next »

[umen]

I ported simple Box2d Explosion simulation the function is called once when there is some event with the "Truck Car"
the problem is that the Explosion working fine but it looks like when the explosion done the "parts" of the car Returning to
the initial car position instead of bean scatter on the floor.
what am i doing wrong here ?

here is the code for the Explosion

Code:

class MyQueryCallback : public b2QueryCallback {
public:
	std::vector<b2Body*> foundBodies;

	bool ReportFixture(b2Fixture* fixture) {
		foundBodies.push_back(fixture->GetBody());
		return true;//keep going to find all fixtures in the query area
	}
};
void BackgroundSprite::explode()
{
 		float m_blastRadius = 40.0f;
		float m_blastPower = 50.5f;
		MyQueryCallback queryCallback;
		b2Vec2 locationPnt = _truck->truckBody->GetPosition();
		b2AABB aabb;
		aabb.lowerBound.Set((locationPnt.x - m_blastRadius), (locationPnt.y - m_blastRadius));
		aabb.upperBound.Set((locationPnt.x + m_blastRadius), (locationPnt.y + m_blastRadius));
		m_world->QueryAABB(&queryCallback, aabb);
	
		for (int i = 0; i < queryCallback.foundBodies.size(); i++) {
			b2Body* pBody = queryCallback.foundBodies.at(i);
			b2Vec2 pBodyPos = pBody->GetWorldCenter();
			b2Vec2 pHitVector(pBodyPos.x - locationPnt.x, pBodyPos.y - locationPnt.y);
			float radDist = pHitVector.Normalize();
			radDist = (radDist * PTM_RATIO);
			if ((pBody->GetType() == b2_dynamicBody) == (radDist <= m_blastRadius)){
				 
				float nHitForce = ((m_blastRadius - radDist) / m_blastRadius) * m_blastPower;
				b2Vec2 appForce(pHitVector.x*nHitForce, pHitVector.y*nHitForce);
				pBody->ApplyLinearImpulse(appForce, pBody->GetWorldCenter(),true);
			}
		}
 }

[BorisTheBrave]

You have to manually undo joints. (though I think there was a patch floating around for doing it automatically).

Either delete all the joints on affected objects, or delete any joints that have large reaction forces on them. The latter means you have to wait a step before you can delete the joints, which has an odd looking effect on the remaingin debris.

[umen]

Thanks for the replay , i trying to delete the joints but i got exception doing it.
Apparently there was still Physics going on while i tried to delete it .
how can i know when is the right time to delete  it ?

[oahda]

Thanks for the replay , i trying to delete the joints but i got exception doing it.
Apparently there was still Physics going on while i tried to delete it .
how can i know when is the right time to delete  it ?

You can't modify stuff within the Box2D update loop. Run all simulations for the current frame first and perhaps within that loop you might have to set flags to indicate which objects should have their joints removed, or put them in a vector or something, and then after the simulation loop, go through those objects and remove their joints.

[umen]

This is exactly what im doing after the world->step(..)
i loop all the joints to remove with this function ( all joints pointer in vector )

Code:

for (int i = 0; i < carJointsVector.size(); i++){
		_world->DestroyJoint(carJointsVector[i]);
	}

but it throws exception

[oahda]

Dunno if it matters since I wrote this such a long time ago now, but I have this after my step loop:

Code:

m_world.ClearForces();

You could try it out if you're not already doing that.

[umen]

ok thanks for the help , in the end the problem was that i updated other parts of the car after i removed the joints
so once fixed that every thing was working.