[Alczem]

Hey everyone,


I am running some reactions with reactingFoam, and I would like to add sink and source terms for several species.


The final goal is to mimic a semipermeable membrane with a custom transfer rate (different from the one available with semiPermeableBaffleMassFraction ) but in a cellzone. So how can we have a massflowrate of one specific species leaving the domain?



I have been trying to use scalarCodedSource in fvOptions to create a sink/source term:

source_mass
    {
        type            scalarCodedSource;

        name        source_h2;

        scalarCodedSourceCoeffs
        {
        
            selectionMode   cellSet;
            cellSet        membrane;
            fields  (H2);
            //name    sourceTime;

            codeInclude
            #{
            #};

            codeCorrect
            #{
            #};

            codeAddSupRho
            #{
                const scalarField& V = mesh_.V();
                scalarField& massSource = eqn.source();
                const labelList& cellIDs = cells();
        
                forAll(cellIDs,i)
                {
                    label cellI = cellIDs[i];
                    massSource[cellI] += -1e-2*V[cellI];
                }
            #};

            codeConstrain
            #{
            #};
        }

        sourceTimeCoeffs
        {
        }
    }

But this affects only the mass fraction in a weird way, and it does not affect the other quantities such as pressure, velocity, temperature etc. Is there another field I could use to change the mass (I tried rho, but I got an error "not implemented").



I have a very poor understanding of all the coded options available in OpenFoam, so I will gladly hear any advice

[Pappelau]

Hey what you want should be the mass source :

massSinkO2
{
    type            massSource;
    selectionMode   cellZone;

    cellZone       membran;
    massFlowRate    -0.1; // kg/s 

    fieldValues
    {
    H2O     0;
    O2         1;
    T         352.15;
    U       (0 0 0);
    }
 }

This is valid for at least OpenFOAM10&11. I think earlyer version have slightly different implementations
Before it was something like this:

massSource
{
    type            semiImplicitSource;
    active          true;
    selectionMode   cellZone;

    cellZone        membran;

    volumeMode      specific;
    sources
        {
         rho
         {
             explicit     -0.1;
             implicit     0;
         }
         O2
         {
             explicit     -0.1;
             implicit     0;
         }


}

Hope this is what u search for

[Alczem]

Thanks for your answer!


I did try massSource, but with OpenFoam v2306, and it raised me an error "unknown fvOption type" so I gave up and did not try with OpenFoam 11. I will give it a try then!


The scalarSemiImplicitSource is also an option, but it results in the sum of mass fractions being more than 1 in a cell, so I am probably using it wrong.


One more thing, what is the influence of the value for T in your massSource example? Since matter is leaving the domain, does it set a fixed value in the cellzone? or does it try to remove energy from the domain to match that temperature? I will try to find out by myself but if you have already some experience with it, I would be grateful for more information.

[Pappelau]

Hey,

when working with species sum of mass fractions over 1 is common problem this can be fixed with limiters in fvSchemes (limitedLinear01). Another Point is the all time topic of using good meshes . Further species solves better with an allready preconverged case (Fix species for the first iterations). Next thing i observed is that you should use GAMG for Yi (seems to be most stable).
The old implementation implicitSource lead to a temperatur increase if mass added and decrease if mass removed.

For the new one (massSource) i can not say to much as i didnt work much with it since it is released. But it will NOT fix the temperature to this value .