[pod]

Hello everyone,
I'm trying to use a continuity equation in chtMultiRegionFoam that looks like it (similarly to the one used in cavitatingFoam):

fvScalarMatrix rhoEqn
    (
        fvm::ddt(rho)
      + fvm::div(phi2, rho)
      ==
        fvOptions(rho)
    );

which basically has the second term implicit instead of explicit like in normal heat transfer solvers.
I get this known problem:

valueInternalCoeffs cannot be called for a calculatedFvPatchField
    on patch ...
    You are probably trying to solve for a field with a default boundary condition

I know this comes out if you set "calculated" b.c. type but I think mine should be fine:
in the 0 folder:

dimensions      [1 -3 0 0 0 0 0];

internalField   uniform 147;

boundaryField
{
    ".*"
    {
        type            zeroGradient;
    }
}

In the fluid part changeDictionary:

    rho
    {
    internalField   uniform 147;

    boundaryField
    {
        ".*"
        {
        type            zeroGradient;
                value           uniform 147;
        }
    }
    }

does anyone have any idea?
thanks a lot

[Tobi]

Hi,

the density field should be set automatically to calculated as the boundary values are "calculated" from other values. It is not a field like a passive scalar and I don´t think that your density field has zeroGradient patch types - even though you defined it. To check it, just plot rho before your continuity equation.

Furthermore, your equation is a transport equation and not a continuity equation:



This equation looks more like an incompressible transport equation for the density.

[pod]

Hello Tobi,
thanks a lot for your reply.
In reality phi2 is not equal to the definition of phi in heat transfer solvers (rho*U), but it is defined like:

surfaceScalarField phi2 = (fvc::interpolate(HbyA) & mesh.Sf())
         + rhorAUf*fvc::ddtCorr(U, linearInterpolate(U) & mesh.Sf()) - rhorAUf*mesh.magSf()*fvc::snGrad(p_rgh)/rhof;

in order to resemble the phi of cavitatingFoam, which is actually more similar to phiHbyA of chtMultiFoam. I forgot to mention that, sorry. So it is a continuity equation, right?


Also, the initialization of rho in cavitatingFoam's tutorials is similar to mine:

dimensions      [1 -3 0 0 0];

internalField   uniform 845;

boundaryField
{
    inlet
    {
        type            fixedValue;
        value           uniform 845;
    }

    outlet
    {
        type            fixedValue;
        value           uniform 835;
    }

    walls
    {
        type            zeroGradient;
    }

    frontBack
    {
        type            zeroGradient;
    }
}

that's why I don't understand what's wrong.
thanks

[mAlletto]

Originally Posted by pod

Hello everyone,
I'm trying to use a continuity equation in chtMultiRegionFoam that looks like it (similarly to the one used in cavitatingFoam):

fvScalarMatrix rhoEqn
    (
        fvm::ddt(rho)
      + fvm::div(phi2, rho)
      ==
        fvOptions(rho)
    );

which basically has the second term implicit instead of explicit like in normal heat transfer solvers.
I get this known problem:

valueInternalCoeffs cannot be called for a calculatedFvPatchField
    on patch ...
    You are probably trying to solve for a field with a default boundary condition

I know this comes out if you set "calculated" b.c. type but I think mine should be fine:
in the 0 folder:

dimensions      [1 -3 0 0 0 0 0];

internalField   uniform 147;

boundaryField
{
    ".*"
    {
        type            zeroGradient;
    }
}

In the fluid part changeDictionary:

    rho
    {
    internalField   uniform 147;

    boundaryField
    {
        ".*"
        {
        type            zeroGradient;
                value           uniform 147;
        }
    }
    }

does anyone have any idea?
thanks a lot

The error you posted usually appears if the boundary condition is unknown. Did you try to remove the value under zeroGradient?