   specify   Hs   and   Tz   and   the   other   JONSWAP   parameters   are   calculated   automatically  and  reported.  For  details  see  Data  for  JONSWAP  and  ISSC  Spectra.  

x

If   ISSC  is  selected   then   you  must   specify   H s  and   Tz  for  each  load   case.   You   can   also  specify  fm  or  Tp  but  since   Tz,   Tp  and  fm  are  tied  together  then  setting  any  one  of  them   changes  the  other  two  to  m atch.  For  details  see   Data  for   JONSWAP  and  ISSC  Spectra.  

x

If   Ochi-­‐Hubble   is   selected   then   Spectral   Parameters   can   be   either   Automatic,   or   Specified.   Again,   this   determines   exactly   how   the   spectral   parameters   for   each   load   case   are   specified.   If   you   select   Automatic   the   program   calculates   the   parameters   of  the   most   probable   spectrum,   based   on   the   overall   significant   wave   height   Hs  that  you  have  specified.  If  you  select  Specified  you  must  specify  all  6  parameters  and  OrcaFlex  then  derives   and  displays  the  corresponding  overall  Hs  and  Tz  values.  For  details  see  Data  for  Ochi-­‐Hubble  Spectrum.  

x

If  Torsethaugen  is  selected   then  you  must  specify  Hs  and  Tp  for  each  load  case.  You  can  also  specify  fm  but  since   Tp   and   fm   are   tied   together   then   setting   either   one   changes   the   other   to   match.   For   details   see   Data   for   Torsethaugen  Spectrum.  

Setting  up  load  cases  for  Spectral  Analysis   When   performing   a   spectral   fatigue   analysis   you   will   typically   have   a   wave   scatter   table   describing   the   relative   probability  of  storm  occurrence.  This  determines  a  number  of  wave  classes,  e.g.  storms  defined  by  H s,Tz  pairs.   The   load   cases   data   should   be   setup   to   match   load   cases   with   wave   classes.   For   example,   suppose   that   you   were   working  with  the  following  (truncated)  wave  scatter  table:     4-­‐5       3-­‐4    

  6  

9  

3    

18  

6    

395  

w

 

Fatigue  Analysis,  Load  Cases  Data  for  Spectral  Analysis    

Hs   2-­‐3    

22   132   117    

  1-­‐2  

3  

  0-­‐1  

15  

       

57   201   249     48  

69  

45    

4-­‐5   5-­‐6   6-­‐7   7-­‐8      

Tz  

 

 

The   values   in   the   table   represent   joint   probabilities   in   parts   per   thousand,   so   that   a   value   of   201   represents   a   probability  of  0.201.   This  wave  scatter  table  gives  16  wave  classes  and  so  the  fatigue  analysis  data  in  OrcaFlex  would  be  setup  with  16   corresponding  load  cases  with  appropriate  Hs  and  Tz  values.   Simulation  files  for  spectral  fatigue  analysis  load  cases  

The  simulation  files  used  to  represent  a  load  case  for  spectral  fatigue  analysis  should  model  all  aspects  of  the  system   and   environment   other   than   the   wave   spectrum.   So   you   must   specify   vessel   offset,   current   profile   and   direction,   wave  direction  and  so  on  which  are  appropriate  for  the  load  case  being  analysed.   The   wave   type   for   the   load   case   simulation   file   must   be   response   calculation.   This   effectively   calculates   system   responses   (i.e.   RAOs)   for   a  range   of   wave   frequencies.  The   spectral   fatigue   analysis  then   combines   these   RAOs   with   the  load  case  wave  spectra  (i.e.  the  Hs,Tz  pairs)  to  produce  fatigue  damage  estimates  for  the  load  case.   Choice  of  Hs  for  response  calculation  simulation  files  

The   Spectral   Response   Analysis   method   which   is   used   to   calculate   system   responses   (RAOs)   includes   non-­‐linear   effects   such  as   hydrodynamic   drag.   In   order  for   these  non-­‐linear  effects   to   be   well   modelled   the   choice   of  Hs  for   the   response   calculation   simulation   files   is   important.   Essentially   the   RAOs   can   be   considered   as   being   dependent   on   wave  height.  How  significant  this  dependence  is  will  vary  from  case  to  case.  Certain  systems  are  dominated  by  linear   physical  effects  and   the  RAOs  may  not  in  fact   be  dependent  on   wave  height.  To  determine  how  significant  this   e ffect   is  we  would  recommend  sensitivity  studies.   In   the   example   above   we   might   choose   to   run   a   response   calculation   simulation   for   each   row   of   the   wave   scatter   table  (assuming  that  the  system  had  significant  non-­‐linearities).   This  would  give   5  simulation  files   for   Hs   ranges  0-­‐1,   1-­‐2,  2-­‐3,  3-­‐4  and  4-­‐5.  There  are  4  wave  classes  corresponding  to  the  0-­‐1  Hs  range.  The  load  case  corresponding  to   each   of   these   wave   classes   would   then   be   represented   by   the   same   simulation   file.   The   other   H s   ranges   are   dealt   with  similarly  and  so  the  load  cases  table  would  look  as  below:  

396  

w

 

Fatigue  Analysis,  Load  Cases  Data  for  SHEAR7  

 

  Figure:  

Example  load  cases  table  

If   the   non-­‐linearities   in   the   system   are   not  so   significant   then   you   may  be   able   to   obtain   accurate   results   with   fewer   simulation  files.  This  may  be  desirable  to  reduce  the  amount  of  time  taken  to  run  the  simulations.  For  example  the   Hs1,   Hs2   and   Hs3   simulations   could   be   combined   into   a   single   Hs2   simulation   etc.   Again,   the   accuracy   of   such   a   simplification  should  be  tested  with  sensitivity  studies.   Response  calculation  simulation  duration  

The   other   decision   to   make   is   over   the   length   of   the  response   calculation   simulations.   You   need   to   simulate   for   long   enough  to  get  accurate  results.  As  for  the  issue  of  H s  discussed  above  we  would  recommend  using  sensitivity  studies   to  determine  how  long  is  required.  

8.6

LOAD  CASES  DATA  FOR  SHEAR7  

Load  Case  File  Name  

The   name   of   the   SHEAR7   .plt   output   file   which   represents   the   load   case.   You   can   either   specify   the   full   path   or   a   relative  path.   Exposure  Time  

The   total   time   the   system   is  exposed   to   this   load   case.   The   damage   for   the   load   case   is   calculated   by   multiplying   the   exposure  time  by  the  damage  rate  read  from  the  load  case  .plt  file.  

8.7

COMPONENTS  DATA  

The  Components  Data  page  is  only  available  when  damage  is  calculated  using  stress  factors.  

397  

w

 

Fatigue  Analysis,  Analysis  Data    

You   define   a   number   of   components   for   which   damage   is   to   be   calculated.   Components   can   be   used   to   represent   different  layers  or  components  in  the  cross-­‐section  of  an  umbilical  or  a  flexible.   Component  Name  

This  is  used  to  identify  the  component  in  the  results.   Tension  Variable  

Specifies  wall  tension  or  effective  tension  is  used  to  calculate  stress.   Tension  and  Curvature  Stress  Factors  

The  stresses  used  to  calculate  damage  are  calculated  according  to  the  formula:   S  =  KtT  +  Kc(Cx•‹Ʌ-­‐  Cy…‘•ɅȌ   where   S  is  stress,   Kt  and  Kc  are  the  tension  and  curvature  stress  factors,  respectively,   T  is  either  wall  tension  or  effective  tension,  as  specified  by  the  tension  variable  data,   Cx  and  Cy  are  the  components  of  curvature  in  the  line's  local  x  and  y  directions,  respectively,  and   Ʌ‹•–Š‡…‹”…—ˆ‡”‡–‹ƒŽŽ‘…ƒ–‹‘‘ˆ–Š‡fatigue  point.   In  effect  this  formula  defines  stress  to  be  the  sum  of  contributions  due  to  direct  tensile  strain  and  bending  strain.   Š‡ …‹”…—ˆ‡”‡–‹ƒŽ ˜ƒ”‹ƒ–‹‘ ȋ‹Ǥ‡Ǥ –Š‡ –‡”• ™Š‹…Š ”‡ˆ‡” –‘ ɅȌ ‹• –‘ ƒ……‘—– ˆ‘” –Š‡ ˆƒ…– –Šƒ– „‡†‹‰ •–”ƒ‹ ˜ƒ”‹‡• ™‹–ŠɅ.   So,  for   a   point   in   the  plane   of  bending,   stress   is  given  by  S  =   K tT   ±   Kc|C|,  where   C   is  the   curvature   vector   (Cx,   Cy).  Similarly,  for  a  point  at  90°  to  the  plane  of  bending,  stress  is  given  by  S  =  KtT.   The   stress   factors   will   typically   be   calculated   from   experimental   data   or   from   detailed   analytic   models   of   the   umbilical  or  riser  cross-­‐section.  Suppliers  of  such  products  are  usually  able  to  provide  the  necessary  stress  factors.   S-­‐N  Curve  

Specifies  which  S-­‐N  curve  is  used  for  damage  calculations  for  this  component.  

8.8

ANALYSIS  DATA  

The  Analysis  Data  page  contains  the  following  data  items,  which  specify  the  parts  of  the  line  to  be  analysed.   Critical  Damage  

Is   a   warning   level.   If   the  total   damage   at   any  fatigue  point   exceeds  the   Critical   Damage   then   that  damage   figure  will   be  highlighted  in  the  results.   Number  of  Thetas  

The   number   of  points   (N)   around   the   pipe   circumference,   at   which   fatigue   analysis   will   be   performed.   There   will   be   N  fatigue  points  uniformly  distributed  at  360°/N  intervals  around  the  pipe  circumference.  A  larger  number  of  thetas   gives  a  more  comprehensive  analysis,  but  takes  a  little  longer.   This  data  is  not  required,  and  hence  not  available,  for  mooring  and  SHEAR7  fatigue.   Radial  Position  (homogeneous  pipe  stresses  only)  

Specifies  whether  the  fatigue  analysis  is  performed  at  the  inner  or  outer  fibre  of  the  pipe.   Line  Length  (SHEAR7  fatigue  only)  

Specifies   the   length   of   the   line   being   consider   by   the   SHEAR7   analysis.   This   data   item   is   implemented   solely   to   provide   some   convenience   to   results   reporting.   A   value   of   '~'   results   in   arc   lengths   being   reported   as   non-­‐ dimensional  x/L  values,  the  native  form  for  SHEAR7.  If  a  value  is  specified  for  the  line  length  then  the  x/L  values  are   re-­‐dimensionalised  using  that  specified  value.   Arc  Length  Intervals  

You   define   the   parts   of   the   line   that   are   to   be   analysed   by   specifying   a   number   of   non-­‐overlapping   Arc   Length   Intervals  in  the  form  of  From  and  To  arc  length  values.  OrcaFlex  will  analyse  cross-­‐sections  at  each  line  end  and   mid-­‐•‡‰‡–™Š‘•‡ƒ”…Ž‡‰–Šœ‹•‹–Š‡”ƒ‰‡ ”‘ζœζ‘Ǥ  

398  

w

 

Fatigue  Analysis,  S-­‐N  and  T-­‐N  Curves  

 

For  simple  cases  you  can  use  just  one  arc  length  interval  covering  the  whole  line.   However  it  is  often  clear   which   part,  or  parts,  of  the  line  are  liable  to  fatigue  problems.  If  calculation  time  is  significant  then  you  can  save  calculation   time  by  analysing  those  parts  of  the  line.   Warning:  

The  included  arc  lengths  must  be  the  same  in  each  load  case,  so  the  line  to  be  analysed   should  have   the  same  number  and  distribution  of  segments  in  each  of  the  load  case  simulations.  

