Weird boolean union behavior, any help appreciated

So I’m making a couple of vases and I’m Flowing the Chicago bears logo onto them. I’m just playing but have had some odd behavior happen when I boolean union them. In the following images, you will see the vase and the big C flowing along the surface. I have the C sticking in approx .01" inside the od of the vase. I have been then offsetting the vase surface inwards .06" with the solid command and trying to union them. The behavior I’m noticing is the surface area right around the Big C becomes see-through. I’ve tried this with multiple methods and even tried trimming. Sometimes the area around the big C does different things, but I’ve noticed in one of the images it appears the normals somehow get flipped for the vase surface and the area just around the big C after the boolean union command. What causes this? I’m missing something somewhere.

Vases.3dm (7.9 MB)

Hello - the problem is the flowed object has a doubled back surface here:

I’ve seen this before - it may be from the refitting that takes place in flowing, I’ll see if I can figure out how to make it better.

-Pascal

Thanks Pascal. I appreciate the help. I’ve seen this before and wasn’t sure if it’s the technique I’m using or not. Any advice is much appreciated!

Hi Jacob - `ExtractSrf` this surface:

`Rebuild` to degree 3 and four points each way, then `MatchSrf` for tangency back to that edge where the original surface gets messy and dense. Join it all up and then Flow again. Any better?
(the folded part of the flowed surface was causing a loop in the trim curve there when Booleaned in, and messing up the display mesh)

Incidentally, @Jacob1 - a way to find these is to intersect the objects then turn on curvature graph for the resulting curve - it is messy but if you lower the scale - sometimes a lot - you can see where there is still a huge spike in curvature - look closely at these locations, they may be looped or folded:

-Pascal

That did the trick! Thanks so much. Also thanks for the tips in finding these issues in the future. I knew those curves were not the best. I should have known better but it never dawned on me to do an intersection and try a curvature graph. Next time I’ll know. Thanks

1 Like