CGTalk - Explode effect!

Als

07-18-2007, 09:07 PM

Take you polygon object, extract with keep faces together set to off, but still keep it one object.
Go!
Put field or whatever, you have explode effect!

Or other method to have goal object:

Copy your object before it becomes ncloth. Then you can constraint facets component to component, with connection order -> components, and weld, then animate target strength.
Force value has to be something low like .01, and negative force will atract object, positive will push it away.
Goal value has to be very low. If you set it to 1, it will instantly smack object to position.
Now animate non ncloth object and ncloth will follow!

Play with spring and rubber options, and different values for force, strength, tangent and glue.

Try following values:
constraint method weld
component relation - all to all
connection method - component order (very important)
connection update - per frame
strenght .05
tangent 0
glue strenht .01
glue scale 1
force - keep zero until you need to draw object, then animate from 0 to -.2
On ncloth object play with input mesh atract value!
Add turbulance field for shaking object effect...
etc.

Enjoy!

Als

Aikiman

07-18-2007, 11:02 PM

kewl, the rigid body one didnt really work for me. I tried animating the mag of a radial field from 500 to 0 from frame 1 - 3 but the solver choked. But if you combine the faces together and create an nCloth, it works like a dream.

You can also now use Mayas crack to break your cube up randomly, keep applying the crack command to the bigger pieces then eventually use 'combine' and apply nCloth! Takes a little more time for nucleus to solve but turn off self collisions and its way faster than the old solver.

Thanks :)

Duncan

07-24-2007, 12:08 AM

Als... it sounds like you are basically doing the equivalent of a "tearable surface" constraint. In terms of the glue strength behaving differently when the constraint strength is greater than 1(as mentioned in another thread), it could happen because the constraint will then internally do another iteration, so there can be jumps when animating strength values, especially when cross over 1.0(higher integer boundaries are more subtle... the difference of 21 iterations vs 20 for example instead of 2 vs 1). In some tests I've not seen much difference in glue values for a stretch of .99 vs 1.01, but your setup could be different.

Duncan

Als

07-24-2007, 11:59 PM

Thanks Duncan for reply.
Main thing with this setup from above, (I will try to post the scene when I greb a bit of time) is that glue value is not changing enough from like .1 to .9, but values above and under make drastic changes. It is quite hard to animate the value therefore.
I will post the file to illustrate...

Thanks

Als

Duncan

07-25-2007, 01:41 AM

A couple of things to keep in mind when using the glue attribute:

It is very sensitive to various changes. The point at which a bond breaks is often close to a critical value, so that one must sometimes fuss with the attribute to get the desired effect. Changing other aspects of the simulation may change the required glue setting.

Also if you animate the glue value up to 1.0 then any bonds that may have broken get rejoined.(for example ripped fabric would get pulled back together... rather quickly)

Duncan

Als

07-25-2007, 10:02 PM

Thanks Duncan.
I'm aware of this, so I'm saying it would be good to have bigger range there so that one can make animation between the two values. It's hard to animate it this way.
I hope this makes sense.
Also I would be nice to have some indicators on all the values in ncloth what are optimum and what "maximum" values... Well, maybe not only in ncloth ;)

Thanks

Als

CGTalk Moderation

07-25-2007, 10:02 PM

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.