Hi, I come across this video:Tool_001: Magnetic Threads Test

http://www.grasshopper3d.com/video/tool001-magnetic-threads-test?xg_source=activity [grasshopper3d.com]

It was done using grasshopper n kangaroo I think… Any one know how to do this in houdini?

I attached my houdini file.

Thanks a lot.

You could do something like that with magnet metaballs & wire solver.

Needs refinement to look good, and if you want precise control of the force fallback etc, you should probably build a force vop with a pointcloud iterator inside.

Hey eetu

Thank you so much for the file. I applied on my model, it works. But what I want is not really every line attracted towards the center to the metal ball, I need each wire itself to be a magnet, so they each have magnetic field (only attraction force), so that they can attract each other according to the distance in between the lines and then achieve an equilibrium state.

Do you think that is possible in Houdini?

Thank you~~~

Well, that's what it's trying to do already, it copies magnet metaballs too all the points of all the curves. I presume it's a matter of falloffs - if the falloff is too gradual, then all the points will be drawn to the general direction of all the rest, eventually the middle.

That's why I brought up the VOP Force, where you could define the falloff more freely. Just loop thru all the points, and determine a force vector for each, and add them all together to get a compound force vector for each point. I tried this very quickly and it didn't function at all - but it sounds good to my ears

LOL, I will try n see whether that works. Another question, how can I make sure that the metal balls on one line do not attract each other but only attract the metal balls on the other lines?

I didn't figure out how that could be done with the metaball approach, but they seem to cancel each other pretty well.

Doing that in VOPs should be doable, but is not straightforward. I have no time to try now, though.