View Full Version : Help me about rigid body simulation!!
06-23-2008, 06:53 AM
I have a complicated shot that about 100 pieces of card gather one point and make 1 card.
I wanted to try to make scene that crash card to pieces, then bake it and reverse animation curve.
Then there's some problems.
1. channel box show me strange value like "1.#QO " and gone somewhere.
2. Simulation is almost good , then I save scene and open again.
Then gravity field dosen't work!! why?
3. Sometime when I run bake simulation, looks very heavy and stop caluclating and make keys with same value.
about Question 1, I guess it caused by ginbal lock.
but I don't have idea to avoid it.
thanks for reading.
I'm not good at using English.
If possible could you teach me simple sentence.
(Of cource I'll try to read however it's difficult to read.)
06-23-2008, 07:08 AM
Turn self collisions off and make sure you dont have extremely high values anywhere if you can help it. I have found if you have values like initialSpin X = 2000 on multiple rigids, then the solver can choke up and stop or cause unexpected results.
06-23-2008, 07:24 AM
thanks for reply.
I truned self collisions off and baked just one piece.
but same result....
I tried the situation that you told me (initialSpin X = 2000 on multiple rigids).
Then as you told the solver stoped.
I write more about detail of my scene.
I use a radial field (magunitude 40) to crash card and gravity(magunitude 3) field for fall them down.
And I changed the rigid body status bounciness 0, damping 1.5, static and dynamic friction to 0 from the default setting.
06-23-2008, 08:58 AM
Dont use any initial spin and lower your magnitude of your radial to 5, see what happens. If it still locks up the solver then perhaps try using nCloth rigids if you have nCloth. I find Mayas rigid solver very unpredictable.
06-23-2008, 09:14 AM
thank you for reply.
I did along your advice, then almost good!
but that motion is not my wish..
I have to convert to 3ds max, so I'm not sure be able to translate completely if I use nCloth.
It's really difficult problem.
06-23-2008, 09:53 AM
Go to your rigid Solver and decrease Step size to 0.10, now increase you radial magnitude until you find a happy medium. I dont like Maya's rigid solver to be honest, its very hard to get good results.
06-23-2008, 10:13 AM
"Step Size had already lower than 0.1.
But I find "Rigid Solver Methods" and change that to Mid point from "Runge Kutta Adaptive".
I didn't know "rigidSolver" tub.
I adjust few statuses then looks good, almost bake animation.
Now all my problem has gone but the problem gravity field has no effect when I start again from saved file!!
Thank you so much Jeremy!!
06-23-2008, 07:47 PM
Your step size could possibly go down as far as 0.01, I think default for 2008 is 0.03.
As for gravity, try creating a new field and connecting it up again. Also make sure the connection is there by checking the Dynamics Relationship Editor.
06-24-2008, 02:11 AM
Thanx for your suggestion.
Finally I changed Step size to 0.0001.
I didn't know the function of stepsize ,now I make sense thank of you.
It's so heavy to set stepsize to 0.0001,but simulation almost good.(some pieces gone right in the middle of simulation but it's really tiny peices so I'm sure that we can't see it after rendering)
About gravity field, It's looks kind a bug?
Ah , thats exactly what I guess...
I have to create new gravity field and connecting everytime when I start up maya...
It's not so hard but draw same falloff curve.
(When I duplicated gravity field and connecting it again , there's different result but I don't know why..)
Anyway I could done this shot last night(almost today's morning (A^_^;)) thank of you.
Thank you so much Jeremy!
06-24-2008, 02:11 AM
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.