SCF  and  Thickness  Correction  Factor  (homogeneous  pipe  stresses  only)  

When   stress   ranges   are   used   with   the   S-­‐N   curve   to   calculate   damage,   the   stress   ranges   are   scaled   by   the   Stress   Concentration  Factor  (SCF)  and  the  Thickness  Correction  Factor  before  calculating  damage.   If  no  stress  correction  is  required  then  these  factors  should  both  be  set  to  1.   Notes:  

To   use   different   stress   correction   factors   for   different   parts   of   the   line,   you   will   need   to   specify   separate  arc  length  intervals  for  those  parts.  

S-­‐N  Curve  (homogeneous  pipe  stresses  only)  

Specifies  which  S-­‐N  curve  is  used  for  damage  calculations  in  this  arc  length  interval.   T-­‐N  Curve  (mooring  fatigue  only)  

Specifies  which  T-­‐N  curve  is  used  for  damage  calculations  in  this  arc  length  interval.  

8.9

S-­‐N  AND  T-­‐N  CURVES  

S-­‐N  Curves  

An   S-­‐N   curve   defines   the   number   of   cycles   to   failure,   N(S),   when   a   material   is   repeatedly   cycled   through   a   given   stress  range  S.  OrcaFlex  uses  the  S-­‐N  curve  to  calculate   the  damage  in  a  fatigue  analysis.  If  needed  you  can  define  a   number  of  different  S-­‐N  curves  and  use  them  at  different  arc  lengths  along  a  line.   With  each  S-­‐N  curve  you  must  also  specify  an  associated  stress  endurance  limit,  FL,  which  is  the  stress  range  below   which  no  damage  occurs.   The  S-­‐N  curve  itself  can  be  specified  either  by  parameters  or  by  a  table.   When   the   curve   is  specified  by  parameters  the   user   specifies   two   parameters,   A   and   b,   and   the   curve   is  then   given   by  either  of  the  following  equivalent  formulae:   N  =  10A  S-­‐b   Log10(N)  =  A  -­‐  b  Log10(S)   When  the  curve  is  specified  by  a  table  the  user  gives  a  table  of  corresponding  values  of  S  and  N.  For  other  values  of  S   we  use  log  linear  interpolation  or  extrapolation  to  find  the  value  of  N.   Mean  stress  effects  can  be  accounted  for  using  Goodman,  Soderberg  or  Gerber  models.   For  details  of  how  the  S-­‐N  curve  is  used  to  calculate  the  damage  see  How  Damage  is  Calculated.   S-­‐N  Curve  Units  

The   S-­‐N   curve   parameters   entered   must   be   consistent   with   the   fatigue   analysis   units.   S-­‐N   curve   parameters   are   typically  quoted  with  respect  to  stresses  in  MPa,  but  you  might  be  doing  the  fatigue  analysis  using  some  other  stress   units.  You  can  handle  this  problem  as  follows.  First  change  the  fatigue  analysis  units  and  set  the  units  system  to  be   'User',   the   length   units   to  be   'mm'   and   the   force   units   to   be   'N'.   This   corresponds   to   stresses   in   MPa,   so   you   can   then   enter  the  S-­‐N  parameters  in  terms  of  MPa.  Finally,  restore  the  units  to  those  that  you  want  for  the  fatigue  analysis.   The  parameters  will  automatically  be  converted  to  allow  for  the  change  in  units.   T-­‐N  Curves  

For  mooring  fatigue  damage  is  calculated  with  T-­‐N  curves.  These  define  the  number  of  cycles  to  failure,  N(T),  when  a   material  is  repeatedly  cycled  through  a  given  effective  tension  range  T.   The  T-­‐N  curve  can  be  specified  either  by  parameters  or  by  a  table.   When   the   curve   is   specified   by   parameters   the   user   specifies   three   parameters,   m,   k   and   the   reference   breaking   strength  (RBS).  The  curve  is  then  given  by  the  following  formula:   N  =  k(T/RBS)-­‐m  

399  

w

 

Fatigue  Analysis,  Integration  Parameters    

When   the   curve   is   specified   by   a  table   the   user   gives   a   table   of  corresponding   values   of  T  and   N.   For   other   values   of   T  we  use  log  linear  interpolation  or  extrapolation  to  find  the  value  of  N.   For  details  of  how  the  T-­‐N  curve  is  used  to  calculate  the  damage  see  How  Damage  is  Calculated.  

8.10

INTEGRATION  PARAMETERS  

Integration  parameters  are   only  available  when  the   spectral  fatigue  analysis  option   is  selected.  Normally  you  should   not   need   to   modify   the   parameters   from   their   default   values.   However,   if   the   spectral   fatigue   calculation   has   problems  with  convergence  then  it  may  prove  productive  to  try  different  values  for  these  parameters.   We   recommend   that   you   contact   Orcina   for   further   details   should   you   encounter   problems   with   spectral   fatigue   convergence.  

8.11

RESULTS  

Fatigue  results  are  presented  in  a  separate  window.  Results  are  available  in  either  graphical  or  tabular  form.   Graphical  output   The  graphical  output  produces  plots  of  fatigue  damage  or  fatigue  life  against  arc  length  Ȃ  range  graphs  of  fatigue.   ‡†‡‘–‡ƒ†ƒƒ‰‡˜ƒŽ—‡ƒ–ƒ”…Ž‡‰–Šœƒ†…‹”…—ˆ‡”‡–‹ƒŽ’‘•‹–‹‘ƒ•ȋœǡɅȌǤŠ‹•…‘—Ž†„‡‘˜‡”ƒŽŽ–‘–ƒŽ†ƒƒ‰‡ or  the  total  exposure  damage  value  for  an  individual  load  case.  The  damage  graphs  all  plot  D(z)  =  maxɅ  ȋœǡɅȌǡ–Šƒ–‹• the  maximum  damage  value  at  a  given  arc  length  z.   The  life  graphs  plot  the  corresponding  life  values  T/D(z)  where  T  is  the  exposure  time.  For  individual  load  cases  T  is   the  exposure  time  for  the  load  case.  For  total  life  T  is  the  total  exposure  time  of  all  the  load  cases.   Graphs  are  available  for:   x

Total  damage/life:  the  overall  total  damage/life.  

x

Individual  load  case  damage/life:  the  total  exposure  damage/life  from  the  specified  load  case.  

x

Worst  cases:  the  total  exposure  damage  value  from  the  five  most  damaging  load  cases.  By  most  damaging  we   mean  the  load  cases  with  the  largest  values  of  maxœǡɅ  ȋœǡɅȌǡ–Š‡ƒš‹—†ƒƒ‰‡˜ƒŽ—‡‘˜‡”ƒŽŽœǡɅǤ  

The  graphs  can  be  customised  in  a  number  of  ways:   x

The   arc   length   axis   can   be   either   horizontal   or   vertical,   the   latter   option   being   more   appropriate   for   vertical   risers.  

x

The   arc   length   axis   can   be   inverted.   When   it   is   inverted   increasing   values   run   from   right   to   left   (if   it   is   horizontal)  and  top  to  bottom  (if  it  is  vertical).  Again  this  is  particularly  useful  for  vertical  risers   with  arc  length   values  that  increase  from  the  top  end  to  the  bottom  end.  

x

Fatigue  damage/life  can  optionally  be  plotted  on  a  logarithmic  scale.  

x

Individual  arc  length  intervals  can  be  plotted.  

Tabular  output   The  tabular  output  is  presented  in  a  spreadsheet  that  has  one  Damage  Tables  sheet,  plus  one  Load  Case  sheet  for   each  load  case.  There  is  also  a  sheet  echoing  the  S-­‐N  curve  data.   Load  Case  sheets  

The  Load  Case  sheets  contain  the  derived  stress   results   for   each  fatigue   point  that  has  been  analysed,   together   with   general  information  such  as  the  environmental  data  that  applied  to  that  load  case.   There   is  one   table   of   stress   results   for   each  arc   length  covered  by  the   specified   arc   length  intervals.   Each  such  table   contains  a  row  of  results  for  each  fatigue  point  in  that  arc  length  cross-­‐section.  These  results  are  the  stress  ranges   (for  each  of  the  stress  components),  the  maximum  stress  range  and  the  resulting  load  case  damage  values.   In   addition   to   the   detailed   tables   a   damage   summary   table   is   presented   which   tabulates   the   load   case   damage   at   each  fatigue  point.   For  mooring  fatigue,  tension  results  rather  than  stress  results  are  reported.  For  SHEAR7  fatigue,  since  the  damage   calculation  has  been  performed  by  SHEAR7,  only  damage  values  are  reported.  

400  

w

 

Fatigue  Analysis,  Fatigue  Points  

 

Damage  Table  sheet  

The  Damage  Table  sheet  starts  with  an  Excessive  Damage  table,  which  lists  any  fatigue  points  at  which  the  overall   total  damage  has  exceeded  the  specified  critical  damage  value.  Details  of  where  on  the  line  the  worst  total  damage   occurred  are  also  reported.   A  table  summarising  the  overall  damage  over  total  exposure  for  all  arc  lengths  is  presented.   Finally,  the  Damage  Table  sheet  provides  damage  tables  for  each  arc  length  cross-­‐section  analysed.  These  report,  for   each   fatigue   point   in   the   cross-­‐section,   the   total   exposure   damage   value   from   each   load   case   and   the   overall   total   damage.   In  all  of  these  tables,  overall  total  damage  values  that  exceed  the  specified  critical  damage  value  are  highlighted  in   red.   Printing  and  Exporting  

To  save  the  results  you  will  need  to  export  the  spreadsheet  as  an  Excel  sheet.  If  you  want  to  print  the  results  then  for   best  results  you  should  first  export  them  and  then  use  Excel  to  do  the  printing.   Customising  results  output   The  tabular  results  output  can  be  customised  using  the  options  on  the  Results  page  of  the  main  fatigue  form.   Output  load  case  tables  

If   this   option   is   deselected   then   the   load   case   sheets   are   omitted   from   the   fatigue   results.   This   can   significantly   reduce  the  time  and  memory  required  to  generate  the  results  tables.   Output  detailed  load  case  tables  

If   this   option   is   deselected   then   the   detailed   tables   on   the   load   case   sheets   are   omitted   from   the   fatigue   results.   This   can  significantly  reduce  the  time  and  memory  required  to  generate  the  results  tables.   Load  case  damage  units  

Load   case   damage  values   can   be  reported   as   damage   per   hour,   damage   per   d ay  or   damage   per   year,   as   specified   by   this  data  item.  

8.12

FATIGUE  POINTS  

Damage  is  calculated  at  a  number  of  line  end  and  mid-­‐segment  cross-­‐sections  along  the  line,  as  specified  by  defining   Arc  Length  Intervals  in  the  Analysis  Data.  Each  included  arc  length  defines  a  cross-­‐section  through  the  pipe.   Damage  calculated  by  homogeneous  pipe  stresses  

Each   cross-­‐•‡…–‹‘‹•†‡•…”‹„‡†„›’‘Žƒ”…‘‘”†‹ƒ–‡•ȋǡɅȌwhere   R   is   the   radial   distance   from   the   centre-­‐Ž‹‡ƒ†Ʌ is  measured  from  the  line's  local  x-­‐axis  towards  its  y-­‐axis.   OrcaFlex  calculates  the  damage  at  either  the  inner  or  outer  fibre  of  the  pipe  and  so  R  is  either  ODstress/2  or  IDstress/2.   ‘”Ʌ™‡…ƒŽ…—Žƒ–‡†ƒƒ‰‡ƒ–‡“—ƒŽŽ›•’ƒ…‡†‹–‡”˜ƒŽ•†‡–‡”‹‡†„›–Š‡•’‡…‹ˆ‹‡†Number  of  Thetas.  For  example,  if   you  specify  16  theta  values,  then  they  will  be  equally  spaced  at  360°/16   =  22.5°  intervals  0°,  22.5°,  45°  etc.   Damage  calculated  by  stress  factors  

For   stress   factor   fatigue   the   radial   position   is   implicit   in   the   stress   factors   and   so   is   not   explicitly   used   in   the   calculation.   Damage   is   calculated   at   circumferential   locations   determined   by   the   specified   Number   of   Thetas,   as   described  above.   Mooring  fatigue  damage  

For   mooring   fatigue   there   is   no   need   to   consider   radial   and   circumferential   variation   and   so   there   is   one   fatigue   point  for  each  arc  length  considered.  

