[Onim]

Thanks for visiting this thread.

I am working on coding numerical algorithm for 3D Euler equation with general coordinate. Since the flow is compressible, I need to utilize upwinding.

After I implement AUSM+ and MUSCL for upwinding/flux limiting, I tried to solve Sod shock tube problem, and got an errorneous result. Rarefaction wave and shock wave are well refined, but contact discontinuity is oscillating.

If I use 1st order upwind difference for numerical flux, I can get a proper solution. I tried to figure out any bug in MUSCL code, but it was unsuccessful.

In this case, which part of the code should be reviewed?