+- 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: A question about colliders. (/thread-86.html)
A question about colliders. - kkansy - 28-08-2017
Hello!
I'm trying to move ObiRope on a staircase model. Until now I was using a mesh collider created from stair mesh for the collisions but the rope tends to pass through it. I guess it's caused by over-complicated collision mesh. I wonder what would be better approach for stairs collider: a composition of Unity boxes, transformed so that they are matching the staircase shape or a custom collision mesh, a simplified staircase model without so much detailed steps.
I would be grateful for an advice in this matter.
RE: A question about colliders. - josemendez - 28-08-2017
(28-08-2017, 09:06 AM)kkansy Wrote: Hello!
I'm trying to move ObiRope on a staircase model. Until now I was using a mesh collider created from stair mesh for the collisions but the rope tends to pass through it. I guess it's caused by over-complicated collision mesh. I wonder what would be better approach for stairs collider: a composition of Unity boxes, transformed so that they are matching the staircase shape or a custom collision mesh, a simplified staircase model without so much detailed steps.
I would be grateful for an advice in this matter.
Hi!
MeshColliders generally are best avoided if it is possible to approximate the shape with primitives. This is because a) they're much more expensive to collide with b) once a particle gets inside of them (either because the collider was transformed, or because there were more contacts for a given particle that frame than collision iterations) it is not possible to project it back outside.
You can also try increasing the amount of collision iterations in the solver, to perform more accurate collision resolution. However in your case, approximating the staircase with box colliders would be a much better approach.
cheers!
RE: A question about colliders. - kkansy - 28-08-2017
(28-08-2017, 09:56 AM)josemendez Wrote: Hi!
MeshColliders generally are best avoided if it is possible to approximate the shape with primitives. This is because a) they're much more expensive to collide with b) once a particle gets inside of them (either because the collider was transformed, or because there were more contacts for a given particle that frame than collision iterations) it is not possible to project it back outside.
You can also try increasing the amount of collision iterations in the solver, to perform more accurate collision resolution. However in your case, approximating the staircase with box colliders would be a much better approach.
cheers!
Great, thank you
