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/)
-   -   Because of skewness (http://www.cfd-online.com/Forums/openfoam-solving/59714-because-skewness.html)

lam May 3, 2007 05:56

Hi all, I need some help fr
 
Hi all,

I need some help from you.

I use the simpleFoam solver with laminar flow, and turbulence off.
And my geomtries are converted by gmsh2ToFoam.
But each time, I make a computation on a complexe geometrie, checkMesh tells me that I have a lot faces which have a "Severe non-orthogonality" (but it's said finally: "Non-orthogonality check OK.") so I do not care (should I?).

But I really get worried about the rate of "Max skewness" which is about 2000% (and more for the other meshes I have)!! checkMesh says to me "This may impair the quality of the result."

And each time, my computation failed (the sum local, and the p initial residual went off!)

And of course, I can not improve the quality of my meshes before doing the conversion into Foam.

My question: is how can I try to use Finit Element discretisation, instead of Finite Volume discretisation in Foam? I think I need something like "fvSchemes" dictionnary but for Finite Element, isn't it?
(I do not know if it may improve my computation, but I really want to try to use it. )

Thanks in advance,

Lam


All times are GMT -4. The time now is 13:29.