An oddity I saved while working.
Subtract the inner surface from the outer surface. It “succeeds” but the areas at the four corners do not get subtracted.
Problem Difference.3dm.zip (438.9 KB)
Only one object in the file…
Let me upload again:
Problem Difference.3dm.zip (271.0 KB)
the problem is that you get ambigious intersections:
I would solve it this way
Problem Difference_solved.3dm (2.8 MB)
Yes, but it silently goes forward to produce really odd results.
I see, yes it involves a bit more work, this one should be better:
2nd_attempt.3dm (1.3 MB)
I was working on fixing that last night by replacing the sweep with a couple of extrudes. Then I was going to make the hole with extrudes as well.
It was just odd to me the behavior I got here and it suggested that their might be an underlying bug.
I see, did the info I posted solve your question? If you are trying to make boolean differences between objects that have curved surfaces that are almost coincident, there will be no clear closed loop intersections, and the boolean will fail.
I see you got the answer. I don’t see how you got it.
I have a solution of my own.
I ended up extracting the surfaces that were created when doing the boolean split with the surface, then joining the two open polysurfaces and capping the end result.
FYI I always prefer to have split surface at tangents/kinks. For this you can use
_DivideAlongCreases SplitAtTangents=Yes
Feel free to post you solution so that others can benefit from it.