CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > CD-adapco

fsi

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   January 14, 2009, 00:16
Default fsi
  #1
sun
Guest
 
Posts: n/a
i just searched the archives for FSI and the last thread found was dated 2005. Is there any update on FSI as of 2008? Which software should i use to do fsi? star-cd or star-ccm? so far i found nothing when searching FSI in star-ccm manuals. Please let me know if anyone has done FSI using starcd or starccm and a few tips or two on doing it. Somehow this area has growing demand from my customers. I'm doing a bit of FSI by passing the force data to Nastran at the moment (old style). Is there a better way (without MPCCI if possible)?

  Reply With Quote

Old   January 14, 2009, 02:52
Default Re: fsi
  #2
ping
Guest
 
Posts: n/a
Both STAR-CCM+ v3.06 and STAR-CD v4 have extensive in-built FSI features using the same FV solver to solve solid stress and thermals, and so are much more efficient than coupling to an external FE solver, either directly or via MPCCI. They can also handle millions of DOFs on a normal PC with modest memory. Direct two-way coupling to Abaqus (ie don't need MPCCI) is in STAR-CD and coming soon to CCM+. Tips - use STAR-CCM+ since it is so easy to learn and all in the GUI - lots of help and guidance available in the online Help.
  Reply With Quote

Old   January 14, 2009, 06:10
Default Re: fsi
  #3
sun
Guest
 
Posts: n/a
I could not find any help or guidance on FSI anywhere in star-ccm online help. Can you tell me where i can find that guidance you mentioned?

  Reply With Quote

Old   January 14, 2009, 08:56
Default Re: fsi
  #4
allan
Guest
 
Posts: n/a
Hi

There aren't any tutorials yet in ccm+ for stress analysis. ccm+ can only solve fsi where one uses a converged cht solution to solve for the stress field.It is uncoupled. Star-cd v4.08 has an interface so that it can be directly coupled to abaqus for true fsi.

Contact support for more info.

allan
  Reply With Quote

Old   January 14, 2009, 21:23
Default Re: fsi
  #5
sun
Guest
 
Posts: n/a
Coupled to Abaqus? Will it do FSI (e.g. fan blade deflection due to wind load or hydro/aero elasticity) without coupled to other software?

  Reply With Quote

Old   January 15, 2009, 05:07
Default Re: fsi
  #6
ping
Guest
 
Posts: n/a
Allan your answer is quite incorrect!

Both STAR-CCM+ and STAR-CD now do two way coupled FSI coupling both thermals and pressure forces between fluids and solids for example within the FV solver. The key limitation is that displacements must be small - for large displacements use the new direct coupler to Abaqas in STAR-CD v4.08.

You can also just do stress analysis - ie whole domain is one or more solids, with typical FEA-type boundaries such as free, fixed, displacement etc.

There is good documentation within the online ccm+ Help at Modeling > Modeling Solid Stress. But I agree there are no tutorials at this point (v3.06). SS and transient studies can be done - eg a vibrating fan blade as requested by Sun, as long as the displacements are small, since mesh movement is not done, only displayed as a field function (coming later)
  Reply With Quote

Old   January 16, 2009, 04:39
Default Re: fsi
  #7
sun
Guest
 
Posts: n/a
thanks ping, will look it up. The fan displacement is expected around about 20 to 30mm. The blade camber thickness is about 2mm. The radius from tip to hub is about 80mm. I suppose if we can get the stress value we can convert that into strain and use the strain as a half-complete FSI calculation in a steady-state fashion. Next stage for a full transient FSI (like a non pre-determined moving mesh), we go with Abaqus or v4.10 perhaps?

  Reply With Quote

Old   January 17, 2009, 23:18
Default Re: fsi
  #8
ping
Guest
 
Posts: n/a
yes this is large displacement, so blade movement will affect fluid flow, so to do it well you need to couple to Abaqas in STAR-CD v4.08 at present. Both STAR-CD and CCM+ will plot strain as a contour plot - the feature missing is that the mesh does not actually move when using their internal FSI (and this accounts for them only being useful for small strains). I imagine this will come soon since the mesh morpher seems to work well within the fluids solver.
  Reply With Quote

Old   January 20, 2009, 00:17
Default Re: fsi
  #9
sun
Guest
 
Posts: n/a
hi Ping, can you tell me briefly step-by-step how to setup the model for FSI in v4? I dont care much about accuracy, i just want to show the strain or stress on the blades after the flow field is calculated (half-complete fsi). i want to compare the strain results with Nastran's. hope you can tell me the steps.

  Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On



All times are GMT -4. The time now is 19:34.