haggi krey

Hi,
I created a simple distributed flip sim with two slices. The basic setup works fine but as soon as I try to use initial states, the sim hangs at first sim frame. I created a separate initial state for every slice. Any idea if distributed sims should work at all with initial states?

Hi, this is a copy of my message on discord's "Think Procedural" server.

Okay, so I tried to create a boiling water effect. Basically I think there are two ways to create discrete bubbles,

1. with incompressible fluid or
   
2. with geometry collisions
   

While my approach with incompressible flip works fine, my approaches with moving colliders will not keep the volume, here are some movies.
Default sim with moving bubbles is this one loosing_volume_default.mp4. You can see how the volume is smaller at the end of the sim
Next I tried to compensate it with a gasequalize microsolver what seems to work even if it looks as if it would first loose volume and then go back to the original again: loosing_volume_equalizer.mp4
And finally I tried the divergence field approach: loosing_volume_divergence.mp4
Seems to work somehow but maybe it's still too much. I attached the hip file as well.

So my questions if anyone is interested:

1. Why does the flip sim loose volume with moving colliders at all?
   
2. Are these methods (maybe I could try to increase the particle radius scale as well) the only way to keep the volume? Feels like cheating. Or am I simply missing some details in my flip sims?
   

Thanks for any answer.

Uppsss... I think I found the problem. The main problem site in front ot the keyboard. The sop crowd import lop node already has everything I need. The important point here is that I have to unpack the agents in the sop import. Now evertyhing seems to work fine.