23-04-2021, 08:43 AM
(This post was last modified: 23-04-2021, 08:46 AM by josemendez.)
You're attaching the rope too close to the collider. If you add a ObiParticleRenderer to the rope, the reason for this will become obvious: the colliders are pushing the rope out, even though it's attached to them.
This situation along with the solution (which is to simply disable collisions between the offending particles and the colliders using phases, as you discovered) are described at the end of this page:
http://obi.virtualmethodstudio.com/tutor...aints.html
-------
The gap in the last image you posted can be due to the fact that input is normally updated after physics (Update() is called after FixedUpdate()). This will result in a 1-frame delay that can cause gaps. The solution is to store input and apply it during the solver's step, in the next frame.
Also, large mass ratios between the rope and rigidbodies attached to it (eg, a heavy object attached to a light rope, or the other way around) can require more pin constraint iterations for attachments to appear stiff enough.
let me know if you need help. cheers!
This situation along with the solution (which is to simply disable collisions between the offending particles and the colliders using phases, as you discovered) are described at the end of this page:
http://obi.virtualmethodstudio.com/tutor...aints.html
-------
The gap in the last image you posted can be due to the fact that input is normally updated after physics (Update() is called after FixedUpdate()). This will result in a 1-frame delay that can cause gaps. The solution is to store input and apply it during the solver's step, in the next frame.
Also, large mass ratios between the rope and rigidbodies attached to it (eg, a heavy object attached to a light rope, or the other way around) can require more pin constraint iterations for attachments to appear stiff enough.
let me know if you need help. cheers!