+- Obi Official Forum (https://obi.virtualmethodstudio.com/forum)
+-- Forum: Obi Users Category (https://obi.virtualmethodstudio.com/forum/forum-1.html)
+--- Forum: Obi Rope (https://obi.virtualmethodstudio.com/forum/forum-4.html)
+--- Thread: Does new update have any freedom for tight knot (/thread-199.html)
Does new update have any freedom for tight knot - chanddu - 17-10-2017
Hi jose,
Chanddu again,I got your rope update and thanks for the update, I was trying the rope for knot which will be used in my suturing thread simulation,Can I use this????
Thanks in Advance.
Chanddu.
RE: Does new update have any freedom for tight knot - josemendez - 17-10-2017
(17-10-2017, 10:45 AM)chanddu Wrote: Hi jose,
Chanddu again,I got your rope update and thanks for the update, I was trying the rope for knot which will be used in my suturing thread simulation,Can I use this????
Thanks in Advance.
Chanddu.
Hi there!
3.2 mainly focuses on the collision system against colliders.
The solver itself is still iterative and particle-based at its core, so knots will always end up being undone if tight enough.
Suturing thread has very specific simulation needs (high tension, tight knot formation, almost inexistent stretch tolerance), so most likely you will need a custom solver to simulate it. General rope solvers such as Obi just won't be able to deal with these requirements.
Academic research on this area seems to be aimed at alternating follow-the-leader type solvers coupled with segment-based collision detection, instead of particle-based. Some even have special cases for knot handling (such as detecting knots and rigidbody-fying them). See:
http://robotics.stanford.edu/~latombe/papers/knotmaking/paper.pdf