Mass doesn't affect Sphere fall speed

[pccsoares]

I've been playing around with the HelloWorld example provided by the latest version of the library.
The demo has a ground plane and a sphere of a certain radius and mass that falls to the ground.

In the example, it has radius of 1.0 and mass of 1.0.
Gravity is set to (0, -10, 0)

Code: Select all

		btScalar mass(1.0f);

		btVector3 localInertia(0,0,0);
		colShape->calculateLocalInertia(mass,localInertia);

		btDefaultMotionState* myMotionState = new btDefaultMotionState(startTransform);
		btRigidBody::btRigidBodyConstructionInfo rbInfo(mass,myMotionState,colShape,localInertia);
		btRigidBody * body = new btRigidBody(rbInfo);

		dynamicsWorld->addRigidBody(body);

If I change the mass to an higher value, the speed of the sphere falling is the exactly same. I can't see any differences between having mass of 1.0 or mass of 10.0.
Although if I modify the radius, I can see the speed changing.

Is this normal? Am I missing something?

[billias13]

Check this video, it explains why

http://www.youtube.com/watch?v=KDp1tiUsZw8

[pccsoares]

you mean there's something wrong with the gravity? It is set as I have seen everywhere in examples: (0, -10, 0)
how can I make it right?

[billias13]

I mean that since there are no other factors such us air resistance (size of object) etc, the rate on which any object falls towards the ground (acceleration of gravity and hence speed) is independent of its mass.

[pccsoares]

you are right!

From what I read, bullet doesn't implement air resistance.
Is there a way to simulate this on my sphere?

[billias13]

I am not quite sure whether or not you can implement air resistance in rigid bodies, but i think that for trivial shapes such as spheres and boxes etc it would be pretty much negligible. However, there is the option of setting slightly different gravity vectors for each falling object to achieve different speeds (scientifically inaccurate, but you can give it a try to see if it produces the effects that you want).

[pccsoares]

thanks.

[billias13]

Sorry, i forgot about damping! You can use this to implement any kind of resistance (linear or angular)
Check btRigidBody::setDamping

[pccsoares]

there is another thing that I'm not understanding. I changed restitution and friction values of my sphere, expecting it to bounce once it gets to the floor. For example:

Code: Select all

		body->setRestitution(0.8f);
		body->setFriction(0.5f);

And these are the results of the fall of my sphere with radius 1.0, in X, Y, Z:

Code: Select all

world pos = 2.000000,2.289296,0.000000
world pos = 2.000000,1.876760,0.000000
world pos = 2.000000,1.463891,0.000000
world pos = 2.000000,1.050691,0.000000
world pos = 2.000000,0.927432,0.000000
...
world pos = 2.000000,0.985486,0.000000
...
world pos = 2.000004,0.999999,0.000014
world pos = 2.000005,0.999999,0.000014
world pos = 2.000005,0.999999,0.000014

It looks like it falls to the ground, compresses and returns to it's initial form.
I was expecting to see a bounce above the floor.
Can you explain this?

[pccsoares]

figured it out.
both bodies (sphere and ground) need to have a restitution set. obviously.