Rigidbody.AddExplosionForce doesn't affect rotational speed

When i use Rigidbody.AddExplosionForce on my rigidbodies with mesh colliders (meshes and colliders are all convex) the effect is a very nice explosion with parts flying and rotating just as expected.
but when i remove the colliders, the rigidbodies only gets a force and flies away but with no rotation/torque. so it looks very bland.

is this the expected behaviour ?

 

Sounds correct to me. A rigidbody is described by position, velocity, angular velocity, mass and moment of inertia. The the first three are subject to simulation, the mass can be set in the rigidbodies inspector and moment of inertia is calculated from all child colliders. If you don't have a collider, the moment of inertia tensor would be 0, which would result in an infinite angular velocity for any given non-zero torque.. i guess this is why physx locks rotation on an axis with zero moment of inertia.

Disclaimer: I didn't take this from the documentation but just from my personal understanding of things.

 

Sorry...,
I actually tried this out and I'm wrong with my assumption, that the moment of inertia would be zero. It actually defaults to (1,1,1), so rotation is not locked. I verified with AddTorque and it works.

I guess AddExplosionForce does something like looking for the closest point from the explosions origin to the collider and adds force there, which in most cases is a ray the does not intersect with the center of mass, so it also induces torque which makes the object spin.

Anyhow, the general behaviour makes still sense to me because without a collider, the "closest point" to the explosion origin probably defaults to the center of mass and adding force there doesn't induce torque.

But.. I was wrong before, so I could be wrong again

 

thank you for your response. yeah i think what you said in your second message makes sense.
Now i use AddForceAtPosition to get even better effect at first i thought AddExplosionForce would work similar to AddForceAtPosition but it's not.