Remeshing error in fluid domain during FSI simulation

[Angelo Puglisi]

Hello everyone,
I was running an FSI simulation on an elastic tube in order to verify that the TetGen remeshing tool implemented in svFSI worked.
The simulation ends with an error that I don't understand, as shown in the images.
I apply a Dirichlet boundary condition with temporal dependence unsteady in inflow.
In outflow apply a Neumann boundary condition with temporal dependence steady equal to 0.
For the solid domain the displacement of the inlet and outlet surfaces in the axial direction is set to 0.
I upload all the data to this drive folder in case anyone wants to consult it.
https://drive.google.com/drive/folders/ ... share_link
Thanks a lot for the help.

[luca cerrato]

Hi Angelo,
you have made a good question because i have the same problem when i try to run a FSI simulation in my pc and requires remeshing. I have got the same output error.
Someone could help us in this forum?
Thanks in advance for your support.

[David Parker]

Hi Angelo and Luca,

I am looking into this now using Luca's data.

Cheers,
Dave

[Vijay Vedula]

@Dave: any update on this? can we create a GitHub issue on the Fortran code? I can help find the issue. We should also have an example in the https://github.com/SimVascular/svFSI-Tests repository for remeshing.

[David Parker]

Hi Vijay,

I looked at the Fortran code a bit but I did not quite understand what was going, note able to map all of the original mesh nodes to the new mesh. If I increase the number of passes then all of the nodes can be mapped, maybe this is a tolerance issue.

I am converting the remeshing code to C++ so I thought I would figure this out when doing that but got sidetracked on some other things.

I am putting together a couple of remeshing tests.

Cheers,
Dave

[Vijay Vedula]

Hi Dave,

Thanks, but usually the algorithm is designed in such a way that it should find all the nodes in a couple of passes. This case appears peculiar. Let me know if I can help in any way. One thing that is unclear to me is that although the baseline fluid mesh has only 11514 nodes (or ~44K elements), after remeshing using the same edge size, TetGen creates 83K nodes (~500K elements). I am not sure why this is the case.

Best,
Vijay