Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion

Tags in this Discussion

Physic's Joints stability.
  • Hello. I have some pipe that is created of small peaces of actors which join by physx joints(spherical). The problem happens when this pipe leans on obstacle(plane or other object). In that case actors start tremble. How can i decrease that trembling? Thanks.
  • 7 Comments sorted by
  • Can you post a PVD capture here?

    Check your gravity to see if they are too big compared to your PxTolerancesScale. For example, you're using gravity(0,-400,0) (instead of 9.8) but you're still using the default PxTolerancesScale(1,1000,10) which should have been PxTolerancesScale(40,1000,400).

    Check if you joints are set up correctly. So it only jitters if you lean against an obstacle? Does it jitter if the pipe doesn't touch anything. You might want to crank up the PxRigidDynamic.SolverIterationCounts if the joint itself is not stable. Check the masses of the objects in the joint, they should not be a huge difference.
  • PVD capture? What is it? We use physx 2.8. now. Yes it jitters only if i lean against obstacle. I set solverIteration of actors in different values but that didn't help. Masses of joint's objects are similar.
  • PVD?
    PhysX Visual Debugger. Can capture PhysX scene and than replay it. Can be downloaded from dev. center, "PhysX Tools" folder.

    Also, there was a great topic about 2.8 joints on old forum
    http://forum-archive.developer.nvidia.com/index.php?showtopic=2420
  • Hello again. I've made dump from Visual Debugger(2.8.3). See attach.
    Physx dump.zip
    940K
  • Any thoughts guys? The problem is that we push a pipe by other object. If it would be free that jitter doesn't happen. Is it possible that pipe doesn't shake in that case or not?
  • You are having one object having a mass of 1.111E-06 jointed to another object with a mass of 8000.

    Please try to keep all the jointed actors' masses within 1 to 10 ratio.

    You can try to set all the masses to 1 first, and see if the pipe still jitters.

  • It's a little bit strange but when we replaced all joints to D6 joints then it became more stable. We used spherical joints before. For resistance we are using parameter of drive. And now it's much better and almost doesn't shake.