[eespi002]

Hi,

I am doing a 3D model with turbulence ( Realizable k-epsilon model) and I am using default Under Relaxation factors for the first 100 iterations. Then at 101 iteration mark, I switch to lower Under relaxation factors to allow the solution to converge. However, when I switch to lower URF number, the solution proceeds to diverge.

Can anyone explain why this is happening? How many iteration should I allows the solution to go through for a 3D model?

Any response is appreciated.

Pleas see link:
http://i1009.photobucket.com/albums/...onvergence.jpg

[-mAx-]

it diverges (you get an error message) or you have a jump?
The jump is normal, if you let iterate, you should retrieve the convergence

[eespi002]

Originally Posted by -mAx-

it diverges (you get an error message) or you have a jump?
The jump is normal, if you let iterate, you should retrieve the convergence

Thanks -mAx-.

I did get an error message. It states:
Turbulent viscosity limited to viscosity ratio of 1.0000000e+6 in 1cells.

Could this have a large impact on the convergence? Also, typically how many iterations should be given to a 3D-turbulent model?

Thanks in advance.

[-mAx-]

this is not an error message, but a warning.
Let iterate, and check if the number of cells involved in this warning is decreasing (in your cas you just have 1 cell )
There is no typical number of iterations untill convergence.
Monitor a variable you want to compute (drop pressure, force , massflow or whatever) and iterates untill you get a flat curve. Then it means that this has converged.
Check also the residuals charts

[eespi002]

Thanks again.