A simple intersection of two pipes fails. The pipes overlap with two intersections. When I use Intersect the curve of one intersection is correct, the other one has a large gap in the middle.
This case requires the tolerance to be increased in Options>Document Properties>Units. At 0.0005 inches a complete intersection is found. Does that help?
Thanks. That helps.
Brian, now the intersection works, I checked I also increasing the tolerance, but the Boolean operation of union fails (I closed surfaces and transformed into solids). Why? Is a bug?
seems that the Boolean union in this case works with the tolerance set to 0.01
If I remember correctly Rhino in these situations changes the tolerance automatically to accomplish the task, but why not in this case?
Thanks violine, I did as you told me, it works! But it is frustrating to have to always act on tolerance and change in small steps to verify that everything is working, do not you think?
In one of my previous model, to operate a boolean I had to try different settings for tolerance, from 0.001 to 0.05, then to 0.1 and finally to 0.2, with this value it worked!
This also applies to other commands, such as the fillet and the shelling, for example: very frustrating!
Yes it is, but the thing that I still don’t understand is why Rhino do not auto changes the tolerance to do the work like in some other cases?
Exact! It should do it automatically, can not we users, from case to case, find the exact tolerance … you lose too much unnecessary time.
The major problem here is that the intersection fails at the given tolerance. I will try to make that work more reliably. As I recall when the Boolean code fails it tries again with double and 1/10 of the original tolerance.
In this case if we tried with 10 times tolerance it works but not at the other tolerance values.
Reporting cases that fail is the best path to getting more reliable operations.
No Intersection Bug.3dm(173.2 KB) @GregArden Okay Greg, here is a case that fails. There are six planer surfaces (five of which are copies of the red one, as SelChildren will demonstrate) and a non-planer surface. Select them all and run Intersect, and the middle two planer surfaces fail to intersect the non-planer surface. If you increase the tolerance to .2 there is some kind of result, but so distorted as to be useless. Even at that level, I can’t use the non-planer surface to split or trim, which was the original intent.