23-11-2020, 08:49 AM
(This post was last modified: 23-11-2020, 08:49 AM by josemendez.)
As stated in the manual, the solver performs all simulation in local space. All contact, particle, and constraint values are expressed in the solver's local space. This:
Will transform a world space position to the rope's local frame. Neither the origin or the destination vector spaces are correct. It should be:
I'll assume the box is a rigidbody: your issue is that you're attaching the rope too close to it (or inside of it), so particles overlap its geometry. The result is that attachment and collisions cause a force feedback loop, since neither can be solved. This situation, along with its solution, is described in the manual. See:
http://obi.virtualmethodstudio.com/tutor...aints.html
The grab script is designed with kinematic rigidbodies or simple colliders in mind. In this case, simply attaching the rope slightly offsetted from the collider surface would suffice.
(22-11-2020, 07:01 PM)datzme Wrote: Vector3 hit = rope.transform.InverseTransformPoint(contact.point); // this didnt work at all.
Will transform a world space position to the rope's local frame. Neither the origin or the destination vector spaces are correct. It should be:
Quote:Vector3 hit = boxCollider.transform.InverseTransformPoint(rope.solver.transform.TransformPoint(contact.point)); // solver to collider space.
(22-11-2020, 07:01 PM)datzme Wrote: I also tried to use the Grab script, it looks good, but it will go crazy kinda fast:
https://i.imgur.com/zogiZZZ.gif
I'll assume the box is a rigidbody: your issue is that you're attaching the rope too close to it (or inside of it), so particles overlap its geometry. The result is that attachment and collisions cause a force feedback loop, since neither can be solved. This situation, along with its solution, is described in the manual. See:
http://obi.virtualmethodstudio.com/tutor...aints.html
The grab script is designed with kinematic rigidbodies or simple colliders in mind. In this case, simply attaching the rope slightly offsetted from the collider surface would suffice.