[Dónal]

I just updated SU2 from version 7.0.0 to 7.2.0 and my solution has gone from slowly converging in 7.0.0, to diverging after 20-50 iterations in 7.2.0. The residuals appear stable and there are no obvious errors when checking the solution. I am trying to work out what changes in the version may be causing this.

The case is an M= 2.5 flow passing over a strut (0 heat flux viscous boundary) which generates vortices. The boundary uses the supersonic boundary condition and the outlet is a Riemann boundary with static pressure equal to the inlet condition. The remaining boundaries of the domain are Euler walls

I have tried switching to first order for both flow and turbulence and tried different meshes. The mesh I'm trying to use has poor orthogonality in some spots but this is a trade off for sufficient refinement at sharp edges. I have tried another mesh with better orthogonality and had the same issue.

I have tried both the FGMRES and BCGSTAB solvers, and the Roe and Lax-Friedrich convective schemes. I have also tried adjusting the CFL and using the adaptive cfl.

I have also tried a much simpler case which I was able to run successfully but only after switching to first order for flow and turbulence

My next step is to try adjusting the slope limiter and the entropy coefficient and hopefully get a first order solution that I can then start a second order solution from.

If anyone has any ideas what changes may have caused this, or just suggestions on how to proceed it would be greatly appreciated

[Dónal]

the version I am currently using is actually 7.2.0, not 7.1.1. I wasn't able to correct this in the title