PyFR robustness and suitability for strongly compressible and wall-bounded turbulent flows

General
1

Hi,

I’m exploring the use of PyFR for strongly compressible flows that may involve shock waves, acoustic waves, and near-wall flow structures.
Could you please share how robust and efficient PyFR is in such cases, especially compared to established compressible solvers like OpenFOAM?
Also, I’d like to know whether PyFR is considered suitable for high-Re DNS or implicit LES of wall-bounded turbulence.

Thanks a lot for your time and insights!

Best regards

2

PyFR’s primary shock capturing method is entropy filtering, see this paper and test cases. It is a robust and efficient method that, like all shock capturing methods, does its best to apply the minimum amount of dissipation to maintain sane physics without smearing out flow structures.

I don’t have numbers in front of me, but OpenFOAM is not known as a performant code in terms of computational throughput, that is not its primary goal, where it is with PyFR.

PyFR is primarily used as an implicit LES → DNS solver with capability for both compressible and incompressible flows (through artificial compressibility). As with all solvers, your computational resources are likely your primary limitation, but the flow configurations you describe are very much in PyFR’s wheel-house.

3

Thanks for your discussion.