[theoME]

Hello

in

OpenFOAM-3.0.1/applications/solvers/heatTransfer/chtMultiRegionFoam/solid/solidRegionDiffNo.C

the calculation of the solid diffusion number is given, which is basically:

DiNum = alpha * deltaT * deltaCoeffs()

with the thermal diffusivity alpha=kappa/rho/cp. If DiNum should be dimensionless, deltaCoeffs() should have the dimension 1/m^2.

The calculation of deltaCoeffs() is given in:
OpenFOAM-3.0.1/src/finiteVolume/interpolation/surfaceInterpolation/surfaceInterpolation/surfaceInterpolation.C

and discussed in
http://www.cfd-online.com/Forums/ope...ltacoeffs.html

From that I expect the dimensions of deltaCoeffs() to be 1/m, so that an additional 1/m is missing somewhere (if DiNum is expected to be dimensionless).

I would really appreciate if someone could shed light on this issue.

Best regards

Florian

[bikooo3878]

It's been so many years but I'm having the same question Have you found an answer?
Also, I'm under the impression that delfaCeoffs() is dimensionless; referring to basicFvGeometryScheme.C, line 156:
"Constructing differencing factors array for face gradient".

The calculation of DiffNo is given as

surfaceScalarField kapparhoCpbyDelta
(
sqr(mesh.surfaceInterpolation::deltaCoeffs())
*fvc::interpolate(kappa)
/fvc::interpolate(Cprho)
);

const scalar DiNum = max(kapparhoCpbyDelta).value()*runTime.deltaTValue ();

So what is the square root of deltaCoeffs() ?

[kanata]

As I know, in OpenFOAM, sqr(x) == x^2. So the dimension is correct.