[lucfd]

I am running a case which is derived from the OpenFOAM/tutorials/periodicHill (steady state). That is, the set-up used which serves as initial conditions for the time-domain simulation using LES.

The main changes are 1) 2D 2) free stream rather than channel flow. Flow conditions are kept the same.



In order to achieve this, I have modified:


p

/*--------------------------------*- C++ -*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  v2312                                 |
|   \\  /    A nd           | Website:  www.openfoam.com                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       volScalarField;
    object      p;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions      [0 2 -2 0 0 0 0];

internalField   uniform 0;

boundaryField
{
    "(inlet|outlet)"
    {
        type            cyclic;
    }
    
    "(front|back)"
    {
        type            empty;
    }


    "(hills)"
    {
        type            zeroGradient;
    }
    
    "(top)"
    {
        type            freestreamPressure;
        freestreamValue $internalField;

    }

}


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

U

/*--------------------------------*- C++ -*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  v2312                                 |
|   \\  /    A nd           | Website:  www.openfoam.com                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       volVectorField;
    object      U;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions      [0 1 -1 0 0 0 0];

internalField   uniform (1e-3 0 0);

boundaryField
{
    "(inlet|outlet)"
    {
        type            cyclic;
    }
    
    "(front|back)"
    {
        type            empty;
    }

    "(hills)"
    {
        type            noSlip;
    }
    
    "(top)"
    {
        type            freestreamVelocity;
        freestreamValue $internalField;

    }
    
}


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

nut

/*--------------------------------*- C++ -*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  v2312                                 |
|   \\  /    A nd           | Website:  www.openfoam.com                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       volScalarField;
    object      nut;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions      [0 2 -1 0 0 0 0];

internalField   uniform 0;

boundaryField
{
    "(inlet|outlet)"
    {
        type            cyclic;
    }
    
    "(front|back)"
    {
        type            empty;
    }


    "(hills)"
    {
        type            nutUSpaldingWallFunction;
        value           $internalField;
    }

    "(top)"
    {
        type            freestream;
        freestreamValue           uniform 0;
    }
}


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

nuTilda

/*--------------------------------*- C++ -*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  v2312                                 |
|   \\  /    A nd           | Website:  www.openfoam.com                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       volScalarField;
    object      nuTilda;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions      [0 2 -1 0 0 0 0];

internalField   uniform 1e-8;

boundaryField
{
    "(inlet|outlet)"
    {
        type            cyclic;
    }
    
    "(front|back)"
    {
        type            empty;
    }


    "(hills)"
    {
        type            fixedValue;
        value           uniform 0;
    }

    "(top)"
    {
        type            freestream;
        freestreamValue uniform 0;

    }
    
}


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

fvOptions

/*--------------------------------*- C++ -*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  v2312                                 |
|   \\  /    A nd           | Website:  www.openfoam.com                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       dictionary;
    object      fvOptions;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

//momentumSource
//{
//    type            meanVelocityForce;

//    selectionMode   cellZone;
//    cellZone        inletCellZone;

//    fields          (U);
//    Ubar            (1 0 0);
//}


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

/*--------------------------------*- C++ -*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  v2312                                 |
|   \\  /    A nd           | Website:  www.openfoam.com                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       dictionary;
    object      fvSolution;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

solvers
{
    p
    {
        solver          GAMG;
        smoother        DICGaussSeidel;
        tolerance       1e-06;
        relTol          0.05;
    }

    "(U|k|nuTilda)"
    {
        solver          PBiCG;
        preconditioner  DILU;
        tolerance       1e-05;
        relTol          0.1;
    }
}

SIMPLE
{
    nNonOrthogonalCorrectors 0;
    consistent      yes;
    //pRefCell        0;
    //pRefValue       0;
}

relaxationFactors
{
    equations
    {
        U           0.9;
        nuTilda     0.7;
    }
}


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

The main modifications are changing "top" to 'patch', and imposing freestream boundary conditions rather than wall conditions. Moreover, I had to turn off momentumSource because this is not longer applicable to external flow problems. I also turned off pRefCell and pRefValue in fvSolution because according to me this is being taken care of by the freestream BC.


So far, I did not modify the blockMeshDict. But, using the original settings this means a very fine resolution for the now freestream BC. The solver returns that I have a very (very) fine mesh near the wall.

yPlus yPlus write:
    writing field yPlus
    patch hills y+ : min = 4.3852231e-05, max = 0.001221246, average = 0.00049500734

Nevertheless, I have a convergent result as shown in the image (pressure) I attached. Of course, I might have to increase the height of the freestream patch. But, I want to focus on the effect of modifying blockMeshDict. There, I changed

simpleGrading (1 ((0.5 0.5 100)(0.5 0.5 0.01)) 1)

to

simpleGrading (1 400 1)

in order to maintain a very fine grid. The results however, do not converge as shown in the second image. The solver returns

patch hills y+ : min = 2.0608024e-05, max = 0.00050223013, average = 0.00019107119

My questions is why? What could possibly cause the stable solution to become unstable only by removing refinement near the freestream patch but remaining very fine near the wall.