Buoyancy restoring calculations are wrong when using graded buoyancy.

[arjo129]

Environment

• OS Version: All
• Fortress and Garden.

Description

Given an example world with restoring moments uniform buoyancy and graded buoyancy provide very different results.

Steps to reproduce

I have attached a simple world here:
https://gist.github.com/arjo129/ce5e553407163f32572bbfd2dc990740

This world has a setup where we have the center of buoyancy below the center of gravity.

As one can see the center of buoyancy (yellow being the buoyancy mesh) is below the actual center of the box which in turn is the center of gravity.

When using graded buoyancy the world seems stable.

In practice according to common knowledge the center of buoyancy of a submarine should always be above the center of gravity for the world should be unstable:

This is consistent with the behavior of uniform buoyancy: