Hi , right now I'm modelling 3D centrifugal fan. I had difficulties to determine what turbulence model which gives accurate and reliable results.

Any suggestions?

In my opinion, finding related papers or experimnetal data would be better.

P.S. Reynolds Stress Model is the best "generally".

goblok , ngono wae gak iso , dancuk koen, asu

Which language is John speaking in?

All the models work but they all suck.

It all depends on your problem, computer power, and desired results.

RSM is good, but I think LES is better if the mesh is fine enough. I think of LES as the next closest thing to DNS.

k-epsilon models and spallart allarmus also work well for certain flows.

What type of flow are you modelling, how much processor power and memory do you have, and what is your end goal?

(1). I think, you are right. (2).Using my definition of CFD, you will have to do "mathematical modeling". (3). And "Modeling" is not "select an option". You will have to fool around with your problem and the corresponding model(ing). (4). So, you don't expect to get accurate results by just picking an option. (unless it has been modelled, fine tuned and validated for your problem)