[kar]

Hi!

U = rUA*UEqn.H();
phi = (fvc::interpolate(U) & mesh.Sf())
+ fvc::ddtPhiCorr(rUA, U, phi);

What fvc::ddtPhiCorr(rUA, U, phi) is there for?

if (nonOrth == nNonOrthCorr)
{
phi -= pEqn.flux();
}

And what ".flux()" is returning?

Kārlis

[schmidt_d]

Karlis,

The .flux() method takes the face fluxes that come from the off-diagonal terms of the equation, in this case pEqn, and returns them directly. An easier-to-understand example would be if I were doing the continuity equation, where rhoEqn is:

fvm::ddt(rho) + fvc::div(phiv,rho)

where phiv is the velocity interpolated to cell faces and dotted with the face normals.

after solving, rhoEqn.flux() would give you the mass flux out of each cell face. Similarly, pEqn.flux() gives you the momentum flux due to the fvm:: terms of the pressure equation.

-DPS
ps. The ddtPhiCorr is a mysterious term that attempts to correct for temporal error in the flux...can't help you there.

[kar]

Good evening David!

>.flux() method takes the face fluxes that come from the off-diagonal terms of the equation

Why it's necessary to distinguish between matrix diagonal and off-diagonal elements?

What papers and www would you suggest to understand this kind of things?

K.

[jaswi]

Hi Karlis

Matrix diagonal elements are the elements for which you are seeking solution and off-diagonal elements are those which will effect this solution.

Checkout the basic Finite Volume, SIMPLE, PISO formulations in this book:

1) An Introduction to CFD: The Finite Volume Method by Versteeg, Malalasekera.

2) CFD by Anderson

With Best Regards
Jaswi