Real-Time Physics Simulation Forum

Thanks, I _think_ I understand -- if you had a chain of bodies ABCDE connected by joints AB, BC, CD, DE you could then solve [AB,CD] in one batch and [BC,DE] in the next? That's pretty smart! You mentioned that you used 10 batches -- the number of batches would have to depend on the number of constr...

In this month's GD mag there's a paid portion from Intel that talks about this simulator: http://www.infernalengine.com/tech_physics.php They specifically mention that it can solve an island of constraints in parallel; collision detection and Jacobian building are more easily split up, because each ...

Erin Catto wrote:In the end they had to write an entire floating point unit in software using integer operations. It was too late to convert the code to fixed point.

I might be showing my ignorance, but would the former really take less effort than the latter? That seems crazy!

Is there any chance of a simple explanation of "variational methods"? So far I've read the thesis twice and have yet to understand it well enough to implement..

Concerning (1), as long as you download the version that's on his site, you should have the most recent one. AFAIK the main difference is that the old version uses cardinality ("n") in the formulas for delta-p.

Box2D's soft distance constraint is using impulses rather than forces, hence the 1/h term. Of course, I'm sorry -- I forgot that ODE was forces/LegrangeMultiplier... I've been away from this stuff for about a year :( With a frequency/damping ratio model you just set the damping ratio to 1 to get cr...

Hi, I'm having trouble trying to add a critically damped harmonic oscillator (spring/PD-controller) constraint to my simulator, and the references I've found are somewhat conflicting. Currently I'm using regular PD controllers outside of the solver (i.e the controllers are driving motor torques) and...

@Rob: AFAIK Box2D does this by using shrunken shapes for CCD.

That makes sense.. I guess I'm just screwed!

We have dealt with this in some games by generating 'future' contact/non-penetration constraints that happen at the time of impact. So they are generated before bodies are penetrating, perhaps similar to some of Trinkle-Steward approaches. Have you considered this? I'm not sure what you mean by &qu...

There should be no difference between 1st order velocity-level or 0-order position-level constraint solving in combination with CCD and CP. In both cases, at the time of impact, a discontinuity happens and contact constraints have to be solved, and the direction of motion typically changes. This me...

Thanks; I have implemented CCD for rope, and it works well, but was simplified by the fact that I'm treating the rigid bodies as read-only until the next frame. Having the rope's topology known (like having a non-changing contact graph/island) also simplifies things. Here's a single simulation step:...

It would be great if someone could explain their CCD process in a bit more detail; I still don't understand how you avoid doing n^2 tests: finding the earliest collision requires a full collision pass, which tests everything vs everything.. and since a full pass is required after each collision (to ...

I'd love to hear how this solver works, I remember trying to read the "regularized variational methods" thesis and failing completely to understand

KenB wrote:Distance maps are used for free-form objects. They are quite efficient in 2D.

It seems like you're calculating them really quickly.. are there any particular tricks, or are computers just fast enough these days?