Joint questions

The use case I have is water buoyancy for boats. I have hulls that are stable on the water surface but need to attach the rest of the boat body.

The main thing is the hull mass needs to not be effected in terms of how it reacts when applying impulses to it. Or another way of looking at it is impulses applied to the hull alone should produce the same result as when a body is attached assuming no collisions are present. Because the buoyancy implementation only works with impulse points spread over a single plane. It doesn't have the same type of stability as say a wheeled vehicle.


An alternative to joints might be finding a way to calculate what the force should be at each impulse point based on mass distribution. But not sure that would even work so joints just seem like a possibly simpler solution.

 

So joints seem to work well for this. A solution based on distribution of force over the mass evenly would be nice if it's simple enough, but that seems unlikely.