8.13

HOW  DAMAGE  IS  CALCULATED  

For  each  load  case  and  fatigue  point  OrcaFlex  calculates  damage  values  as  follows:   x

The   time   history  of  response,   in   that   load   case,   at   that   fatigue   point,   is   calculated.   For  homogeneous   pipes   the   ZZ   stress   variable   is   used.   For   stress   factor   fatigue   then   the   stress   is   calculated   based   on   the   stress   factors.   For   mooring  fatigue  the  response  variable  is  effective  tension.  

401  

w

 

Fatigue  Analysis,  How  Damage  is  Calculated    

x

The   damage   value   corresponding   to  the   response   time   history   is   calculated  Ȃ   see   below   for   details.   This   value   is   the  damage  value  at  that  fatigue  point  due  to  one  occurrence  of  that  load  case.  

x

The  load  case  damage  values  are  scaled  to  allow  for  the  exposure  associated  with  that  load  case.  

x

The  above  step  gives  the  total  exposure  damage  value  from  that  load  case  at  this  fatigue  point.  

x

Finally,   these   total   exposure   load   case   damage   values   are   summed   over   all   load   cases   to   obtain   the   overall   total   damage  value  at  that  fatigue  point.  

Damage  Calculation  using  S-­‐N  curves   The   S-­‐N  curve  defines  the   number   of  cycles  to  failure,  N(S),  for  stress  range  S,  and  also  defines  a  endurance  limit,  FL,   below  which  no  damage  occurs.  OrcaFlex  uses  these  to  calculate  a  damage  value  given  by:   D(S)  =  1/N(S)  if  S  >  FL   ȋȌαͲ‹ˆζ L   This  damage  value  can  be  thought  of  as  the  proportion  of  the  fatigue  life  that  is  used  up  by  1  cycle  of  stress  range  S.   If  the  S-­‐N  curve  is  defined  by  parameters  then  for  S  >  FL  we  have:   Log10(N)  =  A  -­‐  b  Log10(S)   so  D(S)  can  be  expressed  in  the  following  form:   D(S)  =  10-­‐A  Sb.   Mean  stress  effects  

Mean   stress   effects   are   handled   by   modifying   each   stress   range   according   to   a   formula   dependent   on   the   mean   stress   level.   Three   models   of   mean   stress   effects   are   provided:   Goodman,   Soderberg   and   Gerber.   Each   method   is   defined  by  a  formula  for  Se,  the  equivalent  stress,  allowing  for  mean  stress  effects,  to  be  used  in  the  zero  mean  stress   S-­‐N  curve.   The  Goodman  model  is:   Se  =  Sr  /  (1  -­‐  Sm/SMTS)  for  0  <  Sm  <  SMTS   Se  =  Sr  for  -­‐SMTS  <  Sm  ζͲ   where  Sr  is  the  true  stress  range,  Sm  is  the  mean  stress  and  SMTS  is  the  ultimate  tensile  strength  as  specified  in  the  S -­‐ N  data.   The  Soderberg  model  is:   Se  =  Sr  /  (1  -­‐  Sm/SMYS)  for  0  <  Sm  <  SMYS   Se  =  Sr  for  -­‐SMYS  <  Sm  ζͲ   where  SMYS  is  the  yield  strength  as  specified  in  the  S-­‐N  data.   The  Gerber  model  is:   Se  =  Sr  /  (1  -­‐  [Sm/SMTS]2)  for  -­‐SMTS  <  Sm  <  SMTS   The  definition   of   mean   stress,   Sm,   depends  on   the   analysis  method  being   used.   For  regular   analysis,  S m  is  defined   to   be   the   mean   of   the   min   and   max   stress   values   associated   with   the   stress   range   S r.   For   rainflow   analysis,   Sm   is   the   mean   value   of   the   local   turning   points   in   the   stress   time   history.   For   spectral   fatigue,   S m   is   the   mean   of   the   time   history  used  to  determine  the  stress  RAOs.   Regular  analysis  

The   minimum   and   maximum   values   of   stress   over   the   last   simulated   wave   cycle   define   a   stress   range   S.   The   associated   single-­‐‘……—””‡…‡ Ž‘ƒ† …ƒ•‡ †ƒƒ‰‡ ˜ƒŽ—‡ ‹• –Š‡ ‰‹˜‡ „› ȋɈȌ ™Š‡”‡ Ɉ ‹• –Š‡ ’”‘†—…– ‘ˆ –Š‡ stress   concentration   factor   and   the   thickness   correction   factor.   If   mean   stress   effects   are   included   then   the   equivalent   stress  range  Se  is  used.   Rainflow  analysis  

The  stress  time  history  is  analysed  using  the  rainflow  cycle  counting  method.  This  gives  a  number  of  stress  ranges   for  half  cycles,  say  Si  where   i  runs  from  1   to  the  number   of  stress  ranges.  The  associated  single-­‐occurrence  load  case   damage   value   is   then   ‰‹˜‡„›ΦσȋɈi)   where   the   summation   is   over   all   the   half   cycles.   Note   that   the   factor   of   one  

402  

w

 

Fatigue  Analysis,  How  Damage  is  Calculated  

 

half  is   present   because   the   rainflow   algorithm   counts   half   cycles   rather   than   full-­‐cycles.   Again,   if   mean   stress   effects   are  included,  then  the  equivalent  stress  range  Se  is  used.   Spectral  analysis  

For   spectral   analysis   damage   is   calculated   in   the   frequency   domain   using   statistical   methods.   The   calculation   requires   a   power   spectral   density   function   (PSD)   for   stress.   The   PSD   is   obtained   from   a   response   calculation   simulation.   This   calculates   RAOs   for   stress   which   is   then   combined   with   the   load   case   wave   spectrum   to   give   the   PSD   for   stress.   These   PSD   are   then   used   to   calculate   damage   using   either   Dirlik's   formula   or   the   Rayleigh   distribution.  The  stress  concentration   factor,  thickness   correction  factor  and  mean   stress   effects  are  all  accounted   for  in  the  spectral  damage  calculation.   For  detailed  references  on  how  spectral  fatigue  analysis  calculates  damage  from  stress  PSDs  please  refer  to:   x

Barltrop  and  Adams  (1991)  which  has  an  excellent  section  on  spectral  fatigue  analysis.  

x

Dirlik  (1985)  is  the  original  reference  for  Dirlik's  stress  range  distribution  formula.  

Damage  Calculation  using  T-­‐N  curves   T-­‐N   curves   are   handled   in   a   similar   way.   A   T-­‐N   curve   defines   the   number   of   cycles   to   failure,   N(T),   for   effective   tension   range   T.   There   is   no   analog   of   endurance   limit   for   T-­‐N   curves.   Likewise   there   are   no   analogs   of   stress   concentration  factor  and  thickness  correction  factor.   As  for  S-­‐N  curves,  OrcaFlex  defines  damage  as:   D(T)  =  1/N(T)   The  summation  of  damage  is  then  performed  in  an  identical  manner  to  that  performed  for  S -­‐N  curves.   Damage  Calculation  for  SHEAR7  fatigue   SHEAR7  fatigue  is  rather  different  from  the  other  fatigue  methods  because  damage  is  calculated  external  to  OrcaFlex   by   SHEAR7.   OrcaFlex   merely   provides   a   means   to   collate,   sum   and   plot   the   damage   from   a   number   of   different   SHEAR7  load  cases  in  a  convenient  manner.   The  SHEAR7  damage   calculation   is   based   on   the  damage  rate  output   in   the   SHEAR7   .plt   file.  This  value   is   multipled   by  the  load  case  exposure  time  to  produce  the  damage  associated  with  that  load  case.  

403  

w

 

VIV  Toolbox,  Frequency  Domain  Models  

 

9

VIV  TOOLBOX  

The  OrcaFlex  VIV  Toolbox  provides  analysis  of  vortex  induced  vibration  (VIV)  of  lines.  It  offers  a  choice  of  various   alternative  ways  of  modelling  VIV,  including  both  frequency  and  time  domain  approaches,  and  has  been  developed   in  co-­‐operation  with  academics  in  the  UK  and  USA.  The  VIV  Toolbox  is  included  as  standard  in  OrcaFlex.   A  separate  user  guide  for  the  time  domain  VIV  models  is  available  ( Time  Domain  VIV  Models.pdf).   Different  VIV  Models   The  VIV  Toolbox  provides  facilities  for  using  the  following  different  VIV  models:   x

VIVA.  The  VIV  Toolbox  provides  a  fully  integrated  link  to  VIVA.  OrcaFlex  automatically  prepares  the  VIVA  data   from   the   OrcaFlex   data,   calls   VIVA   and   presents   the   results.   To   use   this   you   will   need   a   copy   of   VIVA,   release   2.0.6  or  later.  

x

SHEAR7.  The  VIV  Toolbox  provides  facilities  for  exporting  SHEAR7  structural  data  files  and  SHEAR7  .mds  files   based  on  an  OrcaFlex  model.  SHEAR7  can  then  be  run  manually  using  these  files  as  input,  or  called  directly  from   OrcaFlex.  

x

Two  wake  oscillator  models,  the  Milan  model  and  the  Iwan  and  Blevins  model.  

x

Two  vortex  tracking  models.  

Of   all   these   models,   VIVA   and   SHEAR7   are   the   two   main   programs   in   current   use   in   the   industry.   They   are   both   independent   non-­‐Orcina   programs   written   and   distributed   by   other   companies,   so   to   use   them   you   need   to   purchase  and  install  them  on  your  machine.  They  are  both  frequency  domain  models,  so  they  only  analyse  steady   state  conditions.   The   other   models  are  included   in   the  VIV   Toolbox   within  OrcaFlex,  so   no   further   software  is   needed.  They  are  all   time-­‐domain   models,   so   they   can   analyse   non-­‐steady-­‐state   conditions.   They   do   not   yet   have   a   track   record   in   the   industry.   Using  VIV  Models   You  choose  which  VIV  model  to  use  (if  any)  on  the  VIV  page  on  the  line  data  form.  There  are  separate  choices  for  the   static  and  dynamic  analyses  and  so  you  do  not  have  to  use  the  same  model  for  the  two  analyses.  The  time-­‐domain   models  are  only  applicable  to  the  dynamic  analysis.  

9.1

FREQUENCY  DOMAIN  MODELS  

9.1.1

VIVA  

VIVA  is  an  independent  frequency   domain  program  developed  by  Prof.  M   Triantafyllou  at  MIT  and   distributed   by  JD   Marine  (formerly  DTCEL).   Orcina  is  a  participating  member  of  a  Joint  Industry  Project  run  by  DTCEL  and  Prof.  Triantafyllou  of  MIT.  The  project   began  mid-­‐year  2000.  Phase  1  was  completed  mid-­‐year  2001  and  a  second  phase  is  in  progress.   As   part   of  the   project,   a   software   link   has   been   written   between   OrcaFlex   and   VIVA.   The   software   link   allows   a   riser   model   to   be   built   in   OrcaFlex   taking   advantage   of   the   existing   interactive   user   interface.   OrcaFlex   then   calls  VIVA   which   carries   out   a   frequency   domain   calculation   of  VIV   response   based   on   an   energy   balance.   Results   are   reported   in   spreadsheet   form   for   each   single   excited   mode   of   vibration.   A   combined   multi-­‐mode   response   is   also   reported.   Results   include   amplitudes   of   vibration   and   mean   effective   drag   coefficients.   The   drag   coefficients   can   be   automatically  used  by  OrcaFlex  in  further  static  and  dynamic  analysis.   The   VIVA   calculation   can   be   repeated   at   intervals   through   an   OrcaFlex   analysis   to   update   the   drag   coefficients.   OrcaFlex  runs  as  usual  in  the  time  domain,  but  makes  a  call  periodically  to  VIVA  which  repeats  the  VIV  calculation   for   the   instantaneous   riser   configuration.   The   drag   coefficient   distribution   is   then   read   back   into   OrcaFlex   which   continues  with  the  analysis.   OrcaFlex  VIVA  Interface   The  OrcaFlex  interface  to  VIVA  requires  VIVA  to  be  installed  on  your  machine  and  available  in  the  form  of  a  dynamic   link  library  (DLL)  called  v3ddll.dll.  This  DLL  file,  and  the  VIVA  database  files  that  it  uses,  must  be  located  in  the  same   directory  as  the  OrcaFlex.exe  program  file.  

