Real-Time Physics Simulation Forum

https://github.com/bulletphysics/bullet3/issues/126 I'd like to get Bullet 2 running with CPU multithreading but it has been removed from recent versions of the repository. I looked up the commit that removed it (f06312c632492200ba1478ddc84c5ee474f74e54, comment is "remove most clutter (todo)&...

I've run into the same sort of problem. I managed to tame the shakes somewhat by switching to the btGeneric6DofSpringConstraint for most things and btConeTwistConstraint for shoulders. I had to tune the spring strength and spring damping parameters.

Glad to hear it worked! Yes the default behavior -- colliding with the back faces of triangles -- doesn't seem very useful. It caused me lots of problems. It is really bad when you have a jointed object like a ragdoll and one of its collision objects manages to penetrate the ground or a wall deep en...

The heightfield collision shape is treated as just a thin polygonal surface. The terrain has no thickness to it. This can cause 2 issues: 1. fast moving small objects (without CCD enabled) can fall right through and never even detect a collision (just turn on CCD for such objects) 2. if an object en...

It is already possible to easily determine which collisions are new by inspecting the list of persistent manifolds. For each manifold there are a number of btManifoldPoints, and the manifold point has a "lifetime" value. If the lifetime is 1, then it is new. So, no need to add a flag for t...

The "force" variable could be renamed "accel" or something like that and it would make more sense. If the mass wasn't multiplied in there, then it would effectively need to be folded into the spring and damping constants. That's inconvenient because it would mean that if you chan...

So this code is part of calculating the suspension force for the wheel. Part of the calculation is done in this code, the rest is done elsewhere (nice and confusing). The "inv" variable is basically a scaling factor for the suspension force. When the vehicle is sitting level on flat ground...

Why has this not been included in bullet's main line? If I had to guess, I'd say it's because it doesn't add anything you can't already do, and it's bad for performance. I prefer to walk the list of persistent manifolds after each simulation step, and batch up events to send out to my game objects ...

To fix the objects penetrating your static meshes you could try using CCD (continuous collision detection). See bullet/Demos/CcdPhysicsDemo/CcdPhysicsDemo.cpp for an example.
However, I think you'll find that other problems will crop up if you go down the road of large, variable-sized fixedTimeSteps.