BL + regional refinement

Provides a system for patient-specific cardiovascular modeling and simulation.
POST REPLY
User avatar
Kimberly Stevens
Posts: 15
Joined: Mon Dec 03, 2018 12:06 pm

BL + regional refinement

Post by Kimberly Stevens » Fri Nov 01, 2019 8:26 am

Hello,

I've been running into a problem when trying to use BL meshing with regional refinement. I've seen this now with different meshes of different sizes. I have gotten reasonable results when running the exact same simulation with either regional refinement or BL meshing alone, so it seems the trouble is not with the simulation parameters. The pressure goes negative after just a few hundred time steps, and there is a non-zero velocity at the wall in the area of the regional refinement (see attached). The mesh looks reasonable to my eye. Do you have any suggestions on what the problem is?

Thanks,
Kimberly
Attachments
velocity.PNG
velocity.PNG (40.25 KiB) Viewed 219 times

User avatar
Kimberly Stevens
Posts: 15
Joined: Mon Dec 03, 2018 12:06 pm

Re: BL + regional refinement

Post by Kimberly Stevens » Fri Nov 01, 2019 9:49 am

I've attached the mesh here in case it's useful. I'm using a pulsatile flow with an average flow rate of about 2 mL/s. dt=0.00025.
https://drive.google.com/drive/folders/ ... sp=sharing

User avatar
Weiguang Yang
Posts: 110
Joined: Mon Apr 07, 2008 2:17 pm

Re: BL + regional refinement

Post by Weiguang Yang » Sat Nov 02, 2019 8:13 pm

The aneurysm is missing in your walls_combined.vtp. If you use walls_combined.vtp to set up no-slip boundary condition in your svpre file, the aneurysm wall won't be included resulting in non-zero velocities on that part. You probably need to retag the aneurysm face as wall_aneurysm and remesh the model.

User avatar
Kimberly Stevens
Posts: 15
Joined: Mon Dec 03, 2018 12:06 pm

Re: BL + regional refinement

Post by Kimberly Stevens » Wed Nov 06, 2019 11:28 am

Thanks for your response. I opened up the walls_combined.vtp and I see exactly what you're saying. But in the original model the aneurysm is not designated as a separate face from the rest of the wall--it's all part of the same face, and is designated as a wall. I don't know why it excludes the aneurysm from the walls_combined.vtp file, but it only seems to happen when I'm doing regional refinement and BL meshing both. The excluded section is the area where I was doing regional refinement. If I do only regional refinement, or only BL meshing, the aneurysm is included in walls_combined.vtp.

User avatar
David Parker
Posts: 1677
Joined: Tue Aug 23, 2005 2:43 pm

Re: BL + regional refinement

Post by David Parker » Wed Nov 06, 2019 3:05 pm

Hi Kimberly,

Regional refinement is assigning a separate face ID to the refined mesh. This is a bug and I've opened an SV GitHub issue on it (https://github.com/SimVascular/SimVascular/issues/606).

A work around is to modify the .vtp file created in the project's Meshes directory. In the Meshes directory you will see three files for each mesh you've created. For example, I've created a mesh named model1-bl-refine so there are the following three files under the project's Meshes directory

Code: Select all

model1-bl-refine.msh
model1-bl-refine.vtp
model1-bl-refine.vtu
You can import the model1-bl-refine.vtp file by right clicking on the Models node in the SV Data Manager, I named it combine-mesh. You can then combine the surface faces into a single face by selected the two surface faces (control left mouse button) and selecting the Combine button.

Saving the project creates a file in the Models directory named combine-mesh.vtp. Quit SV and then copy that file to the Meshes/model1-bl-refine.vtp. You can now start SV, read in the model1-bl-refine mesh and create a simulation with it. See pictures below.

You can also modify this file using a Python script. I can explain how to do that if you'd like.

Cheers,
Dave
Attachments
comb-1.png
comb-1.png (190.61 KiB) Viewed 167 times
comb-2.png
comb-2.png (181.85 KiB) Viewed 167 times
comb-3.png
comb-3.png (180.7 KiB) Viewed 167 times

User avatar
Kimberly Stevens
Posts: 15
Joined: Mon Dec 03, 2018 12:06 pm

Re: BL + regional refinement

Post by Kimberly Stevens » Mon Nov 11, 2019 2:54 pm

Great, thank you, that's very helpful!

POST REPLY