[virginie_e]

Hello,

I have a question concerning the SIMPLE loop in interTrackFoam, especially about the velocity equation.

In interTrackFoam, UEqn is implemented this way:

tmp<fvvectormatrix> UEqn
(
fvm::ddt(rho, U)
+ fvm::div(phiNet, U)
- fvm::laplacian(mu, U)
);

UEqn().relax();

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


whereas in other solvers using the SIMPLE algorithm, it is implemented this way:

tmp<fvvectormatrix> UEqn
(
fvm::ddt(U)
+fvm::div(phi,U)
-fvm::laplacian(nu,U)
);
UEqn.relax();
solve (UEqn == -fvc::grad(p));


I understand that the difference is that the interTrackFoam equation is the second one multiplied by rho, but in that case, should the second term -fvc::grad(p) not be multiplied by rho as well? so that the equation should become something like:

tmp<fvvectormatrix> UEqn
(
fvm::ddt(rho, U)
+ fvm::div(phiNet, U)
- fvm::laplacian(mu, U)
);

UEqn().relax();

solve(UEqn() == - rho()*fvc::grad(p));

or something like that?

If what I said is wrong, is there something that I misunderstood in the equation solving in OpenFOAM?

Otherwise, how should I write the equivalent of the line

solve(UEqn() == - rho()*fvc::grad(p));

?

Thank you in advance

[ngj]

Hi Virginie

If you look at 0/p in the hydroFoil test case, then you will see that the pressure has dimensions which differ from those in for instance simpleFoam-tutorials.

Thus rho is incorporated in p.

I hope it did clarify you doubts.

Best regards,

Niels

[virginie_e]

Thank you Niels.

Indeed, I had not pointed out that the p dimensions was different. Thank you for your answer, it helped a lot!

Virginie

[virginie_e]

Hi,

just a quick message to say that I resolved my problems. This high pressure is not a divergence, it is the real result. My problems come from the fact that the cells of my mesh become very flat.

Virginie