Goal Position on packed objects.

[derjcmp 26]

Hello everyone!
So I've seen this vid and was wondering how to create it. 

I'm able to do it with grains and the targetP. The problem with grains is that it only supports spherical objs. 

Is there a way of doing this with RBD or bullted packed prims ? 

Thank you in advance

[Jesus Nieto 7]

Hi João, I would probably try to do this with cone twist constraints and some transforming representative geometry for the dancer, since it would be to heavy to simulate with deforming mesh. You can use a sop solver inside of DOPs to break the constraints if they exceed certain angle.

There is a very good class on cone twist constraints in CG Society's - Destruction in Houdini

http://www.cgsociety.org/training/course/destruction-in-houdini 

[derjcmp 26]

10 hours ago, Jesus Nieto said:

Hi João, I would probably try to do this with cone twist constraints and some transforming representative geometry for the dancer, since it would be to heavy to simulate with deforming mesh. You can use a sop solver inside of DOPs to break the constraints if they exceed certain angle.

There is a very good class on cone twist constraints in CG Society's - Destruction in Houdini

http://www.cgsociety.org/training/course/destruction-in-houdini 

I've taken that couse 1 year ago or something. Im scattering points on a geometry and then doing an attributeInterpolateSOP so they stick to it. Then Im instancing simple cubes.

How'd you go about turning those points on constrains? 

[Atom 1,438]

After playing around with this concept all day I got some help from Jake Rice on Discord on how to manage the RBD collisions much better.

ap_RBD_scatter_person_JR.hiplc

Edited April 28, 2017 by Atom

[derjcmp 26]

18 hours ago, Atom said:

After playing around with this concept all day I got some help from Jake Rice on Discord on how to manage the RBD collisions much better.

ap_RBD_scatter_person_JR.hiplc

Hey atom!
Thanks for wasting time trying to help me, big fan of your work!
The setup you sent me is almost the same as I did before. I used an attribute interpolate to make the particles stick to the surface though.
On the bullet solver I did the same Position expression as you did but on the time I tried it, it didnt go as well... Maybe I need to add a POP speed limit or smth because its spinning alot.

I'm going to try it out with the velocity approach  aswell! Thanks for the file!