I just want to know whether anyone has done transition modeling for separated flows using Low-Reynolds number K-W in CFX 4.2. I tried doing it but the model is not designed to handle transition problem. Some of the coefficient in the source term have to be modified ( Wilcox 1998). This can be done using USRSRC but I just want to know whether anyone has tried doing it successfully for separated confined flow problems. Thanks Athar

Hi Athar,

The k-w model in 4.2 is not the 1988 (i.e. old) Wilcox model. It has w specified differently - making the closure coefficients a function of derivatives of w then becomes problematic. It also has a low-Re function "fudge" which is something Wilcox claims the k-w models dont need.

To implement the 1998 model (without transition), in 2d axissymetric form, I had to write a USRSRC and USRVIS. I'll have to ask my supervisor if I can make them avaliable - I see no problem, but I will have to ask.

I have been looking at turbulent jets, and whether the Pope style and cross diffusion corrections do what they are sposed to in elliptic flow solvers. A copy of a paper is avaliable at http://go.to/rmorgans if you are interested.

I'll get back to you soon about the routines... maybe email me off the list. If you do get something working, let Dave Wilcox know your results - he is always interested!

Best regards

Rick Morgans

__________________________________________________ ___

Rick Morgans Postgrad Student CAE Consultant TEC Group Proen Design Aust Dept. Mech. Eng. 1/157 Burbridge Road Uni of Adelaide 5005 Hilton SA 5033 Ph: +61-8-8303-3157 Ph: +61-8-8152-0800 Fax: +61-8-8303-4367 Fax: +61-8-8152-0888 __________________________________________________ ___