stop tiles penetrating without simulation

[Jordan Walsh]

Imagine you have a pile of random tiles that you want on the ground setup so they are not penetrating each others mesh (the mesh may not just be square, could be broken tiles). Is there a way to set this up without dropping them all from a height in dops and simulating them settle? Hopefully a faster way to do it!

[goldleaf]

You could run an rbd sim with no forces, and a time offset sop, and only cook one frame. Shouldn't be too hard to wrap into a digital asset too.

[Jordan Walsh]

Im not sure I understand. If you had no gravity in the sim how would they settle?

I was kind of wanting a way to to it without a sim.

[goldleaf]

Ah, sorry about that, I misunderstood. I thought you just wanted to resolve any remaining penetrations for already stacked models, whether they were stacked by a sim or hand placed.

[bloomendale]

it is some sort of packing problem and i think rbd sim is the simplest way to do it)

look at what we have to do to pack simple spheres of different size. And you want to "pack" tiles.. even broken ones..

[lukeiamyourfather]

This would be a good use of the Solver node in a SOP network. A rigid body simulation is another option, probably easier than using the Solver node but wouldn't provide as much control over the results.

[kleer001]

Just curious, but what's stopping you from using a sim?

[old school]

You can set up a quick RBD solver engine sim to do this and use the Resolve Penetration parameter on the Rigid Body Solver DOP. Set the Resolve Penetration value to anywhere from half of the objects to all of the pieces you anticipate interpenetrating.

Then continue to run the sim until the pieces start to settle again or get to where you like it. The first frame of the sim from 1 to 2 can solve all the interpenetrations if you set Resolve Penetration high enough. Or run a few frames and let things sort out for themselves. Less energy that way.

This technique works even better if the pieces are roughly in place with not too much penetration as the movement of the pieces to solve the interpenetration are forces that will carry in to the next frame of the simulation.

Here's a worst case where multiple objects are severely interpenetrating with each other. On frame 2, the penetrations are resolved and by frame 6 pieces are settled down enough to call it a day.

Frame 1:

Frame 2:

Frame 6:

resolve_interpenetration_settling_boxes.hip.zip