ObiRigidbodies (or ObiColliders?) causing weird physics

[khalvr]

I have recently had issues with a multimeter object in my scene. It consists of a rigidbody with a mass of 2kg, attached to two probe cords which weigh 10 grams in each end (the closest ends are attached with FixedJoints to the multimeter). The cord visuals are simulated using ObiRopes with a handle attached at each end, and spring joints to enforce the distance constraints. My problem was that despite the large weight difference between the probes and the multimeter, the probes would exert an unrealistically large pulling force on the multimeter, for instance when you dropped it on the table, ultimately pulling it down. After much searching i found that the problem was caused by the fact that the multimeter had an ObiRigidbody and an ObiCollider attached to it (the probes had none). Whenever i removed the Obi components from the multimeter, the problems would go away, and reappear if i added them. Now, i get that the Obi physics world is simulated separately from the PhysX world, but i still don't get why the ObiSolver perpetuates the motion of objects despite this being handled by PhysX. Ideally, ObiRigidbodies would only be used for exchanging forces between a rope and a Rigidbody and not interfere when there are no pin constraints used.

Also, i'm having serious issues with rope particles passing through/getting stuck in walls, despite the wall colliders being up to a meter thick (see attachments).

[attachment=81]
[attachment=82]

[josemendez]

I have recently had issues with a multimeter object in my scene. It consists of a rigidbody with a mass of 2kg, attached to two probe cords which weigh 10 grams in each end (the closest ends are attached with FixedJoints to the multimeter). The cord visuals are simulated using ObiRopes with a handle attached at each end, and spring joints to enforce the distance constraints. My problem was that despite the large weight difference between the probes and the multimeter, the probes would exert an unrealistically large pulling force on the multimeter, for instance when you dropped it on the table, ultimately pulling it down. After much searching i found that the problem was caused by the fact that the multimeter had an ObiRigidbody and an ObiCollider attached to it (the probes had none). Whenever i removed the Obi components from the multimeter, the problems would go away, and reappear if i added them. Now, i get that the Obi physics world is simulated separately from the PhysX world, but i still don't get why the ObiSolver perpetuates the motion of objects despite this being handled by PhysX. Ideally, ObiRigidbodies would only be used for exchanging forces between a rope and a Rigidbody and not interfere when there are no pin constraints used.

Also, i'm having serious issues with rope particles passing through/getting stuck in walls, despite the wall colliders being up to a meter thick (see attachments).

Hi there,

Without taking a look at your scene it is pretty much impossible to figure out what's happening. If the multimeter has no ObiRigidbody attached to it, the ropes should not be able to interact with it at all. It is as you say: ObiRigidbodies are only used to exchange impulses between PhysX and Obi, for collisions and pin constraints.

If you could send the scene to our support address (support (at) virtualmethodstudio.com) I can take a look at it and diagnose the problem.

cheers,

[khalvr]

OK, I will see if i can put together a sample scene for you.