Hello, I'm having some difficulty figuring out when I should use RBD or Bullet as a rigid body solver. I'm noticing that I get drastically different results depending on whether or not the object is pre-fractured or fractured dynamically. The fracturing parameters are identical for both the dynamic and pre-frac'd geometry.

With BULLET as I increase the metaball strength my cube simply flies higher on the pre-frac'd geometry. I don't understand why larger magnet forces pop the box up faster and don't shatter the pre-frac'd geometry. Bullet works fine on my dynamic fractured geometry however.

With RBD, it works as expected for dynamic but pre-frac ONLY works if I set ‘solver per object’ in the options. (Without solver-per, the box does not break at all and with my inner glue strength at 10 and my magnet force at 4000 one would think that it would break apart no matter what.)

Does anyone have any thoughts on why Bullet seems to work so oddly with pre-frac'd geom and why I need a solver per object?

Thanks!

Hi strafer,

I believe the reason is because under the hood, bullet breaks glue and does fracturing based on collision impacts, but the Magnet DOP only applies a force, and since no impact data is being applied to the simulation, bullet is treating the prefractured cube as one object.

I'm still wrapping my head around DOPs myself, but does that make sense?

Chris

*Edit* One work around would be to move the prefracture box ever so slightly into the ground plane, so it starts out colliding… though there must be smarter, more DOP-centric methods to do this…

Ahh I see, thanks for the response! I wasn't aware bullet did not take into account magnet forces for breaking glue. The result makes much more sense now.