Search found 78 matches

by Nathanael
Thu Jul 17, 2008 1:30 pm
Forum: General Bullet Physics Support and Feedback
Topic: Minimal effort for (deterministic) state rewind
Replies: 4
Views: 3017

Re: Minimal effort for state rewind

It depend on your needs, if you just rebuilt your scene from scratch on reset, there must be only few issues (look out for time accumulators, random seeds, etc...) but it should be possible and efficient. related threads: http://www.bulletphysics.com/Bullet/phpBB3/viewtopic.php?f=9&t=1630&hilit=dete...
by Nathanael
Tue Jul 15, 2008 11:14 am
Forum: Research and development discussion about Collision Detection and Physics Simulation
Topic: Alternatives to EPA
Replies: 12
Views: 4766

Re: Alternatives to EPA

Does XenoCollide now return the correct penetration distance and not just an conservative estimate? Last time I looked at it this was the case and Gary also agreed on this on the forum. I'd like to know that too, last time i implemented it, it was an approximation, but to be fair, in most polyhedra...
by Nathanael
Sat Jul 12, 2008 4:13 am
Forum: Research and development discussion about Collision Detection and Physics Simulation
Topic: Alternatives to EPA
Replies: 12
Views: 4766

Re: Alternatives to EPA

I don't think EPA is that complicated, considering that it work on all class of convex shapes (that included quadratics, ie: ellipsoid/ellipsoid ). The core of the algorithm is basically an incremental convex hull builder, the current bullet implementation of EPA (btGjkEpa2.cpp) is less that 300 lin...
by Nathanael
Tue Jul 08, 2008 2:14 am
Forum: General Bullet Physics Support and Feedback
Topic: Coupling SoftBodies with my rendering engine...
Replies: 15
Views: 7175

Re: Coupling SoftBodies with my rendering engine...

Vertex positions are stored in btSoftBody::Node::m_x, so: btSoftBody* body=...; const btVector* positions=&body->m_nodes[0].m_x; const int stride=sizeof(btSoftBody::Node); /* Update positions in your display mesh */ Triangle are stored in btSoftBody::Face, so you can retrieve vertex index with: btSo...
by Nathanael
Mon Jul 07, 2008 11:03 am
Forum: General Bullet Physics Support and Feedback
Topic: Broadphase / Midphase / Narrowphase optimization
Replies: 5
Views: 2688

Re: Broadphase / Midphase / Narrowphase optimization

Bullet already provide a wide range of broad-phases implementations: - btAxisSweep3 (SAP) - btMultiSapBroadphase (multiSAP) - btDbvtBroadphase (dynamic hierarchy) Perhaps you should try existing implementations, then make sure broad-phase is actually a bottleneck. btMultiSapBroadphase especially can...
by Nathanael
Mon Jul 07, 2008 10:52 am
Forum: General Bullet Physics Support and Feedback
Topic: rope elongation
Replies: 2
Views: 1445

Re: rope elongation

What's really important is the mass ratio between the rope and the body, so you can increase the rope mass to improve stiffness, and/or iterations, (rope are fast to solve, try 32).

Hope it help,
Nathanael.
by Nathanael
Sat Jun 28, 2008 1:55 am
Forum: General Bullet Physics Support and Feedback
Topic: winding rope on a windlass
Replies: 9
Views: 2755

Re: winding rope on a windlass

Try increasing rope mass, masses ratio are really the weak point of the type of solver used for soft/rigid interaction.
Also, increase ...m_cfg.piterations , try 16.

I'm interested to try it myself, can you post the code?

Nat.
by Nathanael
Fri Jun 27, 2008 11:26 am
Forum: General Bullet Physics Support and Feedback
Topic: New features of Bullet (dynamic AABB trees)
Replies: 2
Views: 1261

Re: New features of Bullet (dynamic AABB trees)

You need to compile/run CDTestFramework in /extras folder.
More information here: http://www.bulletphysics.com/Bullet/php ... =18&t=2178

hope it help,
Nathanael.
by Nathanael
Fri Jun 27, 2008 3:28 am
Forum: Research and development discussion about Collision Detection and Physics Simulation
Topic: Flutter and Tumble: The Physics of Falling Paper
Replies: 6
Views: 5534

Re: Flutter and Tumble: The Physics of Falling Paper

can you point out some links that say how it all works? Just applying (over)simplified lift and drag on a deformable model do the tricks, of course in the real world the paper is not that deformable, and its motion is mainly driven by the turbulence in the fluid, but it work well enough (run the de...
by Nathanael
Fri Jun 27, 2008 3:02 am
Forum: General Bullet Physics Support and Feedback
Topic: winding rope on a windlass
Replies: 9
Views: 2755

Re: winding rope on a windlass

The rope you created is 50Kg and the cylinder 10Kg, not a fair match under gravity! :wink: In addition, the last parameter of ::CreateRope is 'fixed', its used to attached node to the world (bit mask, 1:first node , 2:last node). try: btSoftBody* pRope = btSoftBodyHelpers::CreateRope( pdemo->m_softB...
by Nathanael
Mon Jun 16, 2008 2:59 pm
Forum: Release Announcements
Topic: Bullet 2.69 BETA1 Physics SDK released
Replies: 5
Views: 13053

Re: Bullet 2.69 BETA1 Physics SDK released

I'm interested in the implementation of the dynamic AABB tree, including the ability to do view frustum culling and front-to-back traversal, but I'm having real problems finding the little sucker. A search for "frustum" in CDTestFramework doesn't result in much useful info, let alone an obvious exa...
by Nathanael
Thu Jun 12, 2008 3:41 pm
Forum: General Bullet Physics Support and Feedback
Topic: Missing factor 2 in PSolve_RContacts (btSoftBody.cpp) ?
Replies: 2
Views: 1462

Re: Missing factor 2 in PSolve_RContacts (btSoftBody.cpp) ?

If I am right, for a (1-dimensional) elastic collision, the difference in impulse is 2(va - vb) / (1/ma+1/mb). va-vb is vr in the code, 1/(1/ma+1/mb) is c.m_c0. So I would expect a factor 2 in front of vr? What is modeled here is inelastic collisions, through in line 1027 the term 'cti.m_normal*(dp...
by Nathanael
Thu Jun 12, 2008 12:14 pm
Forum: General Bullet Physics Support and Feedback
Topic: Sofy body exploding
Replies: 8
Views: 5197

Re: Sofy body exploding

@bob The mesh you describe consist of 24 vertices's, btSoftBody doesn't weld anything, a cube should be made of 8 vertex, 12 triangles. soft bodies expect consistent triangle orientation, normals pointing outward (triangle ABC, normal (B-A)x(C-a) ), and duplicated vertex mean disjoint during simulat...
by Nathanael
Fri Jun 06, 2008 4:11 pm
Forum: General Bullet Physics Support and Feedback
Topic: Sofy body exploding
Replies: 8
Views: 5197

Re: Sofy body exploding

@bob Can you provide some code extract so that we can reproduce the problem? @Bbilz Soft bodies is still a very new feature, it will take some time before things settle in, right now too much changes occur to enable real documentations. Here some help regarding btSoftBody::Config : btSoftBody::Confi...
by Nathanael
Thu Jun 05, 2008 12:34 pm
Forum: General Bullet Physics Support and Feedback
Topic: Softbody-RigidBody Collisions
Replies: 2
Views: 1414

Re: Softbody-RigidBody Collisions

Is this the correct and wanted behavior or it's my fault and I'm missing some option?
Collisions handling for soft body faces / rigid body shapes are currently in active development.

Thanks.