[kaaja]

Hi,
I am having problems with a BC, see the below picture, which displays Uz

strangeUz.png

For the top we have the BC U slip.
For the side facing the screen we have symmetryPlane.

Does anybody know how I can get Uz=0 everywhere on the top?

[Eko]

Originally Posted by kaaja

Hi,
I am having problems with a BC, see the below picture, which displays Uz

Attachment 61080

For the top we have the BC U slip.
For the side facing the screen we have symmetryPlane.

Does anybody know how I can get Uz=0 everywhere on the top?

Can you post your U-file?
Would be easier to help you

[kaaja]

Of course:

/*--------------------------------*- C++ -*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  5                                     |
|   \\  /    A nd           | Web:      www.OpenFOAM.org                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       volVectorField;
    location    "0";
    object      U;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

Umean 0.12;//0.24;
mUmean -0.12;//-0.24;

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

internalField   uniform ($mUmean  0 0);

boundaryField
{
    leftWall //outlet
    {
        type            outletPhaseMeanVelocity;
        alpha           alpha.water;
        Umean           $Umean;
        value           $internalField;
    }
    rightWall //intlet
    {
        type                    fixedValue;
        value                   $internalField;
    }
    lowerWall //part of bottom
    {
        type                    movingWallVelocity;  
        value                   $internalField;
    }
    lowerWallMoving //part of bottom
    {
        type                    movingWallVelocity;  
        value                   $internalField;        
    }
    boat
    {
        type            movingWallVelocity;
        value           uniform (0 0 0);
    }
    atmosphere // This is the top
    {

        type                slip;
    }
    back
    {
        type               slip;
    }
    front
    {
        type               symmetryPlane;
    }
}

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

[luther1990]

1. Why do you use mUmean as internalField but Umean as mean velocity at outlet?
2. The aspect ratio at that location is too big.
3. It also may be caused by the interpolation of visualization in paraview.

[kaaja]

luther1990:
Thanks for answering!

1) mUmean: I don't think it matters, since it is the absolute value that is used in "Umean $Umean;".
2) Aspect ratio: Thank you, we had not thought about this.
3) Interpolation: I have changed the mesh to get better aspect ratio especially near the boundaries. Still I get the strange result in paraFoam. Maybe it has to do with paraFoam interpolation then, as you suggest.

Thanks again!

[kaaja]

Using a good aspect ration, I still get the problem at the corner of the box. The values of U, p, k, and omega all explodes at this point. Does anybody have an idea on how to fix this?

edit: I think there is some conflict between the two BCs. The side BC allows Uz different from zero, while the top BC requires Uz = 0. OpenFoam seems to choose the first. How is it possible to make openFoam choose the top BC over the side BC in the intersectino point?

[luther1990]

Originally Posted by kaaja

Using a good aspect ration, I still get the problem at the corner of the box. The values of U, p, k, and omega all explodes at this point. Does anybody have an idea on how to fix this?

edit: I think there is some conflict between the two BCs. The side BC allows Uz different from zero, while the top BC requires Uz = 0. OpenFoam seems to choose the first. How is it possible to make openFoam choose the top BC over the side BC in the intersectino point?

Sorry to reply late.

OpenFOAM uses FVM. Typically, U p k omega are all stored at cell center and patch face center, not at node. If a boundary cell has two or more patch faces, it will use distance-based weight function to calculate the value.

Check cell value instead of point value in paraview. And check the boundary values in the result files.

Theoretically, cell value at corner should have non-zero Uz, but top face value should have Uz=0.

[kaaja]

Thanks for replying!
I got a really high value at the corner.
I ended changing the whole BC to avoid this.
If I change back the BC, I will certainly have a closer look at what you are prescribing!