Search found 57 matches

by jarno
Fri Sep 14, 2012 5:53 am
Forum: Research and development discussion about Collision Detection and Physics Simulation
Topic: Cloth collision detection-What is the currnt stateof the art
Replies: 12
Views: 22262

Re: Cloth collision detection-What is the currnt stateof the

Edge versus edge can actually be transformed to be similar to vertex versus face. In short, vertex versus face with linearly moving vertices is done by: Convert to vertex versus plane by determining the plane equation for the face Convert to origin versus plane by considering the problem relative to...
by jarno
Wed Sep 12, 2012 3:37 am
Forum: Research and development discussion about Collision Detection and Physics Simulation
Topic: Cloth collision detection-What is the currnt stateof the art
Replies: 12
Views: 22262

Re: Cloth collision detection-What is the currnt stateof the

I've been rewriting a lot of the Bullet softbody collision code redoing the vertex-face collision detection, adding edge-edge detection, supporting self-collision with the same, and sprinkling it with multithreading. Like the Brochu et al. paper I do the constant velocity approximation. I initially ...
by jarno
Tue Jun 12, 2012 6:36 am
Forum: General Bullet Physics Support and Feedback
Topic: Resetting scene
Replies: 3
Views: 4911

Re: Resetting scene

As well as resetting the broadphase pool, I also clean out the overlapping pair cache: btOverlappingPairCache* pair_cache = m_dynamics_world->getBroadphase()->getOverlappingPairCache(); btBroadphasePairArray& pair_array = pair_cache->getOverlappingPairArray(); for(int i = 0; i < pair_array.size(...
by jarno
Fri May 18, 2012 7:16 am
Forum: General Bullet Physics Support and Feedback
Topic: Bug Report with complete code to reproduce it
Replies: 6
Views: 7214

Re: Bug Report with complete code to reproduce it

If the bodies from planet 1 are never going to come into contact with plane 1 (or you don't care if they do), then indeed you do not need that plane. I assumed you were using the planes to confine the objects within their zones.
by jarno
Thu May 17, 2012 12:48 pm
Forum: General Bullet Physics Support and Feedback
Topic: Bug Report with complete code to reproduce it
Replies: 6
Views: 7214

Re: Bug Report with complete code to reproduce it

You could do it by using two planes at (0,500,0), one facing each way. Then use collision masks to exclude the plane the rigid body is expected to be inside of. O <- (rigid body on planet 2) ----------/\-------- <- plane 2 at 0, 500, 0 facing up -------\/----------- <- plane 1 at 0, 500, 0 facing do...
by jarno
Thu May 17, 2012 6:11 am
Forum: General Bullet Physics Support and Feedback
Topic: Bug Report with complete code to reproduce it
Replies: 6
Views: 7214

Re: Bug Report with complete code to reproduce it

Planes have an inside and an outside, like they are infinitely thick. The normal of the plane points to the outside. So your vehicle is actually trapped deep inside the top plane, which is registered as a penetration which it then attempts to get out of. Flip your top plane body so that it points do...
by jarno
Tue May 15, 2012 1:20 am
Forum: General Bullet Physics Support and Feedback
Topic: btConvexHullShape reutilization
Replies: 13
Views: 10967

Re: btConvexHullShape reutilization

Maybe you have a mixup between an object index and a collision shape pointer? I see that for obtenerMatrizRigidBody() you use sim as an index into the object array, while in Java_com_contimita_utils_BulletNdk_insertarDynamicConShape() sim is a collision shape pointer. If the objects are actually app...
by jarno
Sat May 12, 2012 6:45 am
Forum: General Bullet Physics Support and Feedback
Topic: Make btSoftBody more hard/rigid
Replies: 2
Views: 4847

Re: Make btSoftBody more hard/rigid

You could remove the bending links by iterating over all the links and looking for those that have the m_bbending value set. Or freeze them by modifying the linear stiffness of the bending link material to 1 and setting the rest length of the links to the current distance between the nodes, and unfr...
by jarno
Sat May 12, 2012 6:24 am
Forum: General Bullet Physics Support and Feedback
Topic: btConvexHullShape reutilization
Replies: 13
Views: 10967

Re: btConvexHullShape reutilization

As long as the local scaling and margin of the convex hull shape remains the same, I see no reason why it shouldn't work.
by jarno
Sat May 12, 2012 6:02 am
Forum: General Bullet Physics Support and Feedback
Topic: Bug in btRigidBody::applyImpulse's use of m_linearFactor?
Replies: 3
Views: 4313

Re: Bug in btRigidBody::applyImpulse's use of m_linearFactor

It does look like the code is a bit confused about whether the linear factor is a scale factor that is applied to the forces/impulses acting on the body, or only on that part that induces a linear motion. Usually the linear factor isn't changed from its default of (1,1,1), in which case it isn't an ...
by jarno
Mon Apr 09, 2012 7:38 am
Forum: General Bullet Physics Support and Feedback
Topic: Meaning of softbody anchor hardness
Replies: 0
Views: 2981

Meaning of softbody anchor hardness

I'm investigating some issues with the softbody anchor solving. I think there is a bug in the formulae used, but I have some difficulty trying to understand what the intention behind the code is. I understand that setting anchor hardness (softbody->m_cfg.m_kAHR) to 1 means that the node is expected ...
by jarno
Wed Apr 04, 2012 7:21 am
Forum: General Bullet Physics Support and Feedback
Topic: 100m drop at 9.81m/s: tunneling through btHeightfieldTerrain
Replies: 7
Views: 8184

Re: 100m drop at 9.81m/s: tunneling through btHeightfieldTer

After dropping 100m the cylinders will be falling at a speed of about 44m/s. With a timestep of 1/120s, that's about 36cm per timestep. Any collisions between objects with a combined thickness less than that (including collision margins) are likely to be missed. The cylinders are 20cm, while the thi...
by jarno
Sat Jan 21, 2012 6:24 am
Forum: General Bullet Physics Support and Feedback
Topic: Why is "solverConstraint.m_jacDiagABInv" updated again?
Replies: 2
Views: 4244

Re: Why is "solverConstraint.m_jacDiagABInv" updated again?

It's not updated again. They are two different sets of constraints. There are three sets of constraints involved: The contact constraints The friction constraints All other constraints (e.g. hinge, 6dof, point2point, etc.) The ones set up in solveGroupCacheFriendlySetup() are the constraints from th...
by jarno
Tue Jan 17, 2012 12:01 pm
Forum: General Bullet Physics Support and Feedback
Topic: Bullet 2.77 Restitution issue.
Replies: 8
Views: 11027

Re: Bullet 2.77 Restitution issue.

What is the exact definition of restitution in that code context? In that code snippet the restitution is the calculated desired speed after collision. In essence the current speed times the restitution factor set by colobj->setRestitution(). One could change the logic so that if the restitution fa...
by jarno
Tue Jan 17, 2012 6:09 am
Forum: General Bullet Physics Support and Feedback
Topic: Bullet 2.77 Restitution issue.
Replies: 8
Views: 11027

Re: Bullet 2.77 Restitution issue.

As this issue persists in 2.78 and I believe 2.79 as well, and I'm having the same problem, I traced through it all and found the problem. It's in btSequentialImpulseConstraintSolver::setupContactConstraint(), particularly this part: btScalar positionalError = 0.f; btScalar velocityError = restituti...