[Soll]

Greetings!


Recently I'm working on adding source term to rhoSimpleFoam.
I read through the source code and did some test, but still I have some problems unsolved. Please if someone could help:


1. Is it correct to add source term, i.e., mySource (volVectorField, constant but updated at every iteration) to the three locations?
a. In UEqn.H:

solve(UEqn == -fvc::grad(p)+mySource);

b. In pEqn.H, NonOrthogonal correction

fvScalarMatrix pEqn
        (
            fvc::div(phiHbyA)
          - fvm::laplacian(rhorAUf, p)
          + fvc::div(mySource*rAU)

          ==
            fvOptions(psi, p, rho.name())
        );

c. In pEqn.H, corrector

U = HbyA - rAU*fvc::grad(p) + mySource;

I saw a thread

https://www.cfd-online.com/Forums/openfoam/65293-add-source-term-simplefoam.html

related to this. Basically they are discussing adding source term to a. and c., but why not b.?


2. Why in pEqn.H phiHbyA is interpolated like this

surfaceScalarField phiHbyA("phiHbyA", fvc::interpolate(rho)*fvc::flux(HbyA));

instead of

surfaceScalarField phiHbyA("phiHbyA", fvc::flux(rho*HbyA));

like the way rhorAUf did, or they are the same.


3. Why rho*HbyA is interpolated to the cell surface in the first place? may be it is more of a algorithm question but I just cant figure it out.

[Soll]

I think I know the answer to question 3 if I'm correct

post here if anyone is interested:
for any volVectorField V

fvc::div(fvc::flux(V)) equals to fvc::div(V)

with additional options to specify the interpolation schemes