405  

w

 

VIV  Toolbox,  Frequency  Domain  Models    

The   current   version   of   OrcaFlex   is  compatible,   and   has   been   tested   with,   the   VIVA  DLL   version   2.0.6.   Later   versions   of  the  DLL  can  be  used  Ȃ  OrcaFlex  checks  for  interface  compatibility  before  calling  the  DLL,  and  reports  an  error  if   the  DLL  found  is  not  compatible.   To  use  VIVA  for  VIV  analysis  of  a  line,  set  the  line's   Statics  VIV  or  Dynamics  VIV  data  item  (or  both)  to  VIVA.  Then  set   up  the  VIVA  data  that  appears.   VIVA  will  then  be  called   when  you   do  the  OrcaFlex  static   or  dynamic  analysis,  and  the  drag  coefficients  predicted   by   VIVA  (for  the  specified  Target  Mode)  will  be  used  instead  of  the  user-­‐specified  drag  coefficients  from  the  line  type   form.   These   VIVA   drag   coefficients,   and   other   results   calculated   by   VIVA,   are   available   in   OrcaFlex   Ȃ   see   VIVA   Results.   VIVA  in  Static  Analysis  

When   Statics   VIV   is   set   to   VIVA,   OrcaFlex   calculates   the   static   position   of   the   line   using   the   drag   coefficients   calculated   by  VIVA.   A   fully   coupled   statics   calculation   is   performed   since   VIVA's   analysis   depends   on   the   position   of   the  line  and  vice  versa.  The  coupled  static  analysis  is  done  as  follows:   1.

First  OrcaFlex  sets  the  line's  drag  coefficients  to  those  specified  on  the  OrcaFlex   line  type  form.  

2.

OrcaFlex  then  calculates  the  static  position  of  the  line  using  those  drag  coefficients.  

3.

OrcaFlex   then   calls   VIVA   to   analyse   VIV   for   that   position   and   the   line   drag   coefficients   are   updated   to   those   calculated  by  VIVA  for  the  specified  Target  Mode.  

4.

OrcaFlex  then  recalculates  the  static  position  of  the  line  for  those  new  drag  coefficients.  

5.

Steps  3  and  4  are  then  repeated  until  the  static  position  has  converged.  

The   iteration   is   deemed   to   have   converged   when   none   of   the   nodes   has   changed   position   (compared   with   the   previous   iteration)   by   more   than   Convergence   Tolerance   *   NodeLength,   where   NodeLength   is   the   length   of   line   represented  by  that  node.  The  method  usually  requires  only  2  or  3  coupling  iterations  to  converge.   VIVA  in  Dynamic  Analysis  

When   Dynamics   VIV   is   set   to   VIVA,   OrcaFlex   calls   VIVA   at   regular   intervals,   as   specified   by   the   Dynamics   Time   Interval.   After   each   call   the   drag   coefficients   of   the   line   are   updated   to   those   calculated   by   VIVA   for   the   specified   Target  Mode.   Note   that   VIVA   is   a   frequency   domain   program   and   so   can   only   handle   steady   state   conditions.   It   is   therefore   unrealistic   to   use   VIVA   in   dynamic   analysis   unless   the   conditions   change   only   slowly   compared   to   the   Strouhal   period.  See  also  Use  Relative  Fluid  Velocity  and  Include  Wave  in  Fluid  Velocity.   Limitations   When  using  the  OrcaFlex  interface  to  VIVA  please  note  the  following  limitations:   x

The  VIVA  analysis  involves  a  modal  analysis  of  the  line.  Currently  this  modal  analysis  is  done  by  VIVA  but  this   has  limitations  Ȃ  see  VIVA  modal  analysis  limitations.  

x

There  are  difficulties  in  passing  to  VIVA  details  of  attachments  to  the  line  (e.g.  clumps,  buoys,  links,  winches),  so   attachment  are  ignored  by  the  VIVA  analysis.  

x

The  VIVA  DLL  is  not  currently  capable  of  handling  multiple  simultaneous  use,  so  you  should  not  use  the  VIVA   interface  from  more  than  one  copy  of  OrcaFlex  at  a  time  running  on  a  single  machine.  

VIVA  Data   The   VIVA   data   is   the   data   that   VIVA   needs   and   which   OrcaFlex   cannot   deduce   automatically   from   the   ordinary   OrcaFlex   data.   It   is   described   below   and   appears   on   the   VIV   page   on   the   line   data   form   when   you   select   VIVA   for   Statics  VIV  or  Dynamics  VIV.   When  using  VIVA  please  note  the  following:   x

You  should  set  up  the  line  with  End  A  at  the  top  and  End  B  at  the  bottom.  

x

All  the  sections  in  the  line  must  have  the  same  segment  length.  This  is  a  requirement  of  VIVA.  

x

If   the   line   touches   down   on   the   seabed   then   OrcaFlex   passes   only   the   suspended   part   of   the   line   to   VIVA   for   analysis.  The  rest  of  the  line  is  not  passed  to  VIVA.  By  'suspended  part'  here  is  meant  the  part  between  End  A  

406  

w

 

VIV  Toolbox,  Frequency  Domain  Models  

 

and   the   first   node   in   contact   with   the   seabed.   Note   that   if   there   are   multiple   touchdowns   then   the   VIVA   analysis   is  only  applied  to  the  part  of  the  line  up  to  the  first  touchdown  point.   x

When   there   is   seabed   contact   OrcaFlex   tells   VIVA   to   treat   the   touchdown   point   as   pinned.   This   is   as   in   the   touchdown  example  in  the  VIVA  documentation.  

x

You   need   to   use   enough   segments   in   the   suspended   part   of   the   line.   There   is   a   lower   limit   in   VIVA   of   100   segments,  and  if  the  line  has  fewer  than  this  then  VIVA  reports  an  error  and  no  VIV  calculation  is  done.  If  the   line   has   100   or   more   segments   then   VIVA   checks   whether   there   are   enough   segments   to   reasonably   model   VIV;   and  if  not  then  VIVA  gives  a  warning  but  the  calculation  continues.  For  details  see  the  VIVA  documentation.  

x

When   modelling   Vetco   type   risers   (i.e.   with   auxiliary   pipes   attached),   you   need   to   set   the   OrcaFlex   line   end   orientation  to  match  the  orientation  of  the  auxiliary  pipes.  For  details  see  Modelling  Vetco  Risers.  

x

OrcaFlex  passes  VIVA  the  fluid  density  and  viscosity  at  each  node.  For  fully  submerged  nodes  OrcaFlex  passes   the  sea  density  and  kinematic  viscosity  at  that  point.  For   a  node  out   of  the  water  OrcaFlex  passes  the  air  density   specified   in   the   OrcaFlex   data   and   a   hard-­‐coded   viscosity   value   of   1.5e-­‐5  m2/s   (Source:  Batchelor,  page   594,   air   at   20°C).   For   a   partially   submerged   node   OrcaFlex   interpolates   between   the   two,   based   on   the   node's   Proportion  Wet.  

Whole  Line  Properties   Target  Mode  

VIVA  predicts   which  modes   of  oscillation  might   be   excited  and  it  gives  separate  VIV  results  for   each  possible   excited   mode.  In  addition  it  gives  VIV  results  for  'multi-­‐mode'  response,  i.e.  response  that  is  a  mixture  of  the  possible  modes.   You  must  specify  which  set  of  VIVA  Cd  results  to  use,  based  on  the  type  of  response  you  expect,  by  setting  the  Target   Mode   data   item.   The   Target  Mode   can   be   set   to   a   mode  number,   meaning   use   the   single   mode   results   for   that   mode   number.  Or  it  can  be  set  to  '~',  meaning  use  the  multi-­‐mode  results.   If   you   set   the   Target   Mode   to   a   mode  number   that   VIVA  does   not   predict   will   be   excited   then  OrcaFlex   will   report   a   warning  and  use  the  line  type  Cd  values.   Dynamics  Time  Interval  

This  data  item  only  applies  when  Dynamics  VIV  is  set  to  VIVA.  It  specifies  how  often  VIVA  will  be  called  during  the   OrcaFlex  simulation.   VIVA  will  be  called  after  every  T  seconds  of  simulation,  where  T  is  the  specified  dynamics  time  interval.  After  each   call   to   VIVA,   the   drag   coefficients   used   by   the   line   will   be   updated   to   those   calculated   by   VIVA   for   the   specified   Target  Mode.   Convergence  Tolerance  

This  data  item  only  applies  when   Statics  VIV  is  set  to  VIVA.  It  specifies  the  tolerance  used  in  the  statics  iteration.  For   details  see  VIVA  in  Static  Analysis.   Transverse  Damping  Ratio  

The   structural   damping   ratio   that   VIVA  should   use.   It   is  a   single   value   that   applies   to   the   whole   line.   For  details  see   the  VIVA  documentation,  where  it  is  referred  to  as  ZETA.   VIVA  Z  Axis  Direction  

This  data  item  specifies  the  direction  of  the  VIVA  global  Z-­‐axis.   VIVA   uses   its   own   global,   right-­‐handed   frame   of   reference,   in   which   the   X-­‐axis   is   vertically   upwards   (i.e.   in   the   OrcaFlex  global  Z-­‐axis  direction)  and  the  Y  and  Z  axes  are  horizontal.   VIVA   recommends   that   the   VIVA   Z-­‐axis   is   chosen   to   be   in   the   downstream   flow   direction,   if   that   is  well-­‐defined.   For   constant  uniform  flow  there  is  a  unique  flow  direction,  so  it  is  well-­‐defined.  If  the  flow  direction  varies  along  the  line   then   there   is  no  unique   flow  direction.  In  this  case  it  is  probably  best  to  set  the  VIVA   Z -­‐axis  to  be  in  the  average   flow   direction.   A  value  of  '~'  is  interpreted  to  mean  the  current  direction  at  the  sea  surface.  

