[michalRS]

Hello All

I am currently trying to model flow through a tunnel, where a small part of the boundary deforms in time. the deformation is basically a small, solid "bubble" that expands into the flow and collapses back on the wall periodically(prescribed motion, not due to fluid interaction).

I wrote a udf inspired by "define-grid-motion" in the Fluent manual, and this works wonderfuly for one cycle, but in the middle of the second cycle the smoothing of the mesh fails, and negative cell volumes are created.

I tried changing the time step, and the diffusion parameter (under smoothing parameters of the dynamic mesh) but these hardly affected the problem.

1- can anyone suggest why this happens on the second cycle, and what I can do to prevent this?

2- if I save all the case files during the first cycle, can I write a udf that loads all the meshes periodically to create the same motion without having to go through the smoothing process each time?


thank you for any suggestions or comments