I've noticed on several occasions in RH5 that if I have two surfaces that intersect, I'm unable to split one with the other. But if I create a curve from the intersection, I can use that as the cutting object, and the split works okay.
It would seem that Split, using two surfaces, would use the same math as Intersect to find the cutting curve, so why the difference in behavior?
Hi Phil- if you have an example, please send it to me.
Thanks, Pascal. Attached are two surfaces. The side surface cannot split the top one, even though it completely intersects it. This, however, is only a partial example of the observed behavior, since the curve created with Intersect is incomplete and it, also, cannot be used to split the top surface. Unfortunately, I'm unable to backtrack my work to the point where one split failed and the other succeeded. When it happens again, I'll be sure to save my work and post it here.
Both your surfaces are somewhat problematic. Your vertical "side" surface is pretty funky on one end, it folds over on itself - turn on control points and see.
But that doesn't seem to be the principal problem, it's the other surface - it has an internal G1 non-curvature continuous part and the intersection falls on the "joint" (which is where it fails). This type of thing probably should succeed, but that is why it fails currently.
To see, if you run the command DivideAlongCreases on the top surface and chose SplitAtTangents=Yes, you will see that the surface splits into a polysurface at that point. Explode the polysurface and remove the small end part (was that created by ExtendSrf using the line option?). Now re-extend the surface using the Smooth option. Your intersection/split will then succeed. If you re-extend it with the line option, it will fail as before.