Testing Dynamics

[Gene]

Hi All,

 

Attached is a project I started to see how far I could push Dynamics in AM.

 

There are still some issues with it as can be seen in the attached vid. You really have to play with the settings and in some cases I found if I select Constraint (under Target) and try to change it to None, it does not always change....like it becomes stuck??

 

Also, when selecting "Attach to Parent" on or off in bones mode and after Dynamics is turned on...AM crashes

 

Finally, odd things are being simulated, like Objects tend to gain energy after repeated collisions with another object.

 

So far, it appears to work best for really simple stuff.

 

Cheers,

 

Eugene

DynamicsTest_3.mov

[Gene]

and the project...

Dynamics4.prj

[nf1nk]

Its not as strange as it looks, The model is acting as though the wires have a great deal of mass, and the spheres are acting as though they have very little mass, it looks odd because we are used to seeing things set up where the spheres have significantly more mass than the wires, also we expect the wires to flex, or if not, the joint between the wire and the weight would stay rigid.

[Gene]

Hi Nathan,

 

I tried to increase the MASS value of the balls in increments up to 5000 and the wires/rods are at 0.01.

 

You can see the results below...nothing really changes.

 

Also, I noticed a couple of other things...depending on the Mass settings, the Collision radius seems to be change. That is, that the first time the balls hit, everything seems in the right place, but the subsequent hits, the balls surfaces might not be touching or in other cases, the surfaces pass through each other.

 

Can someone tell me if the collision radius is calculated from the start or root of the bone?

 

Also, maybe there is something in my setup which could be improved for better accuracy? I was trying to keep this simple, but perhaps the rig is really important?

 

Thanks,

 

Eugene

DynamicsTest_4.mov

[Gene]

Here is a shot of the bones/rig.

 

Thanks,

 

Eugene

[AngryOne]

I havnt done any Dynamics stuff yet, but it seem like the model has a pointed edge around that is causing it to bounce weird. Maybe try a sphere to see if it bounce right. I coud be wrong.

 

 

Also when I look at ur video in slow motion it seem the left one is hitting the right one above the center, which I think is causin that problem.