407  

w

 

VIV  Toolbox,  Frequency  Domain  Models    

Use  Relative  Fluid  Velocity,  Include  Wave  in  Fluid  Velocity  

VIVA   needs   to  know  the   flow   velocity   at   each   point   along   the   line.   You   can   choose   to   either   use   the   relative   velocity,   including   the  velocity  of  the  line,  or   else  use  only  the  fluid  velocity,  ignoring  any  velocity  of  the  line.   You  can   also   choose  whether  to  include  any  wave  contribution  to  the  fluid  velocity.   Ideally  the  relative  velocity  should  be  used  and  the  wave  should  be  included,  since  in  reality  it  is  the  total  relative   velocity   that   generates   vortices.   However  VIVA  is  a   frequency  domain  program   and  so   can   only  handle   steady   state   conditions,  whereas   OrcaFlex   can   handle   time-­‐varying   conditions.  VIVA  effectively  assumes   that   the   flow   velocity   it   is   given   is   constant   for   long   enough   for   VIV   to   settle.   It   therefore   cannot   correctly   analyse   cases   where   the   flow   velocity  includes  time-­‐varying  components  whose  periods  are  comparable   with  or  shorter   than  the   Strouhal  period.   (The  same  applies  to  SHEAR7,  since  it  too  uses  a  frequency  domain  analysis.)   We  therefore  provide  these   two  switches,  to  allow  you  to  control  whether  line  motion  and  wave  effects  are  included   in  the  flow  velocity  given  to  VIVA.  If  the  line   motion   or   wave  include  significant  velocity  components  that  are   not   'slow'  compared  to  the  expected  VIV  period,  then  you  might  want  exclude  the  line  motion  or  wave.  This  is   still  not   ideal,  since  their  effects  are  then  ignored.  The  alternative  is  to  use  a  time-­‐domain  model,  such  as  a  wake  oscillator   model  or  the  vortex  tracking  models.   Section-­‐Specific  Properties   For  each  section  of  the  line,  you  must  specify  the  VIVA  section  type  and  its  properties.  The  OrcaFlex   line  type  and   length  of  the  sections  are  displayed  for  information,  but  they  are  not  editable  (to  edit  them  see  the  Structure  page  on   the  OrcaFlex  line  data  form).   VIV  Diameter   The  VIV  diameter  specifies  the  diameter  used  by  the  VIV  model.  Separate  values  can  be  specified  for  each  section.   The   value   specified   is   used   for   all   nodes   in   that   section.   For   a   node   at   the   intersection   of   two   sections   the   VIV   diameter  of  the  following  section  is  used.  The  VIV  Diameter  can  be  set  to  '~',  which  is  taken  to  mean  'same  as  the   section  outer  diameter'.   Section  Type  

The  VIVA  section  type  can  be  one  of:   x

Smooth:  No  further  VIVA  section  data  is  needed.  

x

Strake:  You  must  also  specify  the  height  to  diameter  ratio  for  the  strakes.  

x

Fairing:  You  must  also  specify  the  chord  length  and  thickness  of  the  fairing.  

x

Vetco:   No   further  VIVA  section   data   is  needed,  but  the  line  orientation  must   match  the   orientation   of  the  Vetco   riser.  See  Modelling  Vetco  Risers  below.  

For  details  of  this  VIVA  data  see  the  VIVA  documentation.   Modelling  Vetco  Risers   Vetco   risers   have   auxiliary   pipes   attached   and   VIVA   needs   to   know   the   direction   of   the   flow,   relative   to   the   orientation  of  the  auxiliary  pipes.  The  diagram  below  shows  a  section  through  the  line,  looking  towards  End  B.  VIVA   needs  to  know  the  angle  Alpha.   OrcaFlex   calculates   this   angle   automatically   by   assuming   that   the   auxiliary   pipes   are   oriented,   relative   to   the   OrcaFlex  local  x  and  y-­‐direction,  as  shown  in  the  diagram.  You  must  therefore  set  up  the  line   end  orientation  data  so   that   the   line's   local   x   and   y-­‐directions   are   oriented   as   shown.   In   other   words   you   should   set   up   the   line   end   orientation  data  so  that  the  line  local  x-­‐axis  points  through  the  larger  of  the  two  gaps  between  auxiliary  pipes.  

408  

w

 

VIV  Toolbox,  Frequency  Domain  Models  

 

Flow Direction Alpha x

y  

VIVA  Results   The  VIVA  results  are  presented  in  OrcaFlex  as  extra  worksheets  in  the   Full  Results  tables.  The  worksheets  give  the   results  from  the  latest  VIVA  call.   There   is   a   separate   worksheet   for   each   excited   mode,   plus   an   extra   worksheet   for   the   multi-­‐mode   response.   The   drag  coefficients  currently  in  use  are  those  corresponding  to  the  specified   Target  Mode.   The  VIVA  results  are  described  briefly  below.  For  details  see  the  VIVA  documentation.   Note:  

Amplitudes  in  VIVA  results  are  single  amplitudes,  i.e.  measured  from  the  mean  position  to  the  peak.  

Single  Mode  Results   The   worksheet   for   a   given   single   excited   mode   gives   a   table   of  'Single   Mode'   results   as   a   function   of  arc   length   along   the  line.  These  are  results  predicted  by  VIVA  if  single-­‐mode  VIV  response  occurs  in  that  mode.  They  include:   Max  Amplitude  

The  amplitude  of  predicted  VIV  motion  at  that  point.   Max  Bend  Moment,  Max  Stress  

The  dynamic  bend  moment  and  stress  amplitudes  due  to  VIV.  Note  that  they  do   not  include  the  bend  moment  and   stress  due  to  the  mean  position  of  the  line.   Drag  Coefficient  

The  predicted  drag  coefficient,  allowing  for  VIV.   Multi-­‐Mode  Results   The   'Multi-­‐Mode'   worksheet   gives   two   tables.   The   second   table   is   the   multi-­‐mode   equivalent   of   the   single   mode   results  table;  it  gives  the  same  information,  but  this  time  for  the  case  where  multi-­‐mode  VIV  response  occurs.   The  first  table  summarises  the  bending  modes.  It  has  the  following  columns:   Mode  

This   column   shows   which   modes   have   been   analysed.   Those   that   VIVA   calculates   as   possibly   being   excited   are   marked  with  an  asterisk.   Natural  Mode  Frequency  With  and  Without  VIV  

These  columns  give  the  natural  frequencies  of  the  mode,  as  calculated  by  VIVA.  Those  'Without  VIV'  are  calculated   using   VIVA's   own   default   added   mass   coefficients   (not   those   specified   in   the   OrcaFlex   line   type   data).   Whereas   in   those  'With  VIV'  the  VIVA  default  added  mass  coefficients  have  been  modified  to  allow  for  VIV.  

409  

w

 

VIV  Toolbox,  Frequency  Domain  Models    

Warning:  

VIVA's  modal  analysis  calculates  bending  modes  only  and  does  not  allow  for  the  mean  curvature.   This   omission   is   equivalent   to   VIVA   calculating   the   out-­‐of-­‐plane   modes,   since   there   is   no   mean   curvature   component   in   the   out-­‐of-­‐plane   direction.   The   natural   frequencies   in   the   VIVA   results   therefore   correspond   to   the   out-­‐of-­‐plane   bending   modes   predicted   by   the   OrcaFlex   modal  analysis.   These   are   correct   modes   to   use   for   in-­‐plane   flow   (since   VIV   will   then   be   in   the   out   of   plane   direction),  but  they  are  not  the  right  modes  to  use  for  out-­‐of-­‐plane  flow,  since  the  VIV  will  then  be   in-­‐plane.   The   difference   between   the   in-­‐plane   and   out-­‐of-­‐plane   bending   modes   depends   on   the   magnitude  of  the  curvature  and  the  wave  length  of  the  mode.  

Max  Amplitude  

The  largest  offset  of  any  node  in  this  mode.  

9.1.2

SHEAR7  

SHEAR7  is  an  independent  program,  developed  by  Prof.  K  Vandiver  at  MIT  and  distributed  by  SBM  Atlantia.  The  link   to   SHEAR7   is   based   around   OrcaFlex's   ability   to   export   SHEAR7   data   files   (.dat)   and   SHEAR7   mode   shape   files   (.mds).   Note:  

SHEAR7  must  be  licensed  separately  to  OrcaFlex.  Please  contact  SBM  Atlantia  for  details.  

When   exporting   a   SHEAR7   data   file   the   program   assumes   that   you   will   also   be   u sing   a   mode   shape   file   produced   by   OrcaFlex.  The  procedure  for  linking  SHEAR7  to  your  OrcaFlex  model  is  as  follows:   1.

Build  your  OrcaFlex  model  as  normal.  SHEAR7  analyses  VIV  for  one  or  more  of  your  OrcaFlex   Lines.  

2.

Input  the  appropriate  values  on  the  SHEAR7  data  form.  

3.

Calculate  the  static  position  of  the  model.  

4.

Export  a  SHEAR7  data  file  and  a  SHEAR7  Mds  file.  

5.

These  files  can  now  be  used  as  inputs  to  SHEAR7.  

Direct  SHEAR7  interface   As  an  alternative  to  exporting  files  and  running   SHEAR7   manually,  as  described  above,  OrcaFlex  is  capable  of  calling   SHEAR7  directly.  This  capability  is  enabled  by  selecting  the  SHEAR7  statics  VIV  option  on  the  VIV  page  of  the  Line   data  form.   SHEAR7  executable  file  location,  SHEAR7  lift  file  location  

The  direct  SHEAR7  interface  operates   by   executing  the   SHEAR7  executable   file.  In  order  to  do  this  OrcaFlex  must  be   told  where  to  locate  the  executable  file.  This  data  is  specified  on  the  SHEAR7  data  form  and  is  only  available  when   SHEAR7  is  selected  for  the  statics  VIV  option  on  the  VIV  page  of  the  Line  data  form.   In  addition   SHEAR7  requires  a  file  to  specify  lift  coefficients,  usually  called  "common.cl".  Typically  the  lift  file  will  be   the   standard   one   supplied   with   SHEAR7.   If   the   lift   file   location   is   left   blank   OrcaFlex   attempts   to   use   a   file   called   "common.cl"  in  the  same  directory  as  the  executable  file.   These   file   locations   can   be   specified   as   either   full   paths   or   as   paths  relative   to   the   directory   containing   the   OrcaFlex   file.   These   data   are   model-­‐wide   properties.   A   consequence   of   this   is   that   if   you   wish   to   modify   the   values   using   batch   script  then  you  need  to  select  the  General  object  rather  than  a  Line  object.   SHEAR7  coupling  method  

The   SHEAR7   direct   interface   performs   a   coupled   statics   calculation   since   the   SHEAR7   analysis   depends   on   the   position  of  the  line  and  vice  versa.  The  coupled  static  analysis  is  done  as  follows:   1.

First  OrcaFlex  sets  the  line's  drag  coefficients  to  those  specified  on  the  OrcaFlex   line  type  form.  

2.

OrcaFlex  then  calculates  the  static  position  of  the  line  using  those  drag  coefficients.  

3.

OrcaFlex  then  calls  SHEAR7  to  analyse  VIV  for  that  position  and  the  line  drag  coefficients  are  updated.  

4.

OrcaFlex  then  recalculates  the  static  position  of  the  line  for  those  new  drag  coefficients.  

5.

Steps  3  and  4  are  then  repeated  until  the  static  position  has  converged.  

410  

w

 

VIV  Toolbox,  Frequency  Domain  Models  

 

The   iteration   is   deemed   to   have   converged   when   none   of   the   nodes   has   changed   position   (compared   with   the   previous   iteration)   by   more   than   Convergence   Tolerance   *   NodeLength,   where   NodeLength   is   the   length   of   line   represented  by  that  node.  The  method  usually  requires  only  2  or  3  coupling  iterations  to  converge.   The  direct  SHEAR7  interface  offers  a  variety  of  coupling  options  which  differ  in  how  they  handle  the   SHEAR7  .mds   file.   x

The   full   coupling   option   generates   a   new   .mds   file   at   each   coupling   iteration,   that   is   every   time   SHEAR7   is   called.  

x

The   partial,   automatic   .mds   file   coupling   option   creates   an   .mds   file   on   the   first   coupling   iteration   which   is   then  used  in  all  subsequent  coupling  iterations.  

x

The   partial,   user   .mds   file   coupling   option   uses   an   .mds   file   specified   on   the   data   form   for   all   coupling   iterations.  

