Hi All,

I had coded Spalart-Allmaras turbulent model by using finite volume, pseudo time and LUSGS on Unstructred grid. When trying to validate it againest flat plate (Stanford case 1400), I found that the turbulent viscosity always decreasing. Does any one had experince with such problem. your suggestions is more than welcome.

What is the Reynolds number, or incident velocity ?

I think that S-A turb model is good only for high Reynolds number applications, especially aerodynamics. For general purposes, I would advice using a very standard k-epsilon model.

Thanks alot Lionel for your reply, Re = 4.4e6 velocity = 33 m/s,

About your advice using k-epsilon model, I already coded S_A model and I thank K-epsilon model is much more complicated especially for unstructured grid. Also K-epsilon has its problems. So that, I want to compromise between numerical complexity and numerical efficiecny.