Convergence Problems

Carlos

I hope somebody can help me with this one.

I am currently modelling vehicle aerodynamics as part of my PhD and I am having problems in Fluent. The flow is coupled internal-external with some complex geometry in places and convregence is proving difficult. Strangely the standard k-e model gives convergence of 10^-6 for continuity but the Spalart-Allmaras model wont converge below 10^-3 and RSM oscillates all over the place!

I am using a 6 million cell hybrid volume and the quality of the cells is pretty good. I use a central volume of unstructured cells but the rest of the domain is structured. Within the central volume I have about 200 cells between 0.7 and 0.8 equiangle skew but I cant seem to get rid of them! I have been racking my brain for days and I cant work out what the problems are.

I have tried smoothing the mesh, converting the highly skewed cells to polhedra and relaxing the solution but to no avail Does anybody have any suggestions please? I really need some help with this one!

Thanks for your help,

Carlos.