[dschmidt]

Dear Foamers,

I'm trying to implement a passive scalar transport into multiphaseEulerFoam (2.2.x).
The scalar is only transported by one of the phases, so I need to access the corresponding phi.

I got a little confused by phases.phiAlpha(), which is used in the UEqns.H and is calculated within the multiphaseSystem
and phases.phi(), which is used for calculating the overall phi (Peqn.H) and is calculated as:

PEqn.H
...
 phase.phi() =
                    phiHbyAs[phasei]
                  + rAlphaAUfs[phasei]*mSfGradp/phase.rho();
...

Both seem to be described in the phaseModel.H, but that didn't help me either

        //- Volumetric flux of the phase
        surfaceScalarField phiAlpha_;

        //- Volumetric flux for the phase
        autoPtr<surfaceScalarField> phiPtr_;

My scalar transport equation with the two variations looks like this:

 fvm::ddt(alpha, Tracer)     
+ fvm::div(fvc::interpolate(alpha)*phase.phi(),Tracer)
 //+ fvm::div(phase.phiAlpha(), Tracer)  
 - fvm::laplacian(fvc::interpolate(alpha)*DTotal, Tracer)

I would be glad if anyone could clarify this for me.


Thanks,
Dominik

[kwardle]

If the name of the phase you are trying to access the phi for is "Tracer", then you need to define a few things to reference the correct phase.

const phaseModel* tracerPhase = fluid.phases().lookup("Tracer");
const phaseModel& phaseT = *tracerPhase;

then when you include phaseT.phi() in your transport equation, it will be the phi for the phase named Tracer.

Hope this helps,
Kent

[dschmidt]

Thanks for the advice Kent.

As the phase.phi() is the peqn result I also went with that instead of phiAlpha, but my current implementation isn't as neat as yours

I loop over the phases as in UEqn.H and solve the tracer transport equation only for the "tracer transport phase" defined in a tracerProperies dict.

The results look promising, but to keep the Tracer "bounded" to its transport phase I needed to use upwind div scheme for the tracer equation.