CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   Problem with convergence of K-epsilon/LRR model (

bgraves May 20, 2013 19:26

Problem with convergence of K-epsilon/LRR model
Hi Foamers,
I am running an analysis looking at the boundary layer of flow over a two-dimensional bump. I am currently having trouble getting convergence for Launder-Reece-Rodi model, after about a day of running the residuals blow up. I am currently using the following Boundary Conditions:

U fixedValue
P fixedValue
k/epsilon fixedValue
nut calculated

U zeroGradient
P zeroGradient
k/epsilon zeroGradient
nut zeroGradient

U fixed Value 0 0 0
P zeroGradient
k kqRWallFunction
epsilon epsilonWallFunction
nut nutkWallFunction

I tried to use the method found on this site for calculating the value for Reynolds stress, and got that far. If this helps my colleague ran the same analysis using StarCCM+ and his simulation worked, and his results were good.

If anyone has any suggestions of how to improve my simulations it would be greatly appreciated.


Tobi May 21, 2013 09:46


did you have a look at the last timesteps. It should be possible to see the Problem zones there. Maybe you have a Problem in a cell or at a BC (outlet for example).


bgraves May 21, 2013 11:41

Hi Tobi,

Thanks for the speedy reply! I did check, by looking in paraFoam and the velocity goes to infinity by the 20th iteration.


Tobi May 21, 2013 12:23

Then you ll have a problem with your BC i think :)

Can you post a screenshot of that region?

Did you try




bgraves May 21, 2013 12:49

2 Attachment(s)
That did get the code to run however now my residuals are terrible, and the results are bogus. The nice thing is that now the code runs the problem must be in the solving methods right? Since I know the grid is good. Attached is the system and 0 file for my problem if you wouldnt mind looking at it when you have a chance I would greatly appreciate it. I would like to say thank you for helping me!

Tobi May 21, 2013 12:54

would it be possible to upload or send me the whole case?

bgraves May 21, 2013 13:12

I could try to send you the file the constants folder is to big to upload to this website. where should I send it

All times are GMT -4. The time now is 20:53.