Hey,
- When I tried to move the head of a chain/rope (p0.gif), the translation propagates to tail a lot, makes the tail rotate a lot, while the tail should be stable or not be affected too much.
- When moving the tail, the chain simulation works well (p1.gif).
Do you know how to fix this?
Thanks,
(05-05-2020, 05:24 PM)ethan000 Wrote: [ -> ]Hey,
- When I tried to move the head of a chain/rope (p0.gif), the translation propagates to tail a lot, makes the tail rotate a lot, while the tail should be stable or not be affected too much.
- When moving the tail, the chain simulation works well (p1.gif).
Do you know how to fix this?
Thanks,
Can't be fixed, as you cannot control rope twist. Ropes do not physically model torsion resistance, but
rods do. See the comparison table between ropes and rods:
http://obi.virtualmethodstudio.com/tutor...setup.html
Since ropes do not model torsion, the twist of the rope is determined using parallel transport (
https://en.wikipedia.org/wiki/Parallel_transport): an arbitrary twist is selected for the first particle in the rope, then propagated trough all following particles in a way that it minimizes orientation change between consecutive particles. Depending on the rope shape and the orientation of the first particle, this can mean the overall twist of the rope can change quite a lot from one frame to the next, like in your first gif.
In few words: if you need/want to control twist, use rods instead of ropes.