[joshmccraney]

Hi FOAMers!

I am running interFoam version 6. I am looking at an inviscid 2D drop of water in air with zero g. I give the drop a slight disturbance, which causes the drop to oscillate back to equilibrium. When I perform a mesh-convergence study, the results for dx = 40 and 100 are much closer to the analytic prediction than dx = 200: why? I have a limit on the max Courant number, so I don't think it's a time step issue. What do you think?

For your consideration, my controlDict looks like this

/*--------------------------------*- C++ -*----------------------------------*\
  =========                 |
  \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox
   \\    /   O peration     | Website:  https://openfoam.org
    \\  /    A nd           | Version:  6
     \\/     M anipulation  |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       dictionary;
    location    "system";
    object      controlDict;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

application     interFoam;

startFrom       startTime;

startTime       0;

stopAt          endTime;

endTime         0.02;//1000;

deltaT          0.000000001; // INITIAL TIME STEP

writeControl    adjustableRunTime;

writeInterval   0.0002;

purgeWrite      0;

writeFormat     binary;

writePrecision  6;

writeCompression off;

timeFormat      general;

timePrecision   6;

runTimeModifiable yes;

adjustTimeStep  yes;

maxCo           0.2;
maxAlphaCo      0.2;

maxDeltaT       1;//0.0000025;

functions
{  
    waterVolume
    {
        type            volFieldValue;
        libs            ("libfieldFunctionObjects.so");
        log             true;
        writeControl    writeTime;
        writeFields     true;
        regionType      all;
        operation       volIntegrate;
        fields
        (
            alpha.water
        );
    }
}


// ************************************************************************* //

Thanks for everything!

[magnushaese]

I have two suggestions, convergence-wise. I assume this is laminar, thus U-p coupling only with the calculation of the liquid fraction via the MULES solver.


1. As for MULES, you might want to check independence of its subcycles i.e.nAlphaSubCycles and nAlphaCorr.


2. Also more general. It seems like you chose a convective CFL~0. 2 driven solution. I assume you might use the PISO mode at this time step. However, I would recommend lowering the initial linear solver residuals more restrict each time step, thus using the PIMPLE framework to guarantee independence of residuum and a successive lowering of the inital residuums of from 1e-5 down to 1e-8 or so.