Inertia tensors units don't include MASS ??

[Verbo]

Hi,

I have a question about the inertia tensors. From my understanding of moment of intertia, the unit are Kg*m^2 and then angular acceleration = torque / moment of inertia.

So the we get the angular velocity by multiplying by the timestep. However, the Bullet inertia tensors don`t seem to take mass into account, so how can we get angular velocity if we use a moment of inertia tensor that only seems to be in meter? Like in the example below, from btRigidBody :

void btRigidBody::integrateVelocities(btScalar step)
{
...

m_linearVelocity += m_totalForce * (m_inverseMass * step);
m_angularVelocity += m_invInertiaTensorWorld * m_totalTorque * step; ---> how come we don't multiply the inv inertia tensor by the inverse mass?

...
}

Sorry if it sounds retarded to ask, I am just a little confused, and I am more than ready to understand the reason...because from what I experiment, Bullet physics works very well

Thanks guys.

Verbo

[Dirk Gregorius]

You should think of mass as resitance against linear motion and inertia as resistance against angular motion. The inertia tensor already contain the mass. Therefore an extra scaling is not needed.

See here:
http://en.wikipedia.org/wiki/Moment_of_inertia

[Verbo]

Hi,

Actually, I appologize for asking such a question at first, because I realized 10 minutes later that the collision shapes functions ::calculateLocalInertia() gets the mass in input in its first parameter.

Ouch...I thinks it's a clear case of friday+tired+read too much about physics.

But hey, thanks for the answer, you actually gave me one of the many links I kept reading and reading this week...he!he!

CAse close...sorry for the post, my mental inertia is too high today.

Verbo