Real-Time Physics Simulation Forum

Thanks for your reply and sorry for the late reply.
My apologies, but your post appears to be spam since this forum is about rigid- and soft-body dynamics, and users here would be know that Newton Dynamics refers to http://newtondynamics.com/forum/newton.php. You should know this from signing up.

Thanks for posting this! I too have been keeping a watch on this and I really appreciate the folks from Nvidia for their openess and contribution about their key methods - now with even a hands on approach. It's a shame the number of view is just in the thousands and that these forums have kind of d...

Did the Newton developers release any paper that describes their approach? Looking at GitHub and the wiki there isn't really much information available.

So Nvidia introduced a PGS derivative called TGS with much improved convergence rate over their old PGS solver and adding support for much larger mass ratios. Videos look impressive, esp. for difficult constraint types like prismatic, and so far some testing (IIRC our very own Erwin Coumans found so...

The final simplex that GJK/EPA terminates with has a point, an edge or a triangle that faces the origin. If it's a point, then the witness points on the involved shapes maps directly onto each of their individual support points. Contrary, if it is an edge or a face, then you will first have to calcu...

I get a feeling that you are referring to Nvidia's unified particle physics paper, but I'll link to it anyways: http://mmacklin.com/uppfrta_preprint.pdf Section 4.4 discusses the issue in PBD you bring up using a small number of pre-stabilization iterations, specifically how this problem becomes wor...

This might not be popular, but while I have not seen much discussion on these forums about the old Karma engine I think it still deserves a special mention. MathEngine's user guide for Karma gives a thorough, although dated, introduction to dynamical simulation for entertainment purposes. You can st...

From my own experience basing the bounce velocity on the previous frame's velocity makes the simulation much more stable for restitution How do you do this exactly? This sounds like a good solution. Apologies, forgot to emphasize that this is in my own physics engine and not in Bullet. It was done ...

The combined restitution is .4 x .9 = .36, so within the second bounce, not accounting for linear damping, the energy is down to 13% (and by E=0.5mv^2 that is less than 2% of the original velocity) so you need to up the coefficient of restitution. Have you tried enabling continuous collision detecti...