Instability in Hypoelastic Simulations

[ChrisZYJ]

Description

Commit 10b2209 (#727) introduces instability in some hypoelastic simulations compared to previous commit 108805c.

Test Case

• Simulation: Simple acoustic wave propagation with sound waves from left focused at a centered stone

• Case file: case_hypo.txt

Step 300 of the simulation (with commit 10b2209):

Comparison with the previous commit 108805c (top:108805c7 ; bottom:10b2209):

output.mp4

Environment & Configuration

• Ubuntu 22.04.4 LTS
• GNU 11.4.0
• Using CPU with both debug enabled and disabled

[sbryngelson]

Unsure on this, perhaps @mrodrig6 can comment or @mcarcana

[mrodrig6]

@ChrisZYJ Are you running this on two ranks? It might be worth running this on one rank.

If it is ran on 1 rank, we will need to investigate which terms in the hypoelastic formulation are breaking the symmetry.

[ChrisZYJ]

@mrodrig6 Thanks for getting back to me! The same instability occurs with 1 rank or more ranks.

I'm creating a new PR to revert the changes for just the hypoelasticity module for now, if that's alright. It should not affect your PR in progress.