[dai_bach]

Hi,

I'm running an initial test case using the solidDisplacementFoam solver. The problem is a rectangular plate, clamped on three sides and free on the fourth and subject to a hydrostatic pressure on one side. My question is how do I write the boundary conditions for this kind of non-uniform distribution? My sample 0/D file is included below. Thanks for all your suggestions and help.

Dai

/*--------------------------------*
- C++ -
*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  2.3.0                                 |
|   \\  /    A nd           | Web:      www.OpenFOAM.org                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       volVectorField;
    object      D;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions      [0 1 0 0 0 0 0];

internalField   uniform (0 0 0);

boundaryField
{
    left
    {
        type            tractionDisplacement;
        traction        uniform ( 0 0 0 );
        pressure        uniform 0;
        value           uniform (0 0 0);
    }
    right
    {
        type            tractionDisplacement;
        traction        uniform ( 0 0 0 );
        pressure        uniform 0;
        value           uniform (0 0 0);
    }
    down
    {
        type            tractionDisplacement;
        traction        uniform ( 0 0 0 );
        pressure        uniform 0;
        value           uniform (0 0 0);
    }
    up
    {
        type            empty;
    }
    front
    {
        type            tractionDisplacement;
    traction    ( 0 0 0 );
    pressure    uniform 1e5;
    value        uniform 0;
    }
    back
    {
    type        tractionDisplacement;
    traction    uniform ( 0 0 0 );
    pressure    uniform 0;
    value        uniform (0 0 0);
    }
}

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

[Annier]

Hi David,
The code you have presented above is the boundary condition for displacement Field (D). You have to make a new file for pressure field (p) or buoyantPressure field(p_rhogh) within the 0 directory and write a boundary condition on the wall something like this:

{
 ....
    class       volScalarField;
    object      p_rhogh;
}
dimensions      [0 2 -2 0 0 0 0];

internalField   uniform 0;

boundaryField
{
 ...
   Walls
    {
        type            buoyantPressure;
        value          uniform 0;
    }
...

 }

You can study this thread related to hydrostatic pressure:
http://www.cfd-online.com/Forums/ope...sure-cube.html


Since, solidDisplacementFoam won't be using pressure field and so i guess significant workouts and solver modifications have to be accomplished for this solver. Some concepts related to "fan boundary condition" are given in the following thread to describe the pressure jump:
http://www.cfd-online.com/Forums/ope...interfoam.html

Yours
Anil Kunwar

[ron_OFuser]

Originally Posted by dai_bach

Hi,

I'm running an initial test case using the solidDisplacementFoam solver. The problem is a rectangular plate, clamped on three sides and free on the fourth and subject to a hydrostatic pressure on one side. My question is how do I write the boundary conditions for this kind of non-uniform distribution? My sample 0/D file is included below. Thanks for all your suggestions and help.

Dai

/*--------------------------------*
- C++ -
*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  2.3.0                                 |
|   \\  /    A nd           | Web:      www.OpenFOAM.org                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       volVectorField;
    object      D;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions      [0 1 0 0 0 0 0];

internalField   uniform (0 0 0);

boundaryField
{
    left
    {
        type            tractionDisplacement;
        traction        uniform ( 0 0 0 );
        pressure        uniform 0;
        value           uniform (0 0 0);
    }
    right
    {
        type            tractionDisplacement;
        traction        uniform ( 0 0 0 );
        pressure        uniform 0;
        value           uniform (0 0 0);
    }
    down
    {
        type            tractionDisplacement;
        traction        uniform ( 0 0 0 );
        pressure        uniform 0;
        value           uniform (0 0 0);
    }
    up
    {
        type            empty;
    }
    front
    {
        type            tractionDisplacement;
    traction    ( 0 0 0 );
    pressure    uniform 1e5;
    value        uniform 0;
    }
    back
    {
    type        tractionDisplacement;
    traction    uniform ( 0 0 0 );
    pressure    uniform 0;
    value        uniform (0 0 0);
    }
}

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

I found document related to your work

http://www.tfd.chalmers.se/~hani/kur...s_reviewed.pdf

Cheers Chalmers!

[dai_bach]

Hi,

Thank you both for your prompt replies.

@Anil: is adding a pressure field consistent with solidDisplacementFoam? Looking at the .C definition file, only the displacement field, D, is resolved. My question was motivated by how a hydrostatic-like pressure could be applied to one of the front/back walls. Obviously, this is equivalent to a stress normal to the wall that increases linearly with depth and there are both pressure and traction fields in the 0/D file so surely there's some way of including this in the initial/boundary conditions?

[Annier]

Hi David,
-You are correct in mentioning that the plate analysis is related to displacement field variable with Pressure at BC only. So, pressure field cannot be defined for this domain.
-So, I guess that you have to define the value of pressure gradient in the boundary which increases along the depth of the required boundary wall.

Yours
Anil Kunwar

[dai_bach]

@Anil: Indeed, but how...?

[Annier]

Hi David,
-please refer to page 8 of this tutorial prepared by Tian Tang:
http://www.tfd.chalmers.se/~hani/kur...erReviewed.pdf
http://www.tfd.chalmers.se/~hani/kur...s_TianTang.pdf

- The initial description of the hydrostatic bc at the solid domain would be setting the following:
a) In 0/D file

boundaryField
        {
          Wall
     { 
          type            tractionDisplacement; 
          traction        uniform ( 0 0 0 ); 
          pressure        ...; 
          value           uniform (0 0 0); 
      }

The value for traction Force would be zero and that for pressure should be non-zero. The value of pressure specified here represents its gradient if this boundary is defined as a derived type of type name "fixedGradient" in constant/polyMesh/boundary file.


b) Thus,it might be necessary to use derived boundary type for the corresponding boundary than the base (patch) type.
i.e., in constant/polyMesh/boundary file

{
...
object   boundary;
}

...

(
Wall
{
type      fixedGradient;
...
}
...
)

So, you have to specify the gradient of pressure in the ... (blank space) for above displacement boundary field (D) file. In this context, i am not sure how we specify the gradient as vertical or horizontal.

-For more details, please refer OpenFOAM User Guide, Chapter 5 (Mesh Generation and Conversion) and see therein the Boundaries.

Yours
Anil Kunwar

[ron_OFuser]

Hi ALL, check out the link (copyright chalmer university)

http://www.cfd-online.com/Forums/ope...-chalmers.html