[Thangam]

I have tried to modify the buoyantBoussinesqPisoFoam for the LES capability solver which was available in a 1.6.x git repository and followed the instructions on the openfoam wiki site (http://openfoamwiki.net/index.php/Bu...sinesqPisoFoam) to have the LES capability.

Though the code compiled without any warnings, when I tried to run the case with this solver, I got the error :

--> FOAM FATAL IO ERROR: 
keyword laplacian((1|A(U)),p) is undefined in dictionary 
"/home/thangam/Documents/LES_dec23/system/fvSchemes::laplacianSchemes"

file: /home/thangam/Documents/LES_dec23/system/fvSchemes::laplacianSchemes from line 48 to line 56.

    From function dictionary::lookupEntry(const word&, bool, bool) const
    in file db/dictionary/dictionary.C at line 400.

FOAM exiting

The solver was reading laplacian((1|A(U)),p) though I had laplacian((1|A(U)),p_rgh).


Though I had no clue why this was happening, I tried to edit the pEqn.H to its present form which is :

{
    volScalarField rUA("rUA", 1.0/UEqn.A());
    surfaceScalarField rUAf("(1|A(U))", fvc::interpolate(rUA));

    U = rUA*UEqn.H();

    surfaceScalarField phiU
    (
        (fvc::interpolate(U) & mesh.Sf())
      + fvc::ddtPhiCorr(rUA, U, phi)
    );

    phi = phiU + rUAf*fvc::interpolate(rhok)*(g & mesh.Sf());

    for (int nonOrth=0; nonOrth<=nNonOrthCorr; nonOrth++)
    {
        fvScalarMatrix p_rghEqn
        (
            fvm::laplacian(rUAf, p_rgh) == fvc::div(phi)
        );

        if (corr == nCorr-1 && nonOrth == nNonOrthCorr)
        {
            p_rghEqn.solve(mesh.solver(p_rgh.name() + "Final"));
        }
        else
        {
            p_rghEqn.solve(mesh.solver(p_rgh.name()));
        }

        if (nonOrth == nNonOrthCorr)
        {
            phi -= p_rghEqn.flux();
        }
    }

    U += rUA*fvc::reconstruct((phi - phiU)/rUAf);
    U.correctBoundaryConditions();

    #include "continuityErrs.H"

    p = p_rgh + rhok*gh;

    if (p_rgh.needReference())
    {
        p += dimensionedScalar
        (
            "p",
            p.dimensions(),
            pRefValue - getRefCellValue(p, pRefCell)
        );
        p_rgh = p - rhok*gh;
    }
}

The issue is this compiles and runs successfully but it is solving for p_rgh twice(see below for the output). Not clear where im going wrong with the code! Any advise would be appreciated!

Time = 0.01143

Courant Number mean: 0.00258195 max: 0.975465
DILUPBiCG:  Solving for Ux, Initial residual = 0.000280691, Final residual = 4.0324e-07, No Iterations 1
DILUPBiCG:  Solving for Uy, Initial residual = 7.18944e-06, Final residual = 6.81315e-09, No Iterations 1
DILUPBiCG:  Solving for Uz, Initial residual = 0.000280614, Final residual = 3.93437e-07, No Iterations 1
DILUPBiCG:  Solving for T, Initial residual = 3.25608e-05, Final residual = 3.01528e-09, No Iterations 1
DICPCG:  Solving for p_rgh, Initial residual = 0.00346845, Final residual = 3.36784e-05, No Iterations 68
time step continuity errors : sum local = 5.57448e-11, global = 6.73984e-13, cumulative = -1.04705e-10
DICPCG:  Solving for p_rgh, Initial residual = 0.000499402, Final residual = 9.95821e-09, No Iterations 826
time step continuity errors : sum local = 1.64757e-14, global = -5.49e-16, cumulative = -1.04706e-10
bounding k, min: -1.19509 max: 11.2655 average: 0.0156095
ExecutionTime = 53260.1 s  ClockTime = 53494 s

Thanks.

[Bernhard]

I would like to encourage you to read about the PISO algorithm. You will notice the importance of solving the pressure-corrector equation multiple times. Also, I think I explained this in another topic to you as well, with a reference to the release notes of OF 1.7, where for these solvers p has been changed to p_rgh.

It does not makes sense to change p to p_rgh in your pEqn.H file, because it is misleading and inconsistent with all other solvers using p or p_rgh. If I were you, I would rather change this in system/fvSchemes, which is actually what your error message clearly indicates.