K-w implementation

ganesh

Dear Friends,

I jus implemented K-w turbulence model in my 2D solver. The implementation is in terms of logarithmic variables( following Illnca etal.). This helps to guarantee positivity without any undue limiting on k and w. I am able to get the right results on coarse grids, but on finer grids I have problems in reproducing the right skin friction distribution, although I seem to get the pressure distribution ok. I use implicit time stepping with constant reconstruction for turbulence variables and linear reconstruction for the rest. I have directly implemented the Wilcox model, without any further modifications. If anyone has had experience in this, could you please tell about the influence of the freestream values on the solution and of any other problem on fine grids. Is it also really necessary to get a good initial guess for the probelm on fine grids ?

Any suggestions in this regard are most appreciated.

Regards and Thanks in advance,

Ganesh