[MMRC]

Hello community,
Can you please give me your opinion in this implementation of 'retarding shear stress' proposed by Richard & Hoxey (1993) in the row of cells adyacent to wall that helps to minimize the over production of TKE and so the peak of TKE when plotting against vertical profile: shear = rho * (u_friction)²
I'm running simpleFoam, must I remove or keep rho (density) in the calculation of the source?
By using topoSetDict we can catch the first row of cells adyacent to ground patch (wall): first patchToFace and then setToCellZone.

Then, by using fvOptions codedVectorSource:

PHP Code:

retardingShear{    type            vectorCodedSource;    selectionMode   cellZone;    cellZone        areaSet;  // name of my cellZone    fields          (U);    writeToFile     true;    name            retardingShear;        codeCorrect        #{        #};    codeConstrain    #{    #};        codeInclude        #{        #};    codeAddSup            #{                vectorField& retardingShear = eqn.source();  // will asign calculation to source side of fvMatrix         vector Ufriction (-0.57 0 0); // is the friction velocity near wall         vector diskRetardingShear (-1 0 0); // is the direction opposite to normal wind flow                const labelList& cellIDs = cells();                forAll(cellIDs, i)                {                    label cellI = cellIDs[i];                            retardingShear[cellI] += mag(Ufriction)*diskRetardingShear; // the left side with += is thought to be 'a sink', AND the right side is the value of shear with its current direction                }} 


Is it good practice to combine '+=' (sink assigment) with the current direction of this retarding shear ?
I've seen that some user multiplies the sources term by cell volumes, when is this needed?
I apologize but I lost the original code and this code is what I remember that I could run.
Any ideas are welcome