CFD Online Discussion Forums - Non-orthogonal correctors

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- OpenFOAM Pre-Processing (https://www.cfd-online.com/Forums/openfoam-pre-processing/)

- - Non-orthogonal correctors (https://www.cfd-online.com/Forums/openfoam-pre-processing/84834-non-orthogonal-correctors.html)

Non-orthogonal correctors

Okey, lets see how i should explain this. I have a mesh that is (according to my previous knowledge ok regarding orthogonality) but I still want to be sure that the solver dont make to large error on a different mesh, I need therefore an implementation of a non-orthogonal corrector.

Easiest seems to be to use the PISO or SIMPLE loop, which i read corrects for non-orthogonality (http://openfoamwiki.net/index.php/Th...hm_in_OpenFOAM, ... The steps 4 and 5 can be repeated for a prescribed number of time to correct for non-orthogonality. )

The strange thing is that inside the fvSchemes at the "laplacianSchemes":

laplacianSchemes
{
default none;
laplacian((****)),p) Gauss linear corrected;
}

There is something called "Gauss linear correced", which according to the UserGuide means "Explicit non-orthogonal correction". I know also that there are other choices. At the end this leads to my question; is this correction enough or do I have to implement some kind of PISO/SIMPLE loop for correction? It confuses me since it seems to occur some kind of correction already!

Hi Johan,

In fvSolutions you can actually define the number of (non-orthogonal) correctors, so it's probably separated from the schemes like that. in FvSolutions you might have something like

Code:

PISO

{

    nCorrectors     4;

    nNonOrthogonalCorrectors 2;

    pRefCell        0;

    pRefValue       0;

}

Quote:

Originally Posted by gwierink (Post 294598)

Hi Johan,

In fvSolutions you can actually define the number of (non-orthogonal) correctors, so it's probably separated from the schemes like that. in FvSolutions you might have something like

Code:

PISO

{

    nCorrectors     4;

    nNonOrthogonalCorrectors 2;

    pRefCell        0;

    pRefValue       0;

}

Actually I dont have something like that, thats because I havent implemented any PISO/SIMPLE loop at this moment. If i refer to my question i asked more whats the difference between the correction that occurs in the PISO loop compared to the one that seems to occur in the "laplacianSchemes". If there arent any difference, is the PISO loop even necessary?

Quote:

Originally Posted by magjohan (Post 294660)

Hi Johan,

did you ever get an answer to this??? i am trying to understand the same thing, although there still doesnt seem to be much around in terms of an explanation ...

best regards
jonathan

Quote:

Originally Posted by magjohan (Post 294660)

Good Question
I am also struck on to it. Did you find the answer for your query?

Hi guys,

as I understand (and I hope, I do not tell you something wrong ;) ):

In fvSolution you define, that you want to use the PISO iteration, the amount of used iterations, ...

In fvSchemes you have to define additionally, on which quantities or terms you want to apply the correction.

So you have to define the nonOrthoCorrection in both files to properly work. Does this answer your question?

Best regards
Julian