View Issue Details

IDProjectCategoryView StatusLast Update
0001504OpenFOAMBugpublic2015-01-31 14:54
ReporterSvensen Assigned Tohenry  
PrioritynormalSeverityminorReproducibilityalways
Status closedResolutionfixed 
PlatformGNU/LinuxOSUbuntuOS Version14.04
Summary0001504: pimpleFoam unstable with unsteady BC
DescriptionI try to repeat the native experiment using OpenFOAM 2.3.0. Solver pimpleFoam is used. The domain has a complex form with one inlet and one outlet. For boundary conditions it was setted values of velocity at inlet and pressure at outlet. The flow is laminar, fluid is newtonian.
According to experimental data the velocity at inlet and pressure at outlet are trancient.
In the simple case, when I fixed the values for them (using fixedValue), I've got a good results (files "statics_*.png).
But when I tried to make the unsteady boundary conditions, at the inlet there are oscillations for pressure, which doesn't correlate with experimental data (files "trancient_*.png").

BC for velocity:
dimensions [0 1 -1 0 0 0 0];

internalField uniform (0 0 0);

boundaryField
{
    inlet
    {
    type uniformFixedValue;
        uniformValue tableFile;
        tableFileCoeffs
        {
            fileName "$FOAM_CASE/bc/inlet_velocity.bc"
            outOfBounds repeat;
        }
    }

    outlet
    {
        type zeroGradient;
    }

    vessel
    {
        type fixedValue;
        value uniform (0 0 0);
    }
}

BC for pressure:
dimensions [0 2 -2 0 0 0 0];

internalField uniform 9.421284545;

boundaryField
{
    inlet
    {
        type zeroGradient;
    }

    outlet
    {
        type uniformFixedValue;
        uniformValue tableFile;
        tableFileCoeffs
        {
            fileName "$FOAM_CASE/bc/outlet_pressure.bc"
            outOfBounds repeat;
        }
    }

    vessel
    {
        type zeroGradient;
    }
}

file controlDict:
application pimpleFoam;
startFrom startTime;
startTime 0;
stopAt endTime;
endTime 2;
deltaT 0.001;
writeControl adjustableRunTime;
writeInterval 0.1;
purgeWrite 0;
writeFormat binary;
writePrecision 6;
writeCompression uncompressed;
timeFormat general;
timePrecision 6;
runTimeModifiable true;

adjustTimeStep yes;
maxCo 1;
maxAlphaCo 1;
maxDeltaT 0.1;

file fvSolution:
solvers
{
    p
    {
        solver GAMG;
        tolerance 1e-06;
        relTol 0.01;
        smoother GaussSeidel;
        cacheAgglomeration true;
        nCellsInCoarsestLevel 10;
        agglomerator faceAreaPair;
        mergeLevels 1;
    }

    pFinal
    {
        solver GAMG;
        tolerance 1e-06;
        relTol 0;
        smoother GaussSeidel;
        cacheAgglomeration true;
        nCellsInCoarsestLevel 10;
        agglomerator faceAreaPair;
        mergeLevels 1;
    }

    "(U|k|epsilon)"
    {
        solver smoothSolver;
        smoother symGaussSeidel;
        tolerance 1e-05;
        relTol 0.1;
    }

    "(U|k|epsilon)Final"
    {
        $U;
        tolerance 1e-05;
        relTol 0;
    }
}

PIMPLE
{
    nOuterCorrectors 2;
    nCorrectors 4;
    nNonOrthogonalCorrectors 5;
    pRefCell 0;
    pRefValue 0;
}

relaxationFactors
{
    fields
    {
    }
    equations
    {
        "U.*" 1;
        "k.*" 1;
        "epsilon.*" 1;
    }
}
TagsNo tags attached.

Activities

Svensen

2015-01-31 09:43

reporter  

transient_pressure_at_inlet.png (95,106 bytes)   
transient_pressure_at_inlet.png (95,106 bytes)   

henry

2015-01-31 11:09

manager   ~0003632

Spurious pressure fluctuations may be induced by unphysical accelerations caused by non-smoothness of the tabulated velocity data. Check that the acceleration of the inlet flow is smooth by taking the time derivative of the data.

Issues may also arise if the tabulated data causes reverse flow at either the inlet or outlet. Firstly you will find the problems are exacerbated by the tet-mesh and the behaviour is likely to improve by adding extrusion-layers at the inlet and outlet. Secondly zeroGradient on velocity is not appropriate for reverse-flow.

If this does not help and you are unable to resolve the issue you will need to arrange for some support.

Issue History

Date Modified Username Field Change
2015-01-31 09:43 Svensen New Issue
2015-01-31 09:43 Svensen File Added: transient_pressure_at_inlet.png
2015-01-31 11:09 henry Note Added: 0003632
2015-01-31 11:09 henry Status new => closed
2015-01-31 11:09 henry Assigned To => henry
2015-01-31 11:09 henry Resolution open => fixed