The  fully  coupled  approach  does  have  the  disadvantage  that  generating  an  .mds  file  at  each  coupling  iteration  can  be   time   consuming,   especially   for   more   complex   models,   and   so   the   partially   coupled   approach   can   be   significantly   faster.   The   partially  coupled  approaches  are  more  akin  to  running   SHEAR7  manually.  However,  the  ultimate  static  solution   will   differ   because   it   does   include   the   effect   of  drag   enhancement.   When   running   SHEAR7   manually   the   only  way   to   account   for   drag   enhancement   is   to   manually   modify   drag   coefficients   in   the   OrcaFlex   model   which   is   extremely   impractical.   First  and  last  modes  

These  data  items  specify  which  modes  are  to  be  included  in  the  .mds  file.  These  data  are   not  required  if  you  select  the   partial,  user  .mds  file  coupling  option.   A  value  of  '~'  for  the  first  mode  is  interpreted  as  the  lowest  numbered  transverse  mode.  A  value  of  '~'  for  the  last   mode   is   interpreted   as   the   highest   numbered   transverse   mode.   By   transverse   we   mean   that   the   modal   analysis   classifies  the  mode  as  either  Transverse  or  Mostly  Transverse.   These   data   items   are   interpreted   in   exactly   the   same   way   as   the   FirstMode   and   LastMode   parameters   to   the   SHEAR7MdsFile  batch  script  command.   SHEAR7  .mds  file  

The  location  of  the  .mds  file  to  be  passed  to  SHEAR7.  This  data  is   only  required  if  you  select  the  partial,  user  .mds   file  coupling  option.   The  location  can  be  specified  as  either  a  full  path  or  as  a  path   relative  to  the  directory  containing  the  OrcaFlex  file.   The   .mds   file   must   be   generated   separately   which   is   normally   done   either   form   the   modal   analysis   form   or   using   batch  script.   Although  SHEAR7  requires  .mds  files  to  be  named  "common.mds"  that  restriction  does  not  apply  to  OrcaFlex's  direct   interface   to   SHEAR7.   This   is   a   significant   benefit   of   the   direct   interface   to   SHEAR7   because   it   allows   you   to   have   multiple  .mds  files  in  the  same  directory.   Results  output  

When   OrcaFlex   calls   SHEAR7   it  captures   the   SHEAR7   output   file,   .out   and   .plt.   These   output   files   are   included   in   the   Full  Results  tables  for  the  Line.   These  output  files  can  also  be   exported  from  the  SHEAR7  data  form  and  by  using   the   SHEAR7OutFile  and  SHEAR7PltFile  batch  script  commands.   The   .out   and   .plt   files   are   automatically  exported   if   you   run   the   static   analysis  in  batch  mode.   The   files   are   exported   with  file  names  based  on  the  model  file  name  and  the  line  name.   The  enhanced  drag  coefficients  can  be  obtained  from  a  static  state  range  graph  of  x-­‐Drag  Coefficient  (or  indeed  y-­‐ Drag  Coefficient  which  is  identical).   Technical  details  

The  direct  SHEAR7  interface  operates  by  calling  the  SHEAR7  executable  directly  as  follows:   1.

A  temporary  directory  is  created  to  contain  the  SHEAR7  input  and  output  files.  

2.

A  SHEAR7  .dat  file  is  generated  in  the  temporary  directory.  

411  

w

 

VIV  Toolbox,  Frequency  Domain  Models    

3.

The   SHEAR7   .mds   file   is   created   in   the   temporary   directory   and   called   "Common.mds".   How   this   .mds   file   is   created  is  determined  by  the  chosen  coupling  option  as  described  above.  

4.

The  specified  lift  file  is  copied  to  the  temporary  directory  and  renamed  as  "Common.cl".  

5.

The  SHEAR7  executable  is  run  and  the  SHEAR7  .out  and  .plt  files  are  read  by  OrcaFlex.  

6.

The  .plt  file  is  parsed  by  OrcaFlex  to  find  the  drag  enhancements  factors  C f.  

This  process  depends  very  heavily  on  the  current  implementation  details  of  SHEAR7.  We  have  developed  and  tested   the  SHEAR7  direct  interface   with  SHEAR7  versions   4.4  and  4.5.  We  cannot  guarantee  that  other  versions  of  SHEAR7   will  be  compatible  with  the  SHEAR7  direct  interface.   OrcaFlex  uses  linear  interpolation  of  the  the  C f  values   if  the  locations   read  from   the  .plt  file   do   not  correspond  to  the   node   locations   in   the   OrcaFlex   model.   This   can   occur   if   the   discretisation   of   a   user-­‐supplied   .mds   file   differs   from   that  of  the  OrcaFlex  model.  

SHEAR7  data  file   To  export  a  SHEAR7  data  files  you  must  first  provide  extra  data  that  SHEAR7  requires  but  which  is  not  needed  by   OrcaFlex.  This  is   done  from  the  SHEAR7  data  form  which  can  be  opened  from  the   Model  Browser.  Note  that  your   model  must  include  at  least  one  Line  for  the  SHEAR7  data  form  to  be  available.   When  you  have  input  all  the  necessary  data  on  the  SHEAR7  data  form  you  create  the  SHEAR7  data  file  by  clicking  on   the  Export  SHEAR7  Data  File  button.  The  model  must  be  in   Statics  Complete  state  when  you  export  the  SHEAR7   data  file  but  please  note  that  data  on  the  SHEAR7  data  form  is  editable  while  in  this  state.   A   more   convenient   way   to   work   is   to   use   the   direct   SHEAR7   interface   which   can   automatically   save   the   required   SHEAR7  input  files,  run  SHEAR7,  extract  drag  amplification  factors  and  perform  a  coupled  static  analysis.   Include  SHEAR7  data  in  text  data  file  

By  default  the  SHEAR7  data  is  not  included  in   text  data  files  saved  by  OrcaFlex.  You  can  check  this  option  to  override   that  default  and  ensure  that  SHEAR7  data  is  included  in  text  data  files.   If   you   are   using   the  direct   SHEAR7   interface  then   the  SHEAR7   data   is  always   included   in   text  data   files  irrespective   of  this  setting.   Exporting  from  Batch  Script  

The  SHEAR7  data  file  can  be  exported  from  an  OrcaFlex  batch  script  using  the  SHEAR7DataFile  command.   SHEAR7  data   The  majority  of  the  information  in  a  SHEAR7  data  file  can  be  derived  from  the  data  for  an  OrcaFlex  Line.  However,   there   are   a   number   of   other   values   which  SHEAR7   needs   and   these   are  documented   below.   Mostly  these   values  are   given   the   same   name   as   used   by   SHEAR7.   For   full   details   on   how   they   are   used   please   refer   to   the   SHEAR7   documentation.   Line  

Specifies  the  OrcaFlex  Line  to  be  described  in  the  SHEAR7  data  file.   SHEAR7  File  Version  

SHEAR7  changed  the  format  of  its  data  file  with  the  release  of  SHEAR7  version  4.5.  OrcaFlex  can  output  data  files  for   either  version  4.3/4.4  or  version  4.5,  as  specified  by  this  data.   Output  summary  Locations  

These  data  specify  the  locations  at  which  a  summary  of  the  response  is  given  in  the  SHEAR7  .out  file.  A  value  of   '~'   can  be  used  for  the   Arc  Length  Range  Start  to  mean  'End  A'.  Likewise,  a  value  of   '~'  for   Arc  Length  Range  End   means  'End  B'.   Reduced  Velocity  Bandwidths  

For  SHEAR7  version  4.3/4.4  you  specify  the  single-­‐mode  and  multi-­‐mode  reduced  velocity  bandwidth.   For  SHEAR7  version  4.5  you  specify  just  a  single  value  for  reduced  velocity  bandwidth.   Structural  Damping  Ratio  

Corresponds  to  the  SHEAR7  data  item  of  the  same  name.  

412  

w

 

VIV  Toolbox,  Frequency  Domain  Models  

 

Primary  zone  amplitude  limit  (SHEAR7  version  4.5  only)  

Corresponds  to  the  SHEAR7  data  item  of  the  same  name.   Mode  Behaviour  Cut  Off  

Corresponds  to  the  SHEAR7  data  item  called  "Cutoff  level".   Number  of  Lift  Coefficient  Tables  (SHEAR7  version  4.3/4.4  only)  

This  is  the  number  of  tables  you  want  SHEAR7  to  read  from  the  common.cl  file.   Current  Profile  

The  discretisation  option  allows  you  to  control  how  the  current  profile  is  output.   You  can  specify  that  the  current   profile  contains  one  entry  for  each  node  in  the  line.  However  this  can  give  rise  to  numerical  problems  in  SHEAR7,   especially   for   finely   segmented   models,   and   so   it   is   preferable   to   discretise   the   current   more   coarsely   using   the   regular  spacing  option.   You  also  specify  the  probability  of  occurrence  (a  number  between  0  and  1)  of  the  flow  profile  and  the  flow  profile  ID.   Young's  Modulus  

Young's  modulus  for  the  strength  member.  SHEAR7  uses  this  for  computing  stress  and  damage  rate.   Power  Ratio  exponent  (SHEAR7  version  4.5  only)  

Corresponds  to  the  SHEAR7  data  item  of  the  same  name.   S-­‐N  Curve  

Specifies  the  S-­‐N  curve  to  be  used  by  SHEAR7  for  its  fatigue  calculations.  An  endurance  limit  can  be  specified  Ȃ  this  is   called  the  "cutoff  stress  range"  in  the  SHEAR7  documentation.   Stress  Concentration  Factors  

Specify  a  global  stress  concentration  factor  for  the  line  and  optionally  a  number  of  local  stress  concentration  factors.   Section  Data   The   following   data   is   specified   on   a   section   by   section   basis.   That   is   different   values   can   be   specified   for   each   OrcaFlex   line   section.   The   SHEAR7   terminology   for   this   is   sectional   zones.   In   the   SHEAR7   data   file   that   OrcaFlex   produces  there  is  a  one-­‐to-­‐one  correspondence  between  OrcaFlex  line  sections  and  SHEAR7  sectional  zones.   VIV  Diameter   The  VIV  diameter  specifies  the  diameter  used  by  the  VIV  model.  Separate  values  can  be  specified  for  each  section.   The   value   specified   is   used   for   all   nodes   in   that   section.   For   a   node   at   the   intersection   of   two   sections   the   VIV   diameter  of  the  following  section  is  used.  The  VIV  Diameter  can  be   set  to  '~',  which  is  taken  to  mean  'same  as  the   section  outer  diameter'.   Strouhal  Type,  Strouhal  Number  

The   Strouhal   type   and   number   defines   the   relationship   of   flow   velocity   and   cylinder   diameter   to   the   local   vortex   shedding  frequency.  Strouhal  Type  can  be   Rough  Cylinder  or  User  Specified.  Rough  Cylinder  corresponds  to  the   SHEAR7  Strouhal  code  200.  If  User  Specified  is  chosen  then  you  must  also  specify  a  Strouhal  Number.   Lift  Coefficient  Table,  Lift  Coefficient  Factor  

Lift  Coefficient  Table  specifies  which  table  is  used  from  the  common.cl  file.   Lift  Coefficient  Factor  corresponds  to  the  SHEAR7  data  item  called  Lift  Coefficient  Reduction  Factor.   Reduced  Velocity  Damping  Coefficients  

Damping  Coefficients  for  still  water,  low  and  high  reduced  velocity  are  specified.   SHEAR7  structural  zones   When   OrcaFlex   generates   the   SHEAR7   data   file   it   assumes   a   one-­‐to-­‐one   mapping   between   OrcaFlex   sections   and   SHEAR7   structural   zones.   This   has   some   implications   for   how   you   create   your   OrcaFlex   model.   Essentially,   anywhere   on   your   line   where   you   need   SHEAR7   structural   zone   data   to   change   you   must   ensure   that   there   is   an   OrcaFlex  section  boundary.  

413  

w

 

VIV  Toolbox,  Frequency  Domain  Models    

SHEAR7  structural  zone  data  comprises  the  following:   x

Hydrodynamic  and  strength  diameters.  

x

Added  mass  coefficient,  lift  coefficient  and  Strouhal  number.  

x

Area  moment  of  inertia,  mass  per  unit  length  and  tension  variation.  

x

Hydrodynamic  damping  coefficients.  

