CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   SST k-omega convergence (http://www.cfd-online.com/Forums/fluent/48455-sst-k-omega-convergence.html)

 mauri June 9, 2008 12:09

SST k-omega convergence

Hi! I'm simulating a diffuser augmented wind turbine using a SST k-omega model. Residuals of continuity, turbulent kinetic energy and velocity decrease rapidly, but residuals of omega are stable (I mean really fixed on a constant value) on a higher value. Why? What can I do? I've got no problems with k-eps models (and the same mesh)

 Wiebe June 16, 2008 06:59

Re: SST k-omega convergence

Hello Mauri,

There are a few things that you can try to do. What works really depends on your situation (mesh, steady sate or transient simulation, boundary conditions, etc.)

- try to do some iterations while solving the turbulence equation only.

- lower the relaxation factor for omega, or, if it is really low, try increasing it.

Good luck! Wiebe.

 Wiebe June 20, 2008 08:45

Re: SST k-omega convergence

Hello Mauri,

Ironically I had the same problem the past few days. I didn't want to look up turbulence levels for the input boundary condition, so I set it to zero until i would. It turns out that the SST k-omega model doesn't like 0 for both k and omega as inlet boundary condition.

Hopes this helps for you too.

Wiebe.

 mauri June 20, 2008 13:37

Re: SST k-omega convergence

Hello Wiebe,

I tried to run some simulations and thanks to your advice things are better, in terms of convergence rate. And of course another problem has arised: when I refine mesh in order to use enhanced wall functions with y+<5 (using adapt menu in fluent, both in conformal mode and in hanging node), then I get a non-physical solution, even starting iterations from a previous solution (which seems correct)!

Sorry for still bothering you...and thanks a lot!!!

 All times are GMT -4. The time now is 18:13.