CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Computational cost of transient turbulence model (

costinruja March 8, 2012 13:55

Computational cost of transient turbulence model
Good evening everybody,

I've recently began playing with some simulations of ahmed body in steady and transient conditions.For the steady state simulation i have been using 3 mesh densities and at the last case i had some problem with the convergence.This sounds to me like i managed to get past the dissipation limit of the SST model and the flow is flapping wild.So to get used to transient simulations and not to lock my computer the whole weekend i need some info on the computational expense of SAS SST, DES and LES (the main 3 turbulence models i understood are recommended for transient flow around bluff bodies).



ghorrocks March 8, 2012 18:10

You are correct in that the turbulence models you should consider are SAS, DES and LES.

The cost of these approaches depends on the model - the size of the turbulent structures they must model are dependant on the flow Re number, and the mesh must be adequate to resolve this so the mesh density will get higher as the Re number goes up.

But the SAS and DES approaches are much cheaper than LES as they only apply the LES approach in a subsection of the domain. Full LES does it everywhere and is obviously more expensive.

All times are GMT -4. The time now is 13:05.