Real-Time Physics Simulation Forum

Thanks for the quick reply Dirk. I will try to make this as clear and detailed as possible for a simple distance joint For the distance joint constraint, I do this: P.zero(); // Zero the accumulated impulse PreStep() : C = |xa + ra - xb - rb| - InitialDistance A = n.dot(InvMa*n + SkewRaT*InvIa*n*Ske...

So far, I have the following equations for the soft constraint general form JM^-1JT*lambda*dt = - Jv - beta/dt*C - gamma*lambda Lambda is currently a force To turn lambda into an impulse, i multiply by dt (JM^-1JT*dt + gamma)*lambda' = -Jv*dt - beta*C beta = ERP = (k*dt)/(c+k*dt) gamma = CFM = 1/(c ...

For anyone who is interested, just as I was about to give up, I thought of something that fixed the problem to an extent: 1) Increase the thickness of the clipping planes by increasing the plane epsilon to 10^-2 (after experimenting with various epsilon values). 2) DO not clip against the face plane...

Tuning the bias factor didn't do anything. m_cdContactData.contacts[i].bias = k_biasFactor / dt * AMG3D_MATH_MAX(0.0f, m_cdContactData.contacts[i].sPenetrationDepth - k_allowedPenetration); I calculate the bias above where the penetration depth is +ve and the allowed penetration is 0.01, with a bias...

Ok, let's forget about the relative velocity for a minute and use two tangent directions by btPlaneSpace. Now I only have to compute the masses seen by the impulse only once per frame since they no longer depend on the tangent vectors which depend on relative velocity. My stack is stable now with wa...

What about the normal and tangent masses? Are these pre-computed once per frame before the impulse iterations, or for the tangent impulses, they need to be computed at each iteration since the relative velocity changes?

I couldn't paste my applyImpulse code for some reason. I assume you calculate friction at each iteration rather than once per frame as it depends on relative velocity

I have identified a problem related to friction in my impulse solver and don't know what's causing the problem. I disabled friction, enabled warm starting, accumulating impulses and position correction and things work fine. Can someone tell me what the problem is? Implementing friction in 3D is tric...

For my hinge joint, I have one anchor given in world coords and a world hinge axis. From the world anchor, you build two local anchors as follows for each body wt1T.transposeOf(m_pRigidBody1->m_matWorldTransform); wt2T.transposeOf(m_pRigidBody2->m_matWorldTransform); m_vLocalAnchor1 = (vAnchor - m_p...

Converting from world to local just involves multiplying the world transform matrix of the object by the inverse to get the identity matrix. Normally you want to transform a different object B into the local frame of object A, which just involves multiplying B's world transform matrix by A's inverse...

Found a problem with this approach, for anyone who may run into this problem... Testing against 6 triangles generates more contacts than necessary, leading to jittery physics. An example is two contacts on neighbouring triangles are too close and cause double the impulse in close proximity. I have n...

Hi, I am representing the orientation of a rigid body as a quaternion and the angular velocity as a vector. I am attempting to implement an angle constraint directly from the quaternion without having to store a vector alongside the quaternion in the rigid body class to represent the orientation. C ...

Hi, I am having trouble with implementing a revolute joint between two particles. I am representing the particle as having a position, velocity and acceleration and no angular dynamic components. So the constraint derivation is as follows: C = xa + ra - xb - rb CDot = va - vb CDot = [ 1 -1] [va vb]T...

That's true! It is interesting how large games such as GTA handle a large amount of collisions in a huge world. Perhaps things like using really simple primitives like bounding spheres to keep objects from overlapping each other when they're out of view or too far from the camera for the detail to b...

Solved, used a grid to partition the terrain and simple AABB tests to narrow the collision down to 6 triangles and finally applied SAT to generate contacts between the terrain triangles and OBB.

Very fast for anyone who is trying to find a way to do OBB-terrain collisions with physics response