Hello everybody, I have a q
 

Hello everybody,

I have a question concerning source terms. First, I will explain you that I want to solve an equation with a source term in the calculation domain. I write the code like that:

double ttemp;

forAll(mesh.C(), i)
{
ttemp = pow(mesh.C()[i].y(),2)+pow(mesh.C()[i].z(),2);
if (mesh.C()[i].x() >= 2.384e-03 && mesh.C()[i].x() <= 2.634e-03 && Foam::pow(ttemp, 0.5) <= 3.000e-03)
{
Jmax[i] = 1.2e08;
Rayon[i] = Foam::pow(pow(mesh.C()[i].y(),2)+pow(mesh.C()[i].z(),2),0.5);
DeltaZ[i] = 0.25e-03;
}
else
{
Jmax[i] = 0.0;
Rayon[i] = 0.0;
DeltaZ[i] = 1.0;
}
}

/* Veqn solving */
tmp<fvscalarmatrix> VEqn
(
fvm::laplacian(sigma, V)
==
Jmax*exp(-2.080e03*Rayon)/DeltaZ
);

Veqn().solve();

Jmax, DeltaZ and Rayon are declared as volScalarField. The first calculations show great differences between the solution given by OpenFOAM and the solution given by our code. So my question is: in OpenFOAM, do we have to weight the source term by a jacobian or the cell volume like is done in some CFD code? I don't think so by viewing the programming the source term of several solvers in OpenFOAM. But I does not explain the differences I found...

I'm waiting for your explanations and your help. Thank you.

Frédéric LAGO