Generalizing FWH-Based Aeroacoustic Noise Prediction in SU2 for Diverse Flow Configuration

[riddhi1228]

Hi @huozdemir, I'm interested in Project FWH for GSoC 2026. I'm a Mechatronics Engineering student with Python experience. Could you point me to the existing FWH implementation so I can study it before writing my proposal?

[riddhi1228]

Hi @huozdemir — following up on my earlier message. I've now studied the existing implementation at EduardoMolina/SU2PY_FWH and identified the key limitations:

1.Python 2 syntax throughout — won't run on modern systems
2.Freestream values hardcoded for the Lagoon landing gear case (FreeStreamPressure = 5895.49, FreeStreamMach = 0.128)
3.Geometry clipped to l_sim = 3.0 — hardcoded for one wind tunnel model length
4.No validation output or test case with expected results

My proposed approach: port to Python 3, move all case-specific constants to CLI args/config, remove geometry assumptions, and add a validated tandem cylinder test case with documented expected output.
Does this align with what you have in mind for Project FWH? Happy to share my draft proposal.

[riddhi1228]

Since my last message, I've completed the following pre-application work:
9 bugs fixed in SU2PY_FWH (PR #3 now open on EduardoMolina/SU2PY_FWH):

Python 2 → 3 port (24 locations)
SU2 v8.4 config key compatibility
Array sizing fix for 2D cases
Input validation and argparse CLI

End-to-end pipeline validated:

Square cylinder, Mach 0.1, Re 22000, SST, 500 timesteps
Observer 3 (crossflow) shows clear periodic vortex shedding signal
Physically correct: near-zero signal at inline observers

Happy to share the acoustic plot or bug documentation if useful.