Just one more idea on that (having your picture in mind): I would create 6 phantom boxes (deep enough, e.g. r_max*2) that would englobe the red volume but wouldn't interact (in the sense of putting forces on them) with spheres.
Then you would be able to estimate (rather precisely) the overlapping volume: 1. in sphere-sphere interaction, it's sum of volumes of 2 spherical caps: doable analytically, since interaction stores penetration depth and you know radii of both spheres. 2. in sphere-box interaction (spheres only partially within the red volume), do the same. You just take one sphere cap. 3. Disregards spheres that don't overlap with red volume. In this way, you have Yade collider doing all the hard work (contact detection) and you just take what it calculates. The most difficult would be to have colliding boxes that don't create any forces then, but that's entirely programming question. Vaclav _______________________________________________ Yade-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/yade-users
