Negative Pressures for FSI Simulation

[Kewil Bhima]

Good day,

I am currently using the svFSI solver to run a FSI simulation of a heart model consisting of the pulmonary veins, left atrium, left ventricle and aorta. After following the procedure of running a rigid wall simulation and the prestress simulation, I have then run the FSI simulation and I am obtaining negative pressures within my model. I am following the ALE method similar to the tutorials on YouTube but applying it to my model.

I have obtained inflow values from literature for the 4 pulmonary veins and I have used RCR boundary conditions for the 4 outlets in the model (also obtained from literature). I have used various different RCR conditions from different journal articles and all of them result in negative pressures. All of the conditions I have specified are for a normal, healthy heart model.

I have built the svFSI solver from the source code and I am running the simulations on Ubuntu 22. I have also only run the simulations over one cardiac cycle to save time. If I ran the simulation over more cycles, would it normalise and give me the correct pressures? The mesh generated is quite coarse as this is to save time and ensure the boundary conditions are correct before I simulate the finer mesh with more cardiac cycles. I generated the mesh using the boundary layer method in SimVascular and have attached a screenshot of the parameters which I used. The mesh complete was exported and renamed and can be found in the different folders on the Google Drive.

Please can someone look at this link for the uploaded project and assist as I am struggling to figure out what could be the issue. Google Drive link: https://drive.google.com/drive/folders/ ... sp=sharing
The simulations run without any errors generated on my side.

Once I have the correct results from the FSI simulation, the plan is to add the aortic valve and mitral valve into the model in order for my model to be more realistic.

Thank you,
Kewil Bhima

[Aaron Brown]

Hi Kewil,

I'm not an expert on FSI simulations, but I'll try to help. I think the following is the example you have been working from?

https://github.com/SimVascular/svFSI-Te ... /svFSI.inp

I noticed that this input file contains the line

Initialize RCR from flow: t

on line 110, while your input file does not. Looking through the source code, this apparently sets the initial RCR pressure based on the initial velocity and pressure fields of the FSI simulation. If it is not set, the initial RCR pressure is set to 0. This may be causing issues. Could you try adding this line to your input file and rerunning 3) FSI Simulation?

Thanks,
Aaron

[Kewil Bhima]

Hi Aaron,

Thank you for your response. I have since found several errors with the model. The boundary layer mesh for the solid domain was intersecting itself at the inlets and outlets so I have since refined the mesh to remove this. Also, my number of time steps were incorrect and I have calculated RCR values for my specific model instead of using values from literature. I am in the process of re-running the simulations and will add the line that was missing when I run the FSI simulation.

I had a look at the master input file and I see that "Initialize RCR from flow" is not explicitly defined in the code, so I probably overlooked that small detail.

I appreciate the help and I will post an update after the FSI simulation has ran.

Thanks!