Maybe this is a bug, maybe I set something up wrong. I am trying to follow @DanielPiker example of his imagecircles. After running the simulation and slowly bring up the slider until circles are packed, I have found that many circles still over lapped each other while other areas are completely empty, even after solver is marked as converged.
I couldn’t figure out why this was happening until I looked in 3d and found that the circles aren’t colliding because they have begun to spread out in 3d space instead of sticking in the 2d plane.
I’d guess that some part of the input geometry is actually a very tiny amount out of plane, which is enough to cause everything to move vertically once you start the packing relaxation.
Anyway, even if things are slightly out of plane, you can keep things from moving further away in Z with AnchorXYZ with Z set to True.
Data recorder is used advance my simulation step size until all circles are packed and I do find the sasquatch one to be better because of the actual inputs (It would be nice if McNeel gave options to have all setting and inputs be wired instead of a mix match of UI and wires). I have no idea what that other component is as I don’t have that plug in installed…
Is graph mapper not a standard component? I have it saved out in my file… It’s not completely necessary, I just like it to decrease the step size in the late game simulation.