Since   most   of   these   properties   are   constant   within   an   OrcaFlex   section   you   will   naturally   introduce   section   boundaries  at  the  locations  where  they  change.  However,  there  are  a  few  exceptions.   Lift   coefficient,   Strouhal   number   and   hydrodynamic   damping   coefficients   are   SHEAR7   specific   data   which   are   specified  in  OrcaFlex  on  the  Section  Data  page  of  the  SHEAR7  data  form.  You  may  need  to  introduce  OrcaFlex  section   boundaries  which  coincide  with  the  locations  where  these  properties  change.   Area  moment  of  inertia  and  tension  variation  are  used   by  SHEAR7  to  calculate  modal  frequencies.  Typically  these   parameters   would   not   be   used   since   OrcaFlex   produces   a   modes   file   which   is   more   accurate   than   the   internal   SHEAR7  modal  calculations.   The  mass  per  unit  length  is  also  used  by  SHEAR7  to  calculate  modal  frequencies  and  again  this  aspect  of  the  SHEAR7   calculation   would   be   bypassed   when   using   a   modes   file   produced   by   OrcaFlex.   However,   the   mass   per   unit   length   is   used  by  SHEAR7  to  compute  damping  ratio.  For  almost  all  situations  in  OrcaFlex  the  mass  per  unit  length  is  constant   along  a  section.   One   exception   to  this   is   a  free-­‐flooding   line   with   sections   that   cross   the   water   surface.   Because   SHEAR7   has   a   single   mass   parameter   for   a   structural   zone   this   means   that   the   mass   is   effectively   smeared   across   the   zone.   A   free-­‐ flooding  line  has  greater  mass  below  the   water  surface  because   it  is   full  of   water.   Quite  often  this  issue  is  of  little   significance   but  if   you  wish  you   can   also   change   your   OrcaFlex   model   to   arrange   that   there   is  a   section  boundary  at   the  water  surface.  Similar  issues  arise  when  the  slug  flow  contents  method  is  specified.   The  other  modelling  choice  in  OrcaFlex  that  leads  to  properties  varying  along  the  length  of  a  section  is  when  profiled   line  types  (e.g.  stress  joints)  are  used.  Although  it  may   be  tempting  to  use  many  structural  zones  to   represent   the   variation   of   properties   the   SHEAR7   manual   strongly   recommends   not   doing   so.   Accordingly   OrcaFlex   outputs   a   single  structural  zone  for  a  profiled  line  type  section  which  has  the  effect  of  smearing  properties  and  we  believe  that   this  is  the  correct  approach  to  take  in  the  vast  majority  of  cases.  

SHEAR7  Mds  file   The  SHEAR7  Mds  file  export  facility  is  provided  on  the  Modal  Analysis  form.  To  use  it:   x

You  will  need  to  create  a  SHEAR7  data  file.  The  simplest  way  to  do  this  is  to  use  the  built-­‐in  facility  to  export   SHEAR7  data  files.  

x

Calculate  the  static  position  of  the  model.  

x

On  the  Results  menu  select  the  Modal  Analysis.  

x

Do   the   modal   analysis   for   the   line   you   want   to   analyse.   You   must   include   calculation   of   the   mode   shapes   and   include  all  modes  that  might  to  be  needed  by  SHEAR7.  

x

Select   the   VIV   page.   For   each  mode,   the  VIV  page   reports   the  extent   to   which  the   mode   is  transverse   (i.e.   in   the   VIV   direction),   inline   (i.e.   in   the   normal   drag   direction),   axial,   or   some   mixture.   This   helps   you   decide   which   modes  to  export  to  SHEAR7.  See  Mode  Selection  Table  below  for  details.  

x

The   final   column   of   the   table   shows   which   modes   are   currently   selected   to   be   exported   to   the   .Mds   file.   Whenever   you   calculate   the   modes   OrcaFlex   ticks   just   the  Transverse  and  Mostly   Transverse  modes,   but   you   can  then  change  the  selection  as  required.  

x

Use  the  Export  SHEAR7  Mds  File  button.  A  SHEAR7  .Mds  file  is  generated  for  those  modes  that  are  currently   ticked  for  export.  You  are  asked  where  to  save  the  file.  See  Values  Exported  for  details.   Warning:  

OrcaFlex  does  not  calculate  the  mode  slope;  in  its  place  OrcaFlex  outputs  zero.  See  Values  Exported   for  details.  

Exporting  from  Batch  Script  

The  SHEAR7  Mds  file  can  be  exported  from  an  OrcaFlex  batch  script  using  the  SHEAR7MdsFile  command.  

414  

w

 

VIV  Toolbox,  Frequency  Domain  Models  

 

Mode  Selection  Table   VIV  excitation  occurs  in  the  transverse  direction.  In  SHEAR7  mode  shapes  are  1D  since  it  assumes  that  all  the  modes   in  the  .Mds  file  are  purely  in  that  transverse  direction.  The  OrcaFlex  modal  analysis  is  fully  3D,  so  in  principle  you   should  therefore  only  export  modes  that  are  purely  transverse.   In  practice  the  natural  modes  do  not  always   neatly  divide  into  the  transverse,  inline   and  axial  directions,  so  you   will   sometimes   have   to   export   the   modes   that   are   nearest   to   being   transverse.   OrcaFlex   therefore   provides   the   mode   selection  table  to  help  you  decide  which  modes  should  be  exported  to  SHEAR7.   The  table  includes  the  following  columns:   x

Offset   Distribution  displays  a  measure   of  how  inline,  transverse  and  axial  the   mode  is.  For  details  see   Offset   Distribution.  

x

Mode   Type   classifies   each   mode   according   to   the   offset   distribution.   Transverse   means   that   the   transverse   component   is   more   than   90%   of   the   total,   Mostly   Transverse   means   that   it   is   between   50%   and   90%,   and   similarly  for  Inline,  Mostly  Inline,  Axial  and  Mostly  Axial.  Mixed  means  that  none  of  the  components  are  more   than  50%  of  the  total.  

x

Export   to   SHEAR7   Mds   File   determines   which   modes   will   be   exported   to   SHEAR7.   When   you   calculate   (or   recalculate)   the   modes   OrcaFlex   ticks   just   the   Transverse   and   Mostly   Transverse   modes,   but   you   can   then   change  the  selection  as  required.  See  Discussion  and  Examples.  

Values  Exported   The   first   line   in   the   .Mds   file   contains   the   number   of   selected   modes   and   the   number   of   nodes   in   the   line.   Then   follows   a   section   giving   the   angular   frequencies   of  the   selected   modes,   in   radians   per   second.   Finally   there   follows   a   section  for  each  selected  mode,  giving  the  Mode  Offset,  Mode  Slope  and  Mode  Curvature.   Warning:  

OrcaFlex  does  not  calculate  the  mode  slope;  in  its  place  OrcaFlex  outputs  zero.  The  mode  slope  is   only  used   in  SHEAR7  to   correct  the  predicted   RMS  acceleration   for  a  vertical  riser,  in  the  special   case   where   one   is   trying   to   model   the   gravitational   contamination   from   tilt   that   a   transversely   mounted   accelerometer   would   measure.   See   the   SHEAR7   documentation   for   details.   The   RMS   acceleration  is  listed  in  the  SHEAR7  .plt  plotting  files.  

Exported  Mode  Offset  and  Mode  Curvature  

The  OrcaFlex  modal  analysis  gives  vector  values,  but  SHEAR7  requires  scalar  values.  These  are  calculated  as  follows.   Let:   V  =  the  mode  shape  vector  calculated  by  OrcaFlex  at  a  given  node,   Vi  Vt  and  Va  =  V's  inline,  transverse  and  axial  component  vectors,  respectively,   VL  =  lateral  component  vector  of  V,  given  by  VL  =  Vi  +  Vt  (vector  sum),   m  =  maximum  value  of  |VL|  for  any  node,   C   =   curvature   vector   at   the  node,   for   the   mean   position  (this   is  a   vector   in   the  direction  normal   to   the   plane   of  curvature),   dC/dm   =   linear   rate   of   change   of   C   per   unit   maximum   lateral   mode   offset.   This   rate   of   change   (a   vector)   equals  the  change  in  curvature  vector  at  the   node  caused  by  applying  the  mode  offsets  V  to  all  the  nodes  and   then  dividing  by  m.   Then  the  scalar  Mode  Offset  and  Mode  Curvature  values  exported  by  OrcaFlex  to  the  Mds  file  are   Mode  Offset  =  Sign(Vt).|VL|  /  m   Mode  Curvature  =  Sign(inline  component  of  dC/dm).|dC/dm|   The  reasoning  behind  these  formulae  is  as  follows:   x

SHEAR7   assumes   that   the   exported   mode   offset   is   in   the   transverse   direction.   This   assumption   applies   to   the   power-­‐in   zone,   since   in   this   zone   SHEAR7   assumes   that   the   mode   offset   is   in   the   direction   of   VIV   excitation.   However   SHEAR7's   power-­‐out   calculation   will   still   be   valid   providing   the   mode   offset   is   lateral   (i.e.   no   axial   component),  since  fluid  drag  and  damping  occur  in  any  lateral  direction.  

x

OrcaFlex   should   therefore   ideally   export   the   transverse   component   of   mode   offset   for   the   power-­‐in   zone   and   the  lateral  offset  for  the  power-­‐out  zone.  However  OrcaFlex  does  not  know  the  power-­‐in  and  power-­‐out  zones,  

415  

w

 

VIV  Toolbox,  Time  Domain  Models    

flow

  Figure:  

Vortex  Tracking  Plane  

The  model  has  two  main  elements:   x

A   boundary   layer   theory   is   used   to   analyse   the   fluid   very   near   to   the   disc   surface,   where   viscosity   plays   a   dominant   role.   At   each   time   step,   the   boundary   layer   theory   gives   the   positions   of   the   two   separation   points,   and  a  new  vortex  is  created  at  each  of  these  two  points,  to  model  the  vorticity  being  generated  there.  

x

The   vortex   tracking   itself   handles   the   rest   of   the   fluid,   where   viscosity   is   much   less   important.   The   created   vortices   are   tracked   downstream   by   solving   the   inviscid   Navier   Stokes   equations   (which   are   approximately   valid  outside  the  boundary  layer).  This  models  the  wake  development,  i.e.  how  the  vorticity  flows  after  leaving   the  boundary  layer.  

