Real-Time Physics Simulation Forum

Erwin,

What are the conditions in which you trigger activation? Is it just when an active body hits an inactive one, or do you handle separation as well?

My gut feeling is that a triangular or quad mesh with perfectly rigid distance constraints will end up bending only along the edges, corresponding to the tri or quad edges, provided that the elements are regular such that collinear edges may be found along the cardinal tri or quad edge directions. T...

I've seen this algorithm also termed as "sort and sweep" which would imply that "sort" is being used as an alias for "prune". So off the top of my head I'd say that "prune" refers to the sorting of the projection ranges along one or more axes, while "swee...

I have so far tried pre-computing penetration depths (positive inside, negative outside) and associated normals for the eight vertices of each AABB node constituting the BVH I am using to organise the triangles. For each vertex I basically locate the closest enclosed triangle and use its distance fr...

I have a polygon soup geometry object whose elemental triangles are organised using a BVH. I have implemented reasonably robust collision detection code with boxes, spheres, cylinders and even convex polyhedrons, based on the fact that the latter geometry is a fully closed oriented manifold for whic...

The friction law you are applying is known for increasing the energy of a system in some cases. Try using lower coefficients of restitution and friction. Alternatively, you can also try computing the energy before and after the collision, and lower the velocity accordingly. If your linear and angula...

In my own physics engine implementation I sample 8 points on the cylinder end circumference hoping to get a reasonably shaped manifold when placing a cylinder on a box. Seems to work ok in practice.

Assuming you have polyhedral geometry (like the cubes in your case), you could try building a simple aerodynamic model as follows: For every face: 1) compute a face centroid.. for a triangular face this can by done by taking the mean of the vertices 2) compute the body's velocity at the centroid 3) ...

Is centripetal force in your example the same thing as applyTorque()? And how is applyTorque() different from setAngularVelocity()? Also, does applyTorque() use the center of mass to apply the torque around? "outward from the curved path while turning". I am not sure how to translate that...

You could try keeping high wheel friction (to prevent over-realistic drifting) and instead apply a centripetal force at the car's centre of mass, outward from the curved path while turning. The magnitude of this force could be calculated as a function of linear and angular speed and triggered after ...

Yes.. obviously with the downsides I highlighted.

What about a partial approach, that is, just solving unconstrained motion? Arguably you might see jittering or even temporary tunnelling if lag times are large enough, but at least you should be able to get fluid motion between status updates.

I don't wish to deviate this topic further, so I'll close by saying it really looks very impressive and is very reminiscent of the original (missile-equipped jeep, chopper, RIB, submarine inflitration etc.). I wish you guys the best of luck!

Slightly oot, are you the guy who worked on Navy Moves for the Commodore Amiga in the '80s?

Glad to hear it worked :) The trick is to realise that a collision may be formulated as a constraint - one that you apply when two bodies are trying to interpenetrate. This approach is particularly useful when using a simultaneous solver (such as LCP), but it also works fine with a sequential solver...