RBD objects sticking together

[bandini]

Hi there, a couple quick questions...

I am doing a fracture sim and I keep getting some pieces that stick together during the sim. I don't think their SDF's are penetrating and this is not using glue. Wondering if there is a setting I can look at to resolve the stickiness.

Other question is that during the initial force, some smaller pieces fly off at a very large velocity - much larger than seems appropriate for the sim. Is there any kind of way to limit velocity or stop the smaller pieces from going crazy?

[MENOZ]

Hi there, a couple quick questions...

I am doing a fracture sim and I keep getting some pieces that stick together during the sim. I don't think their SDF's are penetrating and this is not using glue. Wondering if there is a setting I can look at to resolve the stickiness.

Other question is that during the initial force, some smaller pieces fly off at a very large velocity - much larger than seems appropriate for the sim. Is there any kind of way to limit velocity or stop the smaller pieces from going crazy?

Hi, I think you can try to set the resolve penetration to 3 or 4 in the RBD solver, or try to push up the substeps. another thing that you can try is to set the collision to edges in the collision tab on the rbd object.

To damp the speed you can use a drag node.

[mawi]

Hi there, a couple quick questions...

I am doing a fracture sim and I keep getting some pieces that stick together during the sim. I don't think their SDF's are penetrating and this is not using glue. Wondering if there is a setting I can look at to resolve the stickiness.

Have you disabled the clustering?

voronoi fracture configure object -> points -> surface/interior/exterior -> clustering

[johner]

Hmm, that's strange. Any chance you can simplify the scene and post a .hip file? Is there are any possibility you're starting out with overlapping objects? I've seen it happen where the RBD Solver does OK simulating multiple copies of the same object, but when they fracture the sim goes crazy.

As menoz noted, a Drag force can help tame crazy velocities, and you can do things like create groups of only the smaller pieces to apply Drag to (see here for an example of this).

Also, under the Fracture tab of the RigidBodySolver, there's a Mininum Piece Volume parameter that culls any pieces below a specified volume, so you can prevent really small, flat shards from ever being created in DOPs, which can limit the number of misbehaving tiny objects.

[Macha]

I had the same problem with large velocities. I solved it in that particular case by applying a drag force with "Ignore mass" ticked.

It would be interesting to link drag to speed, as you can do in pops, but I haven't tried in dops.

Edited November 29, 2010 by Macha

[bandini]

Thank you for all the suggestions. I'll see if I can post my scene file. I kept changing all kinds of parameters to try to fix the problem. Have to see if I have one saved where the problems were easily apparent.

I was using a magnet force to do the explosion. That may have something to do with the pieces flying apart like crazy. I did try playing with drag but found that changed the behavior of the simulation significantly. To counteract that I had to decrease the density of my objects or increase the magnet force - which did not behave the way I thought it should.

All the geometry is pre-fractured. I do have clustering enabled, and that is intentional. Although I thought that clustering merges the geometry together? I was seeing two distinct pieces merged together in the sim.

Also, about the pieces flying off like crazy - I was looking to see if there was a speed limit node or something similar, like in POPS. Is there any equivalent there in DOPS?

I am rendering something right now, so it is a little tough to get to my .hip files. I'll try to post a file in a little bit.

[bandini]

Ok, here is a simplified hip file.

Explosion_ProblemExample_01.hipnc

I know I am using somewhat low settings on my RBD collision solver. Funny thing is that when I increase the collision/contact passes, etc. the results are significantly different. Sometimes it appears the magnet force has almost no effect even. Very strange.

Anyway, I ended up doing my test shot by using an invisible sphere pushed through the scene to create the explosion - which works ok, but isn't exactly what I was looking for. Test movie here. I did not get any of these problems when relying on collisions to create the motion. Seems like there is something in the way I am using forces that are problematic.