[Pratik1810]

I am working on pulsatile FSI simulation in blood vessel. The details of geometry and boundary conditions for fluid (p and U) and solid are mentioned in attached figures (solid geo - 2mm thickness, linear elastic material with E = 7e4 and nu = 0.3). When I start the FSI simulation with coupling of fluid and solid interface at particular time (after 1 or 2 cardiac cycle for stabilize the fluid), it working fine upto only running the fluid simulation, as soon as the coupling starts after certain time, in one or two time step, the fluid forces suddenly increased, thereby increases displacement (1e2 order) and further increases in Courant number and finally simulation crashes (for your ref. figure is also attached).


I also tried a couple of solutions that are listed in the user guide of solids4Foam, such as use of conformal mesh, first try to stabilize fluid solver (Ex: running one or two cardiac (pulsatile velocity) cycle, then after switch on coupling), using higher order stiffness (like 1e07 or 1e09), changing the fluid-solid interface coupling scheme (strong coupling : IQNILS or Aitken), increasing or decreasing outer correction tolerance, increasing relaxation factor etc.


I am also trying to run the simple pipe flow (one inlet and one outlet case) with the same boundary condition. But the same issue rises. !!


Any solution to this problem ???

[bigphil]

Hi Pratik,

Have you tried the different versions of the 3dTube case described at https://solids4foam.github.io/tutorials/tutorial4.html?

Is it possible to take that case and replace the geometry with your geometry?

Also, did you try a solid-only simulation to ensure your solid settings and mesh are suitable?

Philip

[Pratik1810]

Originally Posted by bigphil

Hi Pratik,

Have you tried the different versions of the 3dTube case described at https://solids4foam.github.io/tutorials/tutorial4.html?

Is it possible to take that case and replace the geometry with your geometry?

Also, did you try a solid-only simulation to ensure your solid settings and mesh are suitable?

Philip

I have already tried the other robin tube case with pulsatile veloicty, but it's failed in starting of simulation (it is running fine with pressure pulse case - default case). The fluid and solid simulation separately are working well. But problem start when it couples !!!

[bigphil]

I suggest starting with a working model (like 3dTube) and incrementally making changes to it towards your case. You will then discover what exactly is causing the issue in your case.