CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Blow up of reynolds stress turbulence model (http://www.cfd-online.com/Forums/openfoam-solving/60112-blow-up-reynolds-stress-turbulence-model.html)

newbee July 17, 2006 04:39

Hi, I have a channel with a
 
Hi,

I have a channel with a incompresible flow and high Re number, stedy state. I run my case in a slightly modified version of the simpleFoam application. I want to try the two Reynold stress based turb. models LRR and LaunderGibsonRSTM. I use the following values for R()() in at the internal ande inlet patches of the channel.

inlet & internal
R=[0.564 0 0 0 0.564 0 0 0 0.564];
(Rxx=Ryy=Rzz=0.564)
which is the same as 2*k/3 in my case.

The ploblem is that my time step continuity error blows upp pretty quickly and the calculated value of epsilon follows.

With other turbulence models I experiance many itterations for pressure (500) but the outcome looks nice.

Does anyone have a sugestion on how to avoid the blow upp using Reynold stress based turbulence models?

Thanks
/Erik

barath.ezhilan July 25, 2009 11:22

Similar problem
 
Hey Newbee..

I am experiencing a similar problem here. Any idea how to get rid of the problem?? Did you find the solution to the problem??

svens November 24, 2009 15:09

Hi Erik and Barath

Did you try to use a k-epsilon calculation as initial condition for the RSTM run?

I have good experiences with precalc with an EVM, create the Reynolds-Stress-Tensor with post-processing tool 'R' and than switch to the RSTM. Be aware that the 'R' tool generates good values for the internal field but mixes up the boundaries! Therefore you have to correct the boundary field after operating 'R'.

I hope these information are still helpful. But perhaps you already figured out a way to fix your problems. In that case I would be happy to hear your oppinion about OpenFOAMs RSTMs. Dis they perform satisfactorily? Any further problems?

Thanks pretty much
Sven

natrask November 30, 2009 16:32

Relaxing R also helps a lot

i.e. add the following to your fvSolutions

relaxationFactors{
k 0.5;
epsilon 0.5;
R 0.5;
}

spej August 12, 2010 05:10

Problem with the R-tool
 
i got the problem.


All times are GMT -4. The time now is 21:58.