Added Rigid Body Method "get_inverse_inertia_tensor"

[yrk06]

Exposed inertia tensor to GDScript

How

This was done by creating a variable on the rigidbody class. The variable receives the inverse inertia tensor value each time the RigidBody state gets changed, then the get_inverse_inertia_tensor() returns the the inverse_inertia_tensor. The method is exposed to GDScript with bind_method

Consequences

As a consequence of this, rotations with rigid bodies can be controlled more accurately through the use of Physics Formulas of Torque

Requests

This was proposed in: godotengine/godot-proposals#1114 also in #2125 and kind of related to godotengine/godot-proposals#945

[Calinou]

Also, pull requests should be opened against the master branch unless it can't be applied to the master branch for some reason. We can then cherry-pick the change into the 3.2 branch if it doesn't break backwards compatibility.

[yrk06]

Also, pull requests should be opened against the master branch unless it can't be applied to the master branch for some reason. We can then cherry-pick the change into the 3.2 branch if it doesn't break backwards compatibility.

I tried to make the change using the Master branch, but I couldn't get the engine to run on my local computer. That's why I decided to go with the 3.2 version

[madmiraal]

The inverse of the inertia tensor is used for two reasons:

1. For convenience: To calculate the rotational acceleration, we need to multiply the torque by the inverse of the inertia tensor.
2. For StaticBodies: The inertia tensor needs to be infinite, which is not possible. However the inverse is zero, which is possible.

Inverting the zeroed inertial tensor of a StaticBody, or a RigidBody in Static mode, would result in division by zero.

In short, don't invert the inverse of the inertia tensor. If you want to provide access to the inertia tensor, return the inverse instead.

[yrk06]

I see, thanks for pointing out. I didn't know that. I knew only about the convenience for calculations, but in order to calculate torque, the not-inversed is needed, thus the function was to get the not-inversed. I'll change that and commit it

[madmiraal]

Looks good, aside from my nitpick below, the commits just need to be squashed into a single commit.

[madmiraal]

Nitpicking, but I think this would read better if it didn't include the word "certain" i.e. something like:

This is used to calculate the angular acceleration resulting from a torque applied to the RigidBody.

[yrk06]

I force pushed the commit again in order to change the method name in the commit from get_inertia_tensor to get_inverse_inertia_tensor

[madmiraal]

I'm was just looking at that. This should have been created as a property with a getter and a setter.

[yrk06]

I'm was just looking at that. This should have been created as a property with a getter and a setter.

I didn't do it because the idea was just exposing using the already existent function on class PhysicsDirectBodyState in C++ to RigidBodies in GDscript

I Haven't looked yet on how the engine calculates the Inertia Tensor, but my guess is that it would require a new function inside the Bullet Physics Classes in order to Overwrite it with a Setter

[madmiraal]

Okay, let's leave it as it is then.

One more change, please update the commit message, which still has all the individual commit messages to just the first commit message. Also, commit messages should be written in the present tense; so 'Add Rigid Body Method "get_inverse_inertia_tensor"'.

Finally, please create a second PR against the master branch referencing this one. If you've thoroughly tested it on the 3.2 branch, you should be confident that it will work on the master branch without being able to run a test project.

[akien-mga]

Thanks!