CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   k-e or RSM? (http://www.cfd-online.com/Forums/fluent/40574-k-e-rsm.html)

 kevin April 14, 2006 05:43

k-e or RSM?

i'm simulating turbulent flow over a circular cylinder, then which model is more appropriate? again, standard or non-equilibrium wall function is better for this kind of problem.

any kind of suggestion is sincerely appreciated.

 Mirek April 14, 2006 07:34

Re: k-e or RSM?

RSM is better than RNG k-epsilon + non-equilibrium wall function but I prefer k-omega model with enhanced wall treatment called Transitional Flows + Shear Flow Corrections. I tested all turbulence models with all available additional options in Fluent 6.1.22 seeking static pressure distribution on cylinder surface (unsteady flow of course)at mean velocity w=8-32m/s. The turbulence intensity was i=1%, time step about 5e-5s to 5e-6s, y+=1 for enhanced wall treatment (the best for flows with boundary layer separation)and obtained relative error (for static pressure at the back of cylinder) equals to -2.6%. RSM is also good for example RSM + wall bc + non-equilibrium gives relative error (for w=20m/s) equals to -3.1%. I hope it helps you.

 kevin April 14, 2006 08:14

Re: k-e or RSM?

thanks for your valuble and detailed reply! it is very helpful of course. by the way, what's the range of ur Re? mine is less than 1e4 right now, but wider is hoped.

 Mirek April 14, 2006 10:02

Re: k-e or RSM?

Hello Kevin, it is very easy to calculate. The diameter of my cylinder equals to 12mm so for air Re=6.57e+3 to 2.63e+4. I checked numerical simulation results by myself in my open wind tunnel of low speeds so you can be sure of value of relative errors quoted in my previous message. If you want more information send me e-mail and I'll try to give you more details concerning flow over cylinder. Best regards

 All times are GMT -4. The time now is 20:17.