[Luke99]

Disclaimer: probably already discussed somerwhere can´t find anything though, link would be enough


So I would like to add a electrostatic force term

- fvc::interpolate(rhoE)*fvc::snGrad(Ue) - scalar(0.5)* fvc::interpolate(EE)*fvc::snGrad(eps)

to the UEqn in interfoam and I looked at examples online. Some of them only seem to add it in the momentum predictor and others only in the fvVectorMatrix UEqn part or do i only need to add it in fvoptions or in both? Obviously not fvoptions and as a term in the same equation.
I know that if I add it to the momentum predictor only it needs to be turned on.

MRF.correctBoundaryVelocity(U);

    fvVectorMatrix UEqn
    (
        fvm::ddt(rho, U) + fvm::div(rhoPhi, U)
      + MRF.DDt(rho, U)
      + turbulence->divDevRhoReff(rho, U)


      to add here?



     ==
        fvOptions(rho, U)
    );

    UEqn.relax();
    
    volScalarField EE = magSqr(fvc::grad(Ue));

    fvOptions.constrain(UEqn);

    if (pimple.momentumPredictor())
    {
        solve
        (
            UEqn
         ==
            fvc::reconstruct
            (
                (
                    mixture.surfaceTensionForce()
                  - ghf*fvc::snGrad(rho)
                  - fvc::snGrad(p_rgh)
                  

                   or here?



                ) * mesh.magSf()
            )
        );

        fvOptions.correct(U);
    }

[alundilong]

Originally Posted by Luke99

Disclaimer: probably already discussed somerwhere can´t find anything though, link would be enough


So I would like to add a electrostatic force term

- fvc::interpolate(rhoE)*fvc::snGrad(Ue) - scalar(0.5)* fvc::interpolate(EE)*fvc::snGrad(eps)

to the UEqn in interfoam and I looked at examples online. Some of them only seem to add it in the momentum predictor and others only in the fvVectorMatrix UEqn part or do i only need to add it in fvoptions or in both? Obviously not fvoptions and as a term in the same equation.
I know that if I add it to the momentum predictor only it needs to be turned on.

MRF.correctBoundaryVelocity(U);

    fvVectorMatrix UEqn
    (
        fvm::ddt(rho, U) + fvm::div(rhoPhi, U)
      + MRF.DDt(rho, U)
      + turbulence->divDevRhoReff(rho, U)


      to add here?



     ==
        fvOptions(rho, U)
    );

    UEqn.relax();
    
    volScalarField EE = magSqr(fvc::grad(Ue));

    fvOptions.constrain(UEqn);

    if (pimple.momentumPredictor())
    {
        solve
        (
            UEqn
         ==
            fvc::reconstruct
            (
                (
                    mixture.surfaceTensionForce()
                  - ghf*fvc::snGrad(rho)
                  - fvc::snGrad(p_rgh)
                  

                   or here?



                ) * mesh.magSf()
            )
        );

        fvOptions.correct(U);
    }

Luke99,
do you think pEqn should be changed accordingly?
if you've known how U and pEqn come, you will see what and where changes should be imposed.

[Luke99]

Hi Alundilong,
so I read a bit further into the solver.
Correct me if I am wrong but as far as I understand it the Navier-Stokes Equation is solved as follows:
In the UEqn.H I first construct the U vector Matrix and then have the option to use the momentum predictor to kind off precalculate the velocity field. But the final calculation of the velocity field happens in pEqn.H after I iteratively solved the pressure Equation.
So therefore as you already said I should either implement it in the U vector Matrix or the pressure Equation and in the latter case if I want to use the momentum predictor there as well.

[dlahaye]

You can use source function object to specify a momentum source term.

See e.g. https://www.openfoam.com/documentati...s-sources.html .