[Frisii]

Dear All,


I'm working on performing an 2-way FSI simulation on a human thoracic artery. However I'm having trouble stabilizing the initial steps of the simulation, I've tried many things to try and enhance the stabilization, but to no avail.


The FSI simulation consists of a realistic human thoracic artery through with an inlet flow-rate UDF with a plug-flow profile. The walls have an isotropic linear elastic response and the coupling is between the force data from the fluid simulation and the displacement data from the mechanical simulation.



In the attached images, File1,File2, it can be seen that distortion occurs, but only in certain areas, File1 shows how to distorted geometry looks, and File2 is an overview of all extra edges generated by the re-meshing. The divergence starts at pretty much the first time step, but it takes a while before the simulation finally crashes.


I've done both simulations separately with estimated boundary conditions and they both converged fine. I've used results from the flow simulation to initialize the flow field for the FSI simulation as well. I'm assuming the meshing is not the issue here, bot meshes are reasonably fine, with the mechanical shell being around 4/5 elements thick and the fluid having a boundary layer almost double that resolution. Meshes are tetra-dominated. My assumption that this is not the issue rests on the fact that the distortion if not evenly spread, and mostly concentrated in the flow-caps, which have the highest mesh quality due to their perfect geometry, in contrast to the rest of the artery.

My guess is that the problem lies somewhere with the Fluent dynamic re-meshing, as both sub simulations converged fine, maybe something with the meshing being to slow or too fast for the flow/displacement. Sadly I've reached the limits of my knowledge on the subject of re-meshing. I've included my current settings, and the journal file I use to submit this to the cluster.



Can someone help me figure out what is going wrong here? Do I need to adjust my re-meshing settings, or is there something else I'm overlooking? I've already set the time step to very small sizes, and I am currently trying to increase the stiffness of the wall by a large margin to obtain a stable solution and maybe then transfer to more flexible walls.



Thank you in advance,



P.S. additional data was left out to prevent bloat, but can be delivered on-demand