Boundary  Layer  

When   the   flow   meets   the   disc   it   has   to   flow   around   the   disc   circumference   and   a   boundary   layer   is   formed.   Boundary   layer   theory   is   used   to   model   this   region,   where   viscosity   plays   a   crucial   role.   Sarpkaya   and   Shoaff   originally  used  the  Polhausen  boundary  layer  method,  but  since  then  this  method  has  been  superseded  by  simpler   and  more  accurate  methods.  OrcaFlex  uses  Thwaites'  method  (see   Young  1989)  for  both  models.   Some  of  the  fluid  flows  around  one  side  of  the  disc  and  some  around  the  other,  and  the  point  where  the  flow  splits  is   called   the   stagnation   point.   As   the   fluid   flows   around   the   disc   it   initially   remains   in   contact   with   the   disc,   but   it   typically  then  reaches  a  point  on  each  side  where  the  flow  separates.  These  are  called  the  separation  points,  and  at   these  points  vorticity  is  shed  from  the  disc.   The  boundary  layer  theory  gives  the  position  of  each  separation  point  and  the  strength  of  vorticity  shed  there  in  one   time   step.   A   new   vortex   of   this   strength   is   then   created   at   the   separation   point.   The   new   vortex   is   placed   at   the   •‡’ƒ”ƒ–‹‘’‘‹–„—–ƒ–ƒ†‹•–ƒ…‡ɉǤˆ”‘–Š‡†‹•…•—”ˆƒ…‡Ǥ –Š‡ model  2  ɉ‹•–Š‡Creation  Clearance  specified  by   the   user.   In   model   1ǡ ‘” ‹ˆ –Š‡ ”‡ƒ–‹‘ Ž‡ƒ”ƒ…‡ ‹• •‡– –‘ ̵̵̱ǡ ɉ ‹• …ƒŽ…—Žƒ–‡† –‘ „‡ –Š‡ ˜ƒŽ—‡ –Šƒ– ”‡•—Ž–• ‹ –Š‡ tangential   velocity   contribution   of   the   new   vortex   just   cancelling   out   the   existing   tangential   velocity   at   the   separation  point.   In   the   drawing   above   (and   in   the   OrcaFlex   3D   view)   the   stagnation   point   is   shown   as   a   small   triangle   and   the   separation  points  as  small  blobs,  part  way  around  the  disc  circumference.  The  vorticity  shed  from  the  two  sides  of   the  disc  is  distinguished  by  being  drawn  in  separate  colours,  to  denote  the  difference  in  the  direction  of  rotation   Ȃ   clockwise  for  vorticity  shed  from  the  upper  side  and  anti-­‐clockwise  from  the  lower  side,  as  seen  in  the  drawing.   Vortex  Tracking  

After   being   shed   from   the   boundary   layer,   the   vorticity   then   flows   downstream.   In   reality,   the   vorticity   is   shed   continuously  and  it  is  shed   along  the  neighbouring  parts  of  the  line  at  the  same  time,  so  as  it  flows  away  it   forms   sheets   of   vorticity,   one   on   each   side.   In   the   above   drawing   these   vortex   sheets   are   shown   as   red   and   green   lines,   since   the   drawing   shows   the   intersection   of   the   sheets   with   the   vortex   plane.   The   colour   denotes   the   sign   of   the   vorticity.   For  computation  purposes  the  model  has  to  discretise  the  vorticity  being  shed,  so  in  the  vortex  tracking  model  the   vorticity  is   represented   by   discrete  vortex  points.   The  vortex  sheet  is  therefore   represented  by  a  sequence  of  vortex   points,  each  one  of  which  represents  the  vorticity  of  a  short  length  of  vortex  sheet.  

424  

w

 

VIV  Toolbox,  Time  Domain  Models    

The  Creation  Clearance  can  also  be  set  to  '~',  in  which  case  the  new  vortices  are  placed  using  the  same  algorithm   as  in  model  1ǤŠ‹••‡–•–Š‡ˆƒ…–‘”ɉ–‘–Š‡˜ƒŽ—‡–Šƒ–”‡•—Ž–•‹–Š‡–ƒ‰‡–‹ƒŽ˜‡Ž‘…‹–›…‘–”‹„—–‹‘‘ˆ–Š‡‡™˜‘”–‡š at  the  separation  point  just  cancelling  out  the  previous  tangential  velocity  at  that  point.   Coalesce  Same,  Coalesce  Opposite  

These   are  non-­‐dimensional  thresholds   that   are  only  used   for   coalescing   in  model   2.  They   control   how   close   to   each   other   two   vortices   have   to   be   before   they   are   allowed   to   be   coalesced   into   one   combined   vortex.   For   details,   see   Coalescing.   There   are   separate   threshold   values   depending   on   whether   the   two   vortices   have   equal   or   opposite   signs   of   vorticity.  So  if  one  of  the  two  vortices  is  clockwise  and  the  other  is  anti-­‐clockwise  then  the  opposite  sign  threshold   will  be  used,  whereas  if  they  are  both  clockwise  or  both  anti-­‐clockwise  then  the  same  sign  threshold  will  be  used.   Reducing   the   thresholds  makes   the   model   coalesce   vortices   less   often,   so   the   model   will   have   to   keep   track   of   more   vortices   and   the   simulation   will   therefore   be   slower.   Conversely,   increasing   the   thresholds   makes   the   model   coalesce  more  readily,  so  fewer  vortices  need  to  be  tracked  and  the  simulation  is  faster,  but  less  accurate.   Our   experience   so   far   is   that   the   default   values   of   0.04   for   both   thresholds   gives   a   reasonable   balance   between   performance  and  accuracy.  If  the  mass  ratio  (=   mass  of  line  /  mass  of  water  displaced)  is  low  then  the  fluid  forces   are   more   significant,   and   in   these   cases   lower   coalescing   thresholds   may   be   needed   so   that   the   fluid   behaviour   is   more  accurately  modelled.   Vortex  Decay  Constant,  Vortex  Decay  Threshold  1,  Vortex  Decay  Threshold  2  

These   data   items   set   the   rate   of  vortex   strength   decay   in   both   vortex   tracking   models.   The   decay   model   is   described   below.   It   is   as   in   Sarpkaya   and   Shoaff's   report   (page   79)   and   the   default   values   for   this   data   are   as   given   in   that   report.  We  therefore  recommend  that  the  default  values  be  used  unless  you  wish  to  experiment  with  other  values,   for  example  to  calibrate  the  model.   Vortices  are  created  at  the  separation  points,   with  an  initial  vortex  strength   determined   by  the  tangential  velocity  at   the  separation  point.  The  strength  of  each  vortex  then  decays  at  a  rate  that  depends  on  how  far  the  vortex  is  away   from  the  centre  of  the  disc,  in  the  relative  flow  direction.   Let   R   be   the   disc   radius   (=   half   the   line   outer   diameter)   and   D   be   the   distance,   measured   in   the   relative   flow   direction  from  the  centre  of  the  line  to  the  vortex.  In  model  1,  at  each  variable  time  step  the  vortex  strength  is  scaled   „›ƒˆƒ…–‘”ɉ–Šƒ–†‡’‡†•‘ƒ•ˆ‘ŽŽ‘™•ǣ   x

ˆζ‡…ƒ›Š”‡•Š‘Ž†ͳȗ–Š‡ɉαȋͳ-­‐DecayConstant)  

x

Iˆη‡…ƒ›Š”‡•Š‘Ž†ʹȗ–Š‡ɉαͳ  

x

ˆ‡…ƒ›Š”‡•Š‘Ž†ͳȗδδ‡…ƒ›Š”‡•Š‘Ž†ʹȗ–Š‡ɉ˜ƒ”‹‡•Ž‹‡ƒ”Ž›™‹–Šǡˆ”‘ȋͳ-­‐DecayConstant)  to  1.  

The   effect   of   this   is   that   while   the   vortex   is   less   than   DecayThreshold1   radii   downstream   then   the   vortex   loses   DecayConstant  of  its  strength  (e.g.  DecayConstant=0.01  means  1%  decay)  per  variable  time  step.   While  the  vortex  is   between  DecayThreshold1  and  DecayThreshold2  radii  downstream  its  rate  of  decay  falls  linearly  (as  a  function  of   D)  to  zero.  And  when  the  vortex  is  more  than  DecayThreshold2  radii  downstream  then  there  is  no  decay.   Clearly  DecayConstant   must   be   in   the  range   0   to   1,   and  DecayThreshold1  must  be   less   than   DecayThreshold2.   Note   that   DecayThreshold1   and   DecayThreshold2   can   be   set   to   Infinity.   If   either   oˆ –Š‡ ‹• ˆ‹‹–› –Š‡ ɉ α ͳ-­‐ DecayConstant  always,  so  the  vortices  always  lose  DecayConstant  of  their  strength  per  variable  time  step.   The   same   decay   model   is   used   in  model   2ǡ‡š…‡’––Šƒ––Š‡ˆƒ…–‘”ɉ‹•ƒ†Œ—•–‡†–‘ƒŽŽ‘w  for   the   fact   that   model   2   uses   the   outer   time   step   instead   of   the   variable   time   step   used   in   model   1.   The   adjustment   results   in   the   same   rate   of   decay  per  unit  time.   Drag  Coefficients  

The   vortex   tracking   model   includes   the   drag   effects   in   both   the  transverse   and   inline   directions,   but  not   in   the   axial   direction.   When   the   vortex   tracking   model   is   used,   OrcaFlex   therefore   suppresses   the   components   of   the   usual   Morison   drag   force   in   the   transverse   and   inline   directions,   but   includes   the   component   in   the   axial   direction.   The   drag  coefficients  for  the  normal  directions  are  therefore  not  used,  but  the  axial  drag  coefficient  i s  used.   Results   The  Vortex  Force  is  available  as  line  force  results  variables.  This  reports  the  total  lift  and  drag  force.  Note  that  this  is   the   sum   of   the   force   generated   by   the   vortex   tracking   model,   which   is   in   the   inline   and   transverse   directions   and   already  includes  the  drag  force  in  those  directions,  plus  the  standard  Morison  drag  force  in  the  axial  direction.  

428  

w

 

VIV  Toolbox,  Time  Domain  Models  

 

The  stagnation  and  separation  points  are  available  as  line  angle  results  variables.   Transverse  VIV  Offset  is  also  available  as  a  line  position  results  variable.  

9.2.3

VIV  Drawing  

With   the  time-­‐domain   VIV  models  you   can   control   how   various   aspects   of   VIV   are   drawn   on   the  3D   view,   by  setting   data  on  the  VIV  Drawing  page  on  the  line  data  form.   Arc  Length  Intervals  

You   can   control   which   nodes   have   VIV  detail   drawn,   by  specifying   one   or   more  Arc   Length   Intervals.   An   arc   length   interval  specifies  a  contiguous  length  of  line   From  one  specified  arc  length  To  another.  The  VIV  details  are  drawn   for  all  nodes  whose  arc  lengths  fall  in  any  one  of  the  specified  intervals.  For  convenience  '~'  in  the   From  column   means  End  A  of  the  line,  and  '~'  in  the  To  column  means  End  B.   For   example,   to   view   the   detail   for   one   node   only,  specify  1   arc   length  interval   and   set   both   its   From  and  To  values   equal  to  the  arc  length  of  that  node.  Whereas  to  view  the  detail  for  the  whole  line  specify  1  arc  length  interval  and   set  both  its  From  and  To  values  equal  to  '~'.   What  is  Drawn  

The   following   VIV   details   are   drawn   for   all   nodes   whose   arc   lengths   fall   in   any   one   of   the   specified   arc   length   intervals.   For  all  the  time  domain  VIV  models,  the  vortex  force  is  drawn  as  a  line  radiating  from  the  disc  centre  in  the  direction   of   the   vortex   force.   The   line   length   is   scaled   so   that   a   vortex   force   equal   to   the   standard   Morison   drag   force   with   Cd=1  in  a  relative  velocity  of  1m/s,  is  represented  by  a  line  1  disc  radius  long  (i.e.  just  reaching  the  edge  of  the  disc).   For  the  vortex  tracking   models  only,  and  only  if  the   maximum  number  of  vortices  logged   is  set   greater  than  zero,   then  the  following  extra  details  are  drawn.   x

The   node   is   drawn   as   a   disc   (even   if   you   have   not   specified   nodes   drawn   as   discs)   and   the   stagnation   and   separation  points  are  drawn  on  the  edge  of  the  disc.  

x

The   positive  and   negative   vortices  are  drawn  as  circles  whose  areas  are  proportional  to  the  vortex  strengths.   The  constant  of  proportionality  can  be  controlled  by  setting  Area  per  Unit  Strength.  

x

The  centre-­‐lines  of  the  positive  and  negative  vortex  sheets,  and  the  wake  line  are  drawn.   Note:  

The  vortices  and  sheet  centre-­‐lines  drawn  are  limited  by  the  specified   maximum  number  of  vortices   logged.  

For   all   these   items   you   can   control   the   pen   used   for   drawing.   With   the   vortex   tracking   models,   for   example,   this   allows   you   to   suppress   (by   setting   the   pen   style   to   null)   or   downplay   (by   choosing   a   suitable   colour)   individual   aspects  of  the  detail.  

429