CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Turbulent viscosity ratio (

Jurgen January 28, 2003 12:40

Turbulent viscosity ratio
I am using the standard k-e turbulence model in a quite large grid, about 1 mio. cells. I cannot improve the number on grid cells anymore. During iteration, I get the message that the turbulent viscosity is limited to a viscosity ratio of 1E+5. I know how to change this limit, this is not the point, more how to deal with this problem. Besides, the solution is converging very (!) slowly. Any ideas someone?

Regards, Jürgen

Erwin January 28, 2003 15:56

Re: Turbulent viscosity ratio
Some thoughts:

1) I don't know what a realistic limit for the ratio is. In the past there was some discussion whether such high turbulent viscosities were realistic or not. People disagreed. I guess it depends on what kind of flow problem you have, two-phase or single-phase, material properties etc. What i am trying to say: sometimes a ratio of 1e5 might be realistic ...

2) If you are sure it is not realistic you could try to fool the solver and set the limit to a lower value that is in the right ballpark, something like 5000 or so. That could force better convergence behavior.

3) Also, this message could indicate a mesh problem. Either something simple, like you forgot to scale the grid. Or something more problematic like cell skewness too high.

4) Last but not least: initialize with good values, also for k and epsilon. This can make a huge difference. If the flow field has been initialized poorly it can take ages for the solver to get out of it (as you may be experiencing). Reinitializing can save a lot of time.

Jurgen January 29, 2003 03:23

Re: Turbulent viscosity ratio
Right, skewness, I have a few cells with skewness arounf 0.8. Couldn't get rid of tose without increasing the grid size even further. Thanks for the input, Jürgen

All times are GMT -4. The time now is 12:12.