[DrZoidberg]

Dear CFD community,


I obtained a solution for a compressible airflow on a coarse mesh that looks pretty good: all residuals went down <1e-6 as well as the netto heat-flux and mass-flux went down <0.1%. So I exported this solution in an interpolation file.


Flow characteristic:
- Pipe flow that expands to environment through a straight nozzle.

- subsonic inlet, subsonic outlet, pressure ratio in/out = 100

- Max. Ma=7 in the shock diamond pattern after the outflow of the end of the pipe
- kwSST turbulence model


The pbs coupled solver with pseudo-transient timestepping achieved the solution.



Now, when I load a finer mesh and read the interpolation file, the solver diverges immediately due to turbulence and energy resiudals. The whole solution starts to show "cracks" everywhere, i.e. it cannot be found out what exactly might be the reason (e.g. if there would be odd effects in near outlet or close to walls).



I already tried, but it the problem did not improve:

- switching pbd <> dbs
- starting with spatial discretizaion 1st order

- disabling energy and turbulence in the equations for the first iterations


I would very much appreciate any hints!


Best regards