What is the 'Form factor scales the inertia tensor' option?


Few questions:

  • For what forces it is designed for. Examples, if possible.
  • How to scale it? What defines the amount?
  • How the ‘Damp’ should set versus the ‘Form’?


Form has mainly to do with rigid bodies and their collision with things. You can see its effects best by making a plane and a dynamic cube, rotating the cube a bit, and start the gameengine. Higher values for “form” will have the cube “rotate” to align itself with the ground slower. I, unfortunately, never got what Inertia was in Physics class, so I can’t tell you exactly how the physics works.

The physics definition of inertia is that an object resists changes in its state of motion. Objects which are at rest want to remain at rest (lazy!) and objects which are at constant velocity want to remain at that constant velocity.


just found this thread… I’m also wondering about the EXACT meaning of this form factor thing and I cant find it really explained anywhere. As I studied physics I know well about what a mass and a moment of inertia is:

Mass is, on the one hand something that attracts other masses (gravity…) and on the other hand something that makes a body resist LINEAR motion. That means: heavy things are more difficult to lift AND to move. [that both are described by the same “mass” is btw the “secret” of general relativity theorie, bla, bla :D]

The moment of inertia is the property of an object to resist ROTATIONAL motion.

So: things with more mass are harder to move, things with higher moment of inertia are harder to rotate. The problem is just: in blender i can NOT set the moment of inertia (as I can set the mass), but just this “form factor”.

For example for a solid sphere the moment of inertia I is
I = (2/5) * m * r^2 (m: mass, r: radius)
for a solid cube (rotational axis parallel to some edge) it’s
I = (1/6) * m * a^2 (m: mass, a: edge length)

I see: (2/5) = 0.4 what is the default value for the form factor. So I first thought the form factor is just the forefactor here, m the set mass an r the set radius. But: the set radius has NO effect on the rotational behavior as long as proper bounds are set…

My assumption so far: I = “form factor” * “set mass” * “SOME LENGTH” ^ 2

So for me the remaining question is: what length? For a sphere it’s clearly just the radius, but for a cube it could be half of the edge length or half of the spatial diagonal or anything between.

Another question is still the one of objects with different moments of inertia for different rotations (basically everything thats not a sphere…)

Somebody here who has any ideas who might know this??


I think as long as the shape of the object will be concerned in some wise way in the calculation by the bullet engine its okay set form factors of

0.0 all the mass is in the center
0.4 the body is solid (with homogeneous density)
1.0 all the mass is in the surface (hollow body)

Everything above 1.0 seems to make no physical sense (It would make the body behave as if its mass where located outside its own boundaries :confused:).

The question of “what length” still puzzles me, but I think I will take these values for simulations :slight_smile:

Having a low, but non-zero value helps to keep objects more stable when stacked, having it at the maximum value will mimic the expected behavior you get when you turn off rigid-body, but will still rotate very slightly if it hits something.

If you google for “bullet physics erwin” you will find Erwin Couman’s “Bullet Physics Manual”. That might help you the find answers to you questions (I haven’t read it yet).

For more details you can ask Erwin.