In my opinion, the RNG k-ε model is better. This is due to the fact that better allows to describe a swirling flow of the fluid. Even more accurate model can be Reynolds Stress Model. This model can accurately reproduce the turbulent stress of a highly anisotropic behavior of the fluid (occurring in cyclones).

Hi Mothilal--

I have used CFD to model both the RNG k-ε and RDM models in various swirling flows. The most important thing is to obtain some reliable data to compare and validate with your CFD model. Otherwise, you are merely doing CFD to create a publication, and not to obtain any useful results.

Remember, garbage in, garbage out.

Best,

--Keith

KeithDPatch