Bonded and jiggling

[YoyoBoy]

Does anyone know how to go about doing an animation like the little animated GIF on the right side of this page: http://en.wikipedia.org/wiki/Temperature. Mine will just be a few water molecules, but I'm a bit clueless how to make it happen. Oh, and they need to break apart eventually too.

Thanks.

[super gnu]

well the manual way, if its just a few, would be to simply link each sphere to the one before it, and apply a noise to the rotation of each in trackview.

of course then you would also have to deal with the breaking apart...

[YoyoBoy]

Rotation! Yes, that's the key. I was thinking position, since that's technically what is happening, but couldn't figure out the linking, but it all really boils down to rotation. Thanks.

[YoyoBoy]

Hmm... on second thought, that won't quite work, since they will all be a slave to the root of the chain, and it'll look super weird.

[super gnu]

looks like that is what is happening in the wikipedia example?

[YoyoBoy]

Sort of, but there doesn't appear to be a hierarchy in there. No real center, or branching tree of influence.

[super gnu]

maybe im wrong, but if they are all connected, and all rotating relative to each other there must be a root node..? cant see it though youre right. maybe it could be disguised by putting some random motion on the root..

[super gnu]

but yes i suspect looking more.. its more complex.. if you do small rotations on all of them starting at the root, youd end up with the end ones going mental.

some kind of bones and ik might be a better bet with random motion on each node.... unfortunately youve reached the limit of my paltry animation knowledge here.

[YoyoBoy]

I think the key might be to turn off rotation inheritance for several key nodes.

[cheerioboy]

After linking, try also adjusting the pivot point to the center of the parent - see my tiny test file attached (max 2012) moleclue_test.zip

The main molecule is linked to a dummy, any molecules touching the main one are linked to the main one (and I suppose would just continue onwards) & their pivots are aligned to the molecule down the link chain. Then noise controllers are assigned to the position of the main molecule, and noise controllers are assigned to the rotation of the secondary molecules. And tweak

[YoyoBoy]

Hey thanks! I think I'm onto it now, and this helps a lot!