How about the S-A model for blunt boby flows?
 

Hi,I am using the S-A turbulence modle to compute the compressible flow around a cylinder. But I meet some problems. I found the initial field of eddy-viscosity has a impressive effect on the convergence. Also the eddy-viscosity computed from the S-A model is very small. Sometimes I can attain the convergence if I set the intial field of eddy-viscosity as the results from the k-e model. And It doesn't converge if I set the intial field as the freestream values which is very small(less than mu/10). Meanwhile,in some regions velocity may become ultra large or reverse and doesn't decrease, which leads to the divergence. I have used S-A model for the airfoil flow,and it seems to work well. I don't know if the S-A model is unsuitable for blunt body flows? Anybody has the experience on using S-A model, and I wonder if you can give me some suggestions. Thanks a lot!!! Maximus