Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Partitioning Problems with External Meshing in Cubit? #1174

Closed
ARBolton96 opened this issue Jan 17, 2023 · 2 comments
Closed

Partitioning Problems with External Meshing in Cubit? #1174

ARBolton96 opened this issue Jan 17, 2023 · 2 comments

Comments

@ARBolton96
Copy link

Hello,

I'm running a very simplified test model meshed in Cubit with SpecFEM2D and I'm running into some issues with the partitioning. Using the cubit2specfem2d.py journal file given in the SpecFEM2D utils directory (very helpful by the way!), I was able to create the necessary mesh files and set up a SpecFEM simulation.

However, when I run this model over multiple processors, this occurs:
mymesh

All the energy in the simulation stays within the source layer! There is no transmission into the deeper model, which I have verified with the station configuration in the above image, as well as the output postscript files.

If I run this same exact mesh over a single processor (thus avoiding mpirun commands), I get this:

mymesh_oneproc

Transmission into the subsurface!

I then went to the provided EXAMPLES directory and looked at the salt_dome_CUBIT_mesh_from_Ronan_Madec/Stacey_normal_fluid_solid (full salt dome model with overlying water layer), and edited the Par_file to run over multiple processors. The results were:

saltmesh

Once again perfect transmissions, and this time over multiple processors! Very perplexing to me.

I am assuming the problem I'm encountering is with the partitioning of the simple mesh I've built. Have others encountered this issue, or have any general guidance? I can provide any of the SpecFEM input files or my Cubit mesh and cubit2specfem.py files, if necessary. Thanks!

@AbolfazlKhanMo
Copy link

Hey,
This looks a little bit odd! Can you please provide your mesh's cubit journal file (or .py file)? If not, make sure you're merging all volumes (merge vol all) and imprinting them (imprint vol all) before starting the meshing process.
Hope this helps!

@ARBolton96
Copy link
Author

Hot dog!!! That is exactly what I was missing! Thank you so much @AbolfazlKhanMo, that was so very helpful. I completely missed those commands in the salt dome example.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants