Dynamic mesh distortion
 

This is a question I see has been posed previously without any resolution...

I have a simple 3D rectangular profile channel. The top face oscillates (pimpleDyMFoam, of 2.4) in the normal direction. It works fine in a 2D setup but in the 3D case distortion (severe non-orthogonality) results in divergence. Paraview shows that this is occurring close to the wall: since the wall is static the cells close by are effectively crushed as the top face oscillates inwards.

Wondered if anyone has experience/suggestions for correcting this?

As a first attempt I will try adding motion to the wall.

So running with the extra dynamic BC on the walls (modified patches) does help; the case runs without crashing. There is a limit however to the amplitude to avoid cell crushing. The problem then requires use of layer removal for large amplitudes.