[00Sora]

Hello everyone!


I’ve been trying to implement the EARSM model (as used in ANSYS) in OpenFOAM. I’m basing my work on the version available here:
https://gitlab.com/ejp/BSL-EARSM
I’ve already made a few modifications so that it works with OpenFOAM v2112. One of the main changes is that I replaced the backwardsCompatibilityFunctions with the standard omegaWallFunction. Since the difference for y+ < 1 is small, I don't think thats the main issue.
However, I’m still struggling to match the verification results. In the attached plot you can see a significant discrepancy between my implementation and the reference. Interestingly, the velocity results for the standard kOmegaSST model are actually closer to the verification than the EARSM implementation I’m trying to reproduce. So could this be a general version-related issue in OpenFOAM?

Another problem is stability: for several simple test cases (e.g., a flat plate) the solver becomes unstable and crashes after just a few iterations.
I’ve posted my full code in the thread. If anyone has ideas about what might be going wrong or has experience with EARSM in OF v2112, I would really appreciate your help!

Thank you!!

[klausb]

Start with reasonable defaults for tolerances and convergence like 1e-6 to 1e-8, your values are extremely low. Monitor the residuals and post your observations.