Hi,
I got my first vellum solver simulation running well.
I use vellum configure grain as an input.
I would like to replace my particles/spheres with other geometries.
I'm not really sure where to start, if you have any suggestion it would be great.
Thanks

- use Copy To Points SOP and make sure you either copy variants of packed prims or have Pack and Instance checked

- you can alternatively also use s@instance or s@instancefile attributes


just keep in mind that grains dont compute any rotation or p@orient attribute so the orientation may be based on v@v which is not ideal
so either initialize orient using
before copying or compute the values based on motion or something to fake rolling or whatever you are after

Keen to piggyback on this post!

I'm facing a similar challenge. I'm trying to achieve an object raising up through a bucket of randomly shaped pieces, think a ball lifting up through a basket of lego. I've setup a grain sim (both using vellum grains and pop grains to compare) and have managed to copy to points and replace the geo with my various geometries however the collision boundaries for the solver are still calculated using the original particle sphere when using the copy to points method.

Is there a way to plugin the geo before the solver so that the collisions are calculated using the shape of the new geometry? Otherwise the objects are always clashing or always have gaps.

Perhaps this is a scenario better setup with an RBD solver?

ajmorrison1
Keen to piggyback on this post!

I'm facing a similar challenge. I'm trying to achieve an object raising up through a bucket of randomly shaped pieces, think a ball lifting up through a basket of lego. I've setup a grain sim (both using vellum grains and pop grains to compare) and have managed to copy to points and replace the geo with my various geometries however the collision boundaries for the solver are still calculated using the original particle sphere when using the copy to points method.

Is there a way to plugin the geo before the solver so that the collisions are calculated using the shape of the new geometry? Otherwise the objects are always clashing or always have gaps.

Perhaps this is a scenario better setup with an RBD solver?

thinking pscale before making them grains should work? or maybe https://youtu.be/iCxkGeiXtfc [youtu.be]