[ethanw]

Im trying to simulate the flow through a S-Duct intake with an inlet Mach of 0.8 and a Re of approx 1.4x106.
I created the mesh ensuring that skewness < 0.8, max included angle is approx 145 and that the boundary layer cells are created from a structured mesh to maintain wall orthogonality. The initial cell spacing in the boundary layer is approx 1x10-6 which gives a y+ of less than 1.
From what I have read online these should all be fine for Fluent.
For the first part of the simulation I use inviscid flow and first order for all discretizations. URFs are set to default. I have also tried this with constant density and without constant density. However, the simulation diverges after about 35 iterations.

Is there any additional settings I could try that will help the solution to converge?