i can absolutely not see anything that could cause the boolean split to fail, it is a simple extruded ring with 2 simple planes trying to split the ring.
simple boolean split.3dm (67.0 KB)
i can absolutely not see anything that could cause the boolean split to fail, it is a simple extruded ring with 2 simple planes trying to split the ring.
simple boolean split.3dm (67.0 KB)
The way BooleanSplit currently works it needs consistent normals for the splitters. Join the two planes first and then it splits fine.
I understand this is not optimal. RH-53103 is the issue to address this.
FWIW you can also just use the curves you used to create your splitter surfaces to do the cutting with the gumball instead of first extruding them to surfaces then BooleanSplitting. Saves you a step
thanks @Joshua_Kennedy
@nathanletwory nice, thanks. i have so neglected the new functions in the gumball that it hurts. cutting stuff like that is indeed a neat treat.
so this is why it fails with other booleans the bloody directions where facing inside, now i know how i should approuch booleans thanks Nathan!
How do you do this with the gumball?
With your planar curves selected on the gumball you’ll notice lines crossing the axes, click and drag with that
Check the section on the Gumball widget in our docs: Gumball | Rhino 3-D modeling
this forum is gold learned 2 things today
Really? Not in all cases. For example any of the green objects can be BooleanSplit by any number of the intersecting planes. Normal direction of the planes doesn’t matter…
BS-BS.3dm (5.0 MB)
Actually you’re right, the normals aren’t the problem in this case. Normals do need to be consistent for booleans. Inconsistent normals would be if those two planes were joined and one didn’t have it’s direction flipped. Rhino is pretty good about doing this correctly. BooleanSplit, and really all the booleans, process objects in pairs. A single planar surface may not split the torus into distinct pieces.
So when BooleanSplit processes the pairs neither plane on its own splits the ring but when processed together they would. The proposed fix in the issue wants to process all the splitters together. In all your examples each plane on its own splits the object into distinct pieces.
Sorry about that, I should have taken more time when responding to this. Thanks for challenging me here.