Accretion

[kleer001]

Ok, basic geophysics here.

Particles are moving towards a collision. Once a particle collides it is sticky. A moving particle will stop when its within a small threshold distance from a sticky particle.

Think coral where polips stick to established colonies but not to other free floating particles.

How the frak do I do this?

[pclaes]

Is this the same as aggregation?

Sounds like a sopsolver with pointclouds to me. two groups, active points and static points. You deform the points within the sopsolver to make them move around. You do a lookup against the static points with a certain minimal radius.

If I have some time I would give it a go, but not at the moment.

[kleer001]

I'm not exactly sure about the technique you're describing, but it's something like that.

I did figure it out though. It always amazes me how quickly a solution comes just from asking.

It's pretty straightforward, if a little unintuitive.

The key is to write out the generated collision geometry and then read back in the previous frame and use that as a collision.

See attached.

aggregation.hip

[pclaes]

The key is to write out the generated collision geometry and then read back in the previous frame and use that as a collision.

This is exactly what the sop solver does, except there is no IO, the geometry of the previous frame is kept in ram. Sop solvers, vops and python sops can be a very powerful combination to tackle growing effects. Vops for speed, python for adding points and sop solver for keeping a "history". (In some of my previous posts I tend to use sop solvers for a variety of things - I'm sure if you invest some time in them it will open up a world of possibilities.)