In the attached I have a sweep2 in blue and a closed extrusion in purple. The extrusion passes through the sweep twice. However, if I split (or Trim) the purple using the blue, I would expect to get 3 pieces. Instead I only get 2. The upper surface passing through the purple does not split.
This should intersect, but it is getting an incomplete result at the file tolerance of 0.00001. If you reduce your tolerance to 0.001 you will get a result.
Hi Jim - this is apparently a hard case for the intersector - it looks like there is a knot line that exactly matches the intersection for a short distance and this is where the gap appears in the intersection curve. I’ll get this on the pile for the developer to look at - it has the same problem in V6.
The real bug is that Rhino fails to identify object with knots piled close together as Bad Objects.
The failure to identify the bad objects leads to the failure to identify bugs in the Rhino code that create such objects.
I regularly check for near-degenerate knot spans in boundary curves by using a macro that duplicates all surface boundaries, converts to Beziers and checks for short curves. You have to do this every couple of hours and repair any edges that are bad to keep a model clean. There is just no excuse for McNeel stubbornly refusing to address this problem.