I have an RBD and FLIP sim interacting with each other and I'm using the Shelf auto fracture node to have the two rocks shatter properly on impact. This creates so many small pieces that I think make the FLIP sim freak out around the frame 50-60. I currently have both solvers at 4/12 substeps and have tweaked the FLIP solver to handle high speed collisions. I tried the swirly kernel and it helped a bit but also didn't look as good… Do I need to simply add substeps at the DOP level? Has anyone had a similar reaction and is there anything to keep it in check? I've checked and the collision SDFs are being properly generated for the shattered pieces but they don't seem to go well into the water (they seem to create empty air bubbles which might then blow up). I really don't know. The reaction on this last run is so extreme that it makes me think there is something much more basic I'm not getting right / is going wrong.
I've attached the project file if you want to run it locally. (OpenCL is on)
Thanks!
I also forgot to mention that I recently reopened the project which I originally created in H17 and had put in the backburner because of this issue. I wanted to run it w/ a slightly larger tank, more detail/substeps and a better render. Here's what I got out the first time. There are hints of the problem on the last frames (probably why I cut it short…)
I think I figured it out after running it a couple more times and getting out some diagnostic flipbooks. I think the feedback scale (1.2) was making the angular velocity on the small pieces go haywire. You can see it clearly in the video I've attached. I'm resimming with 2 substeps at the DOP level, slightly more particles, tweaked densities for the rocks and the boat and higher rotational stiffness on the rocks. I'll report back with the results tomorrow
