Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

FSI in a closed system

Please login with a confirmed email address before reporting spam

Dear all

We are having some erratic results when trying to model the variation in pressure and flow (using FSI) in a fluid which is fully encapsulated in a polymer membrane. The idea is that by applying a pressure on the polymer we cause the fluid to move and apply a pressure elsewhere (as determined by the architectue of the polymer case).

First of all we set up the problem by defining:

1) The pressure applied to the polymer
2) The initial pressure (an arbitary value) and the intial flow = 0

This did not seem to solve, following which we were advised to set a point in the fluid with a known pressure (constant throughout the simulation) This solved but does not really represent the true physics.

Is anybody able to advise on how best to set this problem up ? (usually in the FSI tutorials, the systems modelled are open flow with a known output and input flow, however it appears to be more complicated when setting up a closed system in which the fluid pressure and flow are dependent on the boundary conditions applied to the solid (polymer membrane)

Any help direction or advice would be very much appreciated.


Regards, Duncan

7 Replies Last Post Nov 15, 2013, 10:31 a.m. EST
Nagi Elabbasi Facebook Reality Labs

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 14, 2013, 9:28 a.m. EST
Dear Duncan,

It is typically harder to get convergence when the system is closed but should be doable without the pressure point setting since you have a flexible solid. Try putting some compressibility in the fluid (even if more than the physical value at least initially) and also try using a fully coupled solver.

Nagi Elabbasi
Veryst Engineering
Dear Duncan, It is typically harder to get convergence when the system is closed but should be doable without the pressure point setting since you have a flexible solid. Try putting some compressibility in the fluid (even if more than the physical value at least initially) and also try using a fully coupled solver. Nagi Elabbasi Veryst Engineering

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 15, 2013, 7:27 a.m. EST
Dear Nagi

Thanks for the informative reply, i tried adjusting the solver to a fully coupled one and the problem solves.

Do you know of any documentation that explains the capabilites, limitations etc of each solver type ?. I am not a specialist on the numerical theory behind these things but it would be good to try and educate myself. Thanks agains for the help
Dear Nagi Thanks for the informative reply, i tried adjusting the solver to a fully coupled one and the problem solves. Do you know of any documentation that explains the capabilites, limitations etc of each solver type ?. I am not a specialist on the numerical theory behind these things but it would be good to try and educate myself. Thanks agains for the help

Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 15, 2013, 8:31 a.m. EST
Hello Duncan,
Maybe an approach similar to that used in the "hyperelastic seal" model could be used, circumventing the need to solve for the fluid flow itself (This would seem reasonable if you don't have major dynamic effects in the sstem -only quasi-static). See www.comsol.com/model/hyperelastic-seal-206
Just my $.02.
Jeff
Hello Duncan, Maybe an approach similar to that used in the "hyperelastic seal" model could be used, circumventing the need to solve for the fluid flow itself (This would seem reasonable if you don't have major dynamic effects in the sstem -only quasi-static). See http://www.comsol.com/model/hyperelastic-seal-206 Just my $.02. Jeff

Nagi Elabbasi Facebook Reality Labs

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 15, 2013, 9:38 a.m. EST
Dear Duncan,

I'm glad to hear that the problem now solves! I am sure there is documentation or COMSOL training material describing the solver capabilities and limitations but I couldn’t find it right now. In general, the full coupling is more likely to work without adjustment for strongly coupled problems compared to the segregated solver since it is an iterative scheme. It is however in many cases not a feasible option because it requires more memory. You can find more info if you do a literature search for monolithic vs. partitioned solvers.

Nagi Elabbasi
Veryst Engineering
Dear Duncan, I'm glad to hear that the problem now solves! I am sure there is documentation or COMSOL training material describing the solver capabilities and limitations but I couldn’t find it right now. In general, the full coupling is more likely to work without adjustment for strongly coupled problems compared to the segregated solver since it is an iterative scheme. It is however in many cases not a feasible option because it requires more memory. You can find more info if you do a literature search for monolithic vs. partitioned solvers. Nagi Elabbasi Veryst Engineering

Nagi Elabbasi Facebook Reality Labs

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 15, 2013, 9:48 a.m. EST
Thanks Jeff, a valuable 2¢ :) This approach really helps simplify quasi-static FSI problems where the fluid is just providing a pressure vs. volume constraint.
Thanks Jeff, a valuable 2¢ :) This approach really helps simplify quasi-static FSI problems where the fluid is just providing a pressure vs. volume constraint.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 15, 2013, 10:29 a.m. EST
Dear Jeff
Thanks for the info ! ...... unfortunately we have the task of anaysing the dynamics of the system, but i take note of the quasi static approach as this may come in useful in the future ! Thanks !

Regards, Duncan
Dear Jeff Thanks for the info ! ...... unfortunately we have the task of anaysing the dynamics of the system, but i take note of the quasi static approach as this may come in useful in the future ! Thanks ! Regards, Duncan

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Nov 15, 2013, 10:31 a.m. EST
Dear Nagi

Thanks !

Regards, Duncan
Dear Nagi Thanks ! Regards, Duncan

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.