[GregorAlan]

Hi all,

I am simulating a conjugate heat transfer problem with chtMultiRegionSimpleFoam. For this purpose I need to use the SIMPLE algorithm to solve the flow eqs. I am using a mesh imported from FLUENT (fully unstructured). Do I need to impose limiters in the fvSchemes or should I use nonOrthogonal correctors to account for the nonOrthogonal mesh?

Here is the checkMesh log:

Checking geometry...
    Overall domain bounding box (-0.015 -1.4083438e-17 -0.005) (0.03 0.01 0.02)
    Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
    Mesh has 3 solution (non-empty) directions (1 1 1)
    Boundary openness (1.0582429e-16 8.4474521e-16 -5.4978726e-16) OK.
    Max cell openness = 4.2135059e-16 OK.
    Max aspect ratio = 8.6457664 OK.
    Minimum face area = 5.4296002e-09. Maximum face area = 9.4163136e-07.  Face area magnitudes OK.
    Min volume = 2.3686376e-13. Max volume = 2.9847725e-10.  Total volume = 1.05e-05.  Cell volumes OK.
    Mesh non-orthogonality Max: 64.489763 average: 15.592416
    Non-orthogonality check OK.
    Face pyramids OK.
    Max skewness = 2.5141688 OK.
    Coupled point location match (average 0) OK.

Mesh OK.

And the fvSchemes for the fluid region:

ddtSchemes
{
    default         steadyState;
}

gradSchemes
{
    default         Gauss linear;
}

divSchemes
{
    default         none;
    div(phi,U)      bounded Gauss linearUpwindV grad(U);
    div(phi,h)      bounded Gauss linearUpwind grad(T);
    div(phi,K)      bounded Gauss linearUpwind grad(T);
    div(((rho*nuEff)*dev2(T(grad(U))))) Gauss linear;
}

laplacianSchemes
{
    default         Gauss linear corrected;
}

interpolationSchemes
{
    default         linear;
}

snGradSchemes
{
    default         corrected;
}

Regards,

Gregor Alan

[giammy92]

hi, you can give a look here to set a suitable fvSchemes: http://www.dicat.unige.it/guerrero/o...sandtricks.pdf
Cheers,
Gianmichele

[Kina]

Your snGrad already is a non-orthogonal correction for the computation. You can limit the values in case your computation diverges. In that case, I recommend to change div(phi,U) to:

div(phi,U)            bounded Gauss linearUpwindV linearLimited 1;