CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Exploding h/rho in rhoSimpleFoam (https://www.cfd-online.com/Forums/openfoam-solving/72763-exploding-h-rho-rhosimplefoam.html)

bastil February 17, 2010 10:41
Exploding h/rho in rhoSimpleFoam
 
Dear Solver-Experts,

I am running a case in rhoSimpleFoam and I get exploding values for temperature and especially density after few iterations, but no divergence. This even happens for turbulence turned off. I have to blame a few cells in my prism layer (low-Re, layer next to wall) for this.
I am trying to bound rho and h in some way, but without success so far. I tried:

Code:
interpolationSchemes
{
      interpolate(rho) limitedLimitedLinear phi 1.0 1.0 1.25;

...
}
to bound rho between 1.0 and 1.25 but this does not work at all. I get higer densities after 4 iterations... How can these values be bounded? Thanks.

Regards

olesen February 18, 2010 05:05
Quote:
Originally Posted by bastil (Post 246314)
Dear Solver-Experts,

I am running a case in rhoSimpleFoam and I get exploding values for temperature and especially density after few iterations
Have you tried severely underrelaxing rho? For example, a value of 0.05 might get you over the rough spots.

bastil February 18, 2010 05:53
Quote:
Originally Posted by olesen (Post 246427)
Have you tried severely underrelaxing rho? For example, a value of 0.05 might get you over the rough spots.
Thanks Mark. Yes I do this. I still have problems in some critical areas. I am also wondering about BCs at walls. We use for walls:
fixedValue, 1e-20 for k and epsilon (for LaunderSharma-k-epsilon)
zeroGradient for temperature
I am not sure for temperature... The only similar tutorial case is NacaAirfoil from turbFoam. This uses zeroGradient for temperature but also for k and epsilon which is obviously wrong?
Any other ideas? Thanks.

Regards BastiL


All times are GMT -4. The time now is 13:33.