[kumar2]

Hello Friends

I was wondering if there is a mistake in the PISO loop for the interFoam , rasInterFoam solvers.

In Dr. Rusche's PHD page 148,

(Del.V) = 0 , but at the interface density is rapidly changing so shouln't (Del.(rho*V)= 0);

Looking at the PISO loop in rasInterFoam we have laplacian(rUAf,pd)==div(phi) .

should this be changed to something like

laplacian(rUAF*rho,pd) == div(phi*rho)

Thanks a lot in advance

Kumar

[kumar2]

Dear Friends

Could someone comment on this post ?

thanks

kumar

[eugene]

No, because the interFoam solver conserves volume not mass (both fluid are assumed incompressible). This is done specifically to get around the problem of large gradients at the interface.

[kumar2]

Hi Eugene

Thanks a lot for your reply. I have just one more question.

In equation of 4.28 of henrik rusche's phd thesis we have phi=phi* - (1/A_D)_face * |S|*faceGradient(P)

But in 4.30 we have something like

div( (1/A_D)_face , gradient(P) ] = diver.(phi*)

How is |S|*faceGradient(P) converted to gradient(P) ??

In all openfoam solvers the equation 4.30 is implemented and they are dimensionally correct.

Another question if rUA = 1.0/UEqn.A()

is rUAf = rUA interpolated on the face and multiplied with surface area or it is just rUA interpolated to the faces

Thanks a lot once again

Kumar