[CeesH]

Dear all,

I'm running a stirred tank simulation with the following method:

- cyclic domain (60 degrees modelled)
- sliding mesh simulation
- parallel solver (4 cores, typically)
- massless DPM of ~500k particles

In fluent 15.07 there is the option to assign a `DPM domain' which clusters the particles in a separate partition, rather than assigning them to the partition their cell location is in. Theoretically, this should be helpful in unsteady flows where the local particle density may differ over time. In practice, I simply wanted to test if my simulations would speed up by this.

But, so far I have not been able to get my simulations to run. After some timesteps (2-20 typically) I get a segmentation violation error.
I'm in contact with ANSYS support on this, they have not been able to consistently reproduce the error, but it seems to be a sliding mesh-DPM interplay.

I was wondering if anybody else has problems with this, and if you perhaps already found some solution.

Best,
Cees