|
|
physical property - behaving madly
About Truespace Archives
These pages are a copy of the official truespace forums prior to their removal somewhere around 2011.
They are retained here for archive purposes only.
physical property - behaving madly // Archive: Tech Forum
Post by Burisman // May 24, 2007, 12:52pm
|
Burisman
Total Posts: 128
|
I was experimenting with physical properties and created an oddly shaped object. Gave it the iron property, and reduced its flexibility in LE. I drop a block on it (also iron). The large object (which must have an eccentric centre of gravity) tumbles and starts to bounce a bit. But despite the fact that I expected it to stop and drop dead, it just starts turning and bouncing like a rubber ball gaining more energy. But where from? Perhaps there is someone who likes to analyze which parameter is responsible for this odd and counter intuitive behaviour? I'm puzzled. :o |
Post by weaveribm // May 24, 2007, 10:30pm
|
weaveribm
Total Posts: 592
|
...it just starts turning and bouncing like a rubber ball gaining more energy. But where from?
Gaining or perhaps retaining?
If no friction from the floor or air resistance that might explain the object retaining its kinetic energy which would be forever without friction
Anvils On Ice
Wait- bouncing? That's not on was your last scene a rubber room Jack Nicholson iron blocks have no lift component :)
Peter |
Post by frootee // May 25, 2007, 4:37am
|
frootee
Total Posts: 2667
|
Hi Burisman. Couple things to try.
Increase the polygon count on your objects (subdivision). Also, try creating a thin, very wide cube which is right on top of the truespace grid. I do not think the grid itself has physical properties, but if you assign physical properties to the cube (i.e. iron with high friction as weaveribm pointed out), your object should lose energy as it interacts with the cube (ground surface). Also make sure the cube you make is subdivided. More vertices/polys = improved physics simulation, since I believe the simulation relies on this information for calculations.
I have not looked at the phys properties yet, but (as in 6.6), there may be an set of checkboxes for Vertices or Vertices/Faces. Use the checkbox that will enable both.
Also, when you make changes to the physical properties, rewind your simulation and delete the old one prior to replaying. Otherwise, if your new sim is say, 500 frames long, but the old one was 1000 frames long, the frames 501-1000 from the old sim will still replay if you let it play long enough, which would also result in funny looking behaviour.
HTH!
Frootee |
Post by Burisman // May 25, 2007, 7:38am
|
Burisman
Total Posts: 128
|
Peter & Frootee, thanks both of you :)
I just took the grid as being a floor, but it iisn't, :o it has indeed no physical properties. Somehow the effect is that of a super-trampoline :rolleyes:
Adding a real floor to the scene immediately resulted in a more predictable and natural behaviour.
Also increasing the polygon count, as you predicted.
And of course increasing friction.
:) |
Post by frootee // May 25, 2007, 9:24am
|
frootee
Total Posts: 2667
|
Burisman,
I am glad to be of assistance, so .... TOUCHDOWN!
One other very important thing to consider; I just remembered this from earlier forum posts: if you want to use a ball for whatever reason in phys sims, use a geosphere, not a regular sphere. I think the reason is, the vertices on a geosphere are uniformly distributed, as opposed to a regular sphere. Add a geosphere and a regular sphere to confirm this (I should do this as well so I can make sure I am telling the truth here). If they look the same, I am wrong, but use a geosphere anyway. It works better. :p
I think the Simulating the World Course topics still hold true in 7.5.
HTH!
Frootee |
|
