CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Convergence criteria (http://www.cfd-online.com/Forums/fluent/67405-convergence-criteria.html)

PrandtlGlawert August 13, 2009 18:12

Convergence criteria
 
Dear all:
I´m modelling a WORTMANN profile and studying its behaviour (Re=3x10^5). K-w (SST) model seems to reach convergence quick, but when switching to Reynolds Stress turbulence modelling, convergence tends to take far more time. That wouldn´t matter if it weren´t because i can never get a single value, just a range of them, and i can only take a semivalue...Is that correct? Any ideas to reach convergence? :confused::confused:

(By the way, i have to study later the same model with k-kl-w turbulence modelling and i´ve run a few cases... OMG convergence seems impossible there!!!! :D:D:D)

Thx in advance...

tstorm August 13, 2009 19:06

The problem with the RSM is the tight coupling of the Reynolds Stress transport equations. It's very difficult to get that model to converge. I don't know what a wortmann profile is, but if it involves strong pressure gradients, separation, transition, or other complicated features it might be tough. In my experience, the best way to get RSM cases to converge is to start with a very nice mesh - smaller cells than you'd use with an eddy viscosity model and more attention to skewness.
Good luck!


All times are GMT -4. The time now is 05:29.