Please see the link

https://www.researchgate.net/post/How_do_you_solve_the_problem_of_negative_cell_volume_generation_in_Fluent

This discussion might help you out..

Cheers!

Check the size of the smallest volume. It can happen that also a very small motion of the boundaries implies the collapse of one element. 

Furthermore, check the diffusivity of the mesh motion.

Franco

I would suggest you check your mesh quality, mostimes these type of problem comes as a result of bad mesh quality. Both for the unstructured and structured mesh,  do a quick check of quality and remesh if found to be poor.

I experienced the same problem while moving particular zone. Problem is remeshing, remeshing happening only in triangular zone that may be fluent limitation.

So solution,

1.Use triangular mesh in the dynamic zone (moving).

2.Give smoothing parameter appropriate.

3.And check time step size, distance moved in one time step must be less than the mesh size in dynamic zone (moving).

PS: Remeshing interval must be one (default: it will be 5)

I may give you suggestions in terms of the unstructured mesh. Using quadrilateral cells (just like prism layers) around the airfoil and proper transition to triangular cells, I am very confident you will be able to generate a good mesh in this case.

Assuming you used the described strategy, I suggest you to move the near body layer of cells together with the airfoil, and the remaining mesh elements, with larger cells should accommodate the deformations properly. If it still doesn't work, you should use remeshing options, but it will increase your computational cost.