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/)
-   -   interFoam Stability Issues (https://www.cfd-online.com/Forums/openfoam-solving/149751-interfoam-stability-issues.html)

Dazzler March 7, 2015 17:18

interFoam Stability Issues
 
I am trying to simulate an outlet chamber using interFoam but no matter what combination of boundary conditions I choose - I seem to run into a stability issue after a varied amount of time into the simulation.

InterFoam decreases the time step continually until a floating point error occurs. I have based the case set up on the waterChannel tutorial file - but i have tried numerous alternative configurations of boundary conditions and not yet found anything that actually works.

The mesh was constructed using Salome and represents a fairly simple situation where two pumped high velocity inlets enter a small outlet chamber, the chamber is equipped with two simple gravity outlet pipes.

I have included a drop box link to the case files below.

https://www.dropbox.com/sh/m06gbu64j...aBwSqGcfa?dl=0

If anyone out there can shed any light on this - it would make my year.


Thanks

Svensen March 8, 2015 04:15

1) You mesh have some non-ortho faces, but you have no correctors in your fvSolution: "nNonOrthogonalCorrectors 0;"
2) It seems that you initial time step is too big, try to reduce it
3) Always check your mesh with checkMesh. In the output have to be no-errors or warnings.

Dazzler March 11, 2015 16:46

Stability Issues
 
I have tried increasing the number of orthoCorrectors and this helps the simulation limp on a little further.

But I still get the same problem little later in the process.

I have tried adusting the max Courant number and the tolerance in the Fvsolution file - but none of this actually seems enough to push this over the edge and develop a stable solution.

Anyone got any other ideas ?

Svensen March 12, 2015 02:04

Try to simulate with "adjustTimeStep no" and small value of deltaT.

Adjusting of the time step in some situation may not work correctly.


All times are GMT -4. The time now is 16:02.