nParticle self collision event

[John Cassella]

after a cursory search,  I can't seem to find any documentation on if  Nparticles  self collisions will actually show up as  collision events to the maya collision event system.

it seems not to,  so I was just wondering if I'm doing something wrong? or if this is not  something that is supported. 


thx

-johnc.

[Jeremy Raven]

Never seen that one myself. Collision events normally refer to
particles colliding with objects rather than self collision.

[Duncan Brinsmead]

You can’t get self collision even info currently. However perhaps this thread will help you:

 

http://forums.cgsociety.org/showthread.php?f=86&t=925162

 

Duncan

[John Cassella]

Cool. thats what I figured duncan,  thanks for the verification.  I can't use some of the  other examples  in that cgSociety  thread because I'm dealing with geometry that is goaled and moving.   I think I'm  just going to  create  another nparticle object  and use that to throw collisions  that should do the trick.

thanks for the info

-john

[a a]

I was about to make feature request about exactly this early today (but forgot, now i remember).
The problem here is when using the SPH solver (particles self-collide there all the time) and trying to detect collisions with other geometry - it is impossible, but badly needed.
It is essential to be able to emit detailed secondary splashes where the main liquid mass smashes onto something.

[Jeremy Raven]

can you use the emit command to emit particles into its own shape?
probably not but that would be ideal in this case at collision world
position.

[Duncan Brinsmead]

You should be able to do what you are after… it sounds like you want collision events between the water and the geometry it collides with, rather than the internal sph “self collisions”(which aren’t exactly self collisions anyways).  Collisions between particle systems or particles and rigids or cloth all generate events. Self collisions could generate an enormous number of events if we included them(especially if sph self interaction was included). It could be a drag on simulation speed. However it might be feasible at some point to optionally flag particles that had collided the last step.

 

Duncan

[a a]

Thanks Duncan.
It does not work, at least here.
I expect the event to happen only when particles collide with other objects, but i get it for all SPH particles that contact or overlap.
Or maybe i don't understand you right.

[Duncan Brinsmead]

Does the attached (maya2011) file work for you?  It is a simple scene with a collision event where some sph nParticles collide with a torus and die and emit into a second nParticle system where they collide. Note that it helps to keep the two particle systems from colliding with each other(either by turning off collisions on the second system or using a collide exclude constraint), otherwise the collision between systems will generate more events.