Reactor Buoyancy or Flotation - How to control Center of Mass (CG)?

[klzsailing]

Hi, I'm trying to simulate flotation of boats using Reactor. I need to control the local position of the Mass Center (or Center of Gravity) for each object separately, to represent this phenomena in a more realistic way. Reactor locates the CG of each volume in its Volume Center, regardless of its internal mass distribution, that leads to a non realistic representation.

What I do to correct this situation is pretty nasty: Adding some little, heavy volumes nested into the main volume, to correct by "try and fail" the way this volume will float in reality.

So the question is: How do you correct or change the Mass Center (CG) position of any volume in Reactor?

Here a video to understand what I need:

https://www.youtube.com/watch?v=aFTQByZ4cx0

Thanks!

[samab · Accepted solution]

I've done this kind of thing before, though I don't have a version of Max with Reactor any more 😞

As you have already worked out Reactor treats any single object as a homogenous mass, which fixes the position of the COM.

So to alter the COM, as with a real boat, add some ballast. For example, in a motor boat there would be a big heavy engine somewhere low in the hull which lowers the COM. Or like the sail boat in the Youtube clip, it has a big weight at the bottom of the keel for stability.

The way reactor works is actually quite realistic, if you stick to a realistic scale with dimensions and weights and construct the model accordingly it should work out right.

So to get an uneven COM in practice you must construct the model from multiple parts of varying density, then group them.

So you have the hull modeled to the corect volume and assign a realistic weight to it (the hull only, not the entire boat). Add the engine, just a simple box of realistic size and weight. Add a weighted keel if there is one. Some boats may have ballast in the bilge, concrete or metal weights. Add any super-structure there may be too and whatever else there is.

If you keep it all to realistic size and weight, it should just work.

Of course use lo-poly proxy meshes for all this and link the render meshes to it.

[klzsailing]

Thanks samab, that's what I actually do.
So, volumes having non homogeneous mass distribution must be corrected the way you explain.
A movable COM could have been a better solution, anyhow.