[de_Broglie]

The title pretty much says it. Imagine we have a physically moving wall, for which the motion is prescribed using pointDisplacement as boundary condition. But we need to also specify an additional tangential wall velocity for which no mesh movement is necessary. How do we go about this?


Thanks a bunch in advance for any help.

[al.csc]

For ESI-v2206, you find the answer in:

- If rotating, ie cylinder: OpenFOAM-v2206/src/finiteVolume/fields/fvPatchFields/derived/rotatingWallVelocity/rotatingWallVelocityFvPatchVectorField.H

Description
    This boundary condition provides a rotational velocity condition.

Usage
    \table
        Property     | Description             | Required    | Default value
        origin       | origin of rotation in Cartesian coordinates | yes|
        axis         | axis of rotation        | yes         |
        omega        | angular velocty of the frame [rad/s] | yes    |
    \endtable

    Example of the boundary condition specification:
    \verbatim

// Here what you need:

    <patchName> // for U
    {
        type            rotatingWallVelocity;
        origin          (0 0 0);
        axis            (0 0 1);
        omega           100;
    }

- If moving, ie flat wall: OpenFOAM-v2206/src/finiteVolume/fields/fvPatchFields/derived/movingWallVelocity/movingWallVelocityFvPatchVectorField.H

Description
    This boundary condition provides a velocity condition for cases with
    moving walls.

Usage
    Example of the boundary condition specification:
    \verbatim

//What you need: 

    <patchName> // for U
    {
        type            movingWallVelocity;
        value           uniform (0 0 0);    // Initial value
    }
    \endverbatim

Cheers!

[de_Broglie]

Thank you, al.csc, for replying. But I believe this is a BC that should not be used when the wall is physically moving. Say you have a cylinder that is rotating around its center, then you wouldn't need a mesh movement and could use rotatingWallVelocity safely.


My question relates to when the wall is actually moving. Think a linearly oscillating cylinder that is also rotating at the same time. The oscillation needs mesh movement, whereas the rotation doesn't.

[al.csc]

Originally Posted by de_Broglie

Thank you, al.csc, for replying. But I believe this is a BC that should not be used when the wall is physically moving. Say you have a cylinder that is rotating around its center, then you wouldn't need a mesh movement and could use rotatingWallVelocity safely.


My question relates to when the wall is actually moving. Think a linearly oscillating cylinder that is also rotating at the same time. The oscillation needs mesh movement, whereas the rotation doesn't.

Now I get your point, sorry. But I do not understand why noSlip BC is not sufficient.

[de_Broglie]

Ok, on movingWallVelocity, this BC works well (and should be used) when the wall is physically moving. The problem is that sometimes you don't want the wall to have an actually movement in the tangential direction. In such situations if movingWallVelocity is used, the mesh actually moves in the tangential direction, which is not desirable.

[de_Broglie]

Originally Posted by al.csc

Now I get your point, sorry. But I do not understand why noSlip BC is not sufficient.

No worries, at all. Sometimes you would want to prescribe a specific velocity on the wall in tangential direction, which simulates a wall movement (without the actual movement and hence no need for mesh motion).

[de_Broglie]

Just to add to this, in commercial codes like Fluent or CFX, and when you have dynamic mesh, you could specify a "relative wall velocity" on wall boundaries in addition to wall mesh motion, which would take care of this problem.