[despaired student]

Hi,

could someone help me to interprete the attached logfile.

Information about my mesh, solver:
1) It is created with ICEM 14.0 and transferred to OpenFoam with "fluent3DToFoam"
2) There are some realy, realy bad cells.
3) Mesh consists of about 23 million cells. Some cells are very fine - 0.05 millimeters.
4) Solver is simpleFoam


My OpenFoam batch-script:

#!/bin/bash
#PBS -N first_mesh
#PBS -l walltime=150:00:00
#PBS -l select=3s=rhel:ncpus=6
#PBS -M soeren.schillerwein84@gmx.de
#PBS -m abe
source /nfs/rzpool/OpenFOAM/OpenFOAM-2.0.1/etc/bashrc
cd /apistemp/sm1/ss/Meshs/First_Mesh
foamUpgradeFvSolution
foamUpgradeTurbulenceProperties
fluent3DMeshToFoam ICEM_CRAP_33.msh -scale 0.001
decomposePar
mpirun -np 18 foamExec simpleFoam -parallel > Log1
reconstructPar -latestTime
foamCalc mag U
sample
foamLog Log


To run checkMesh I used the command: "checkMesh > checkMeshLog33"


OpenFoam is not calculating this mesh. Could the reason be that the mesh is too bad? Or where do you see the problem?

[olivierG]

hello,

Your mesh is realy realy bad, you will not be able to get any correct results with, so change the mesh.

regards,
olivier

[despaired student]

...I was afraid that that's the reason...

Thanks for the answer Olivier

[despaired student]

Hi,

if one runs checkMesh and the answer is not "Mesh OK" is it impossible for OpenFoam to use this mesh? Or are there exceptions?

[olivierG]

hello,

Basically, yes.
... but sometimes, like with max orthogonality at 80 instead of max 70, it's fine.
bad mesh lead to bad and wrong result ... and you don't want that.

NB: if you get a good mesh with commercial software, which become very bad in openfoam, then you may have a problem with the way you import mesh.

regards,
olivier