Hi MA, I haven't found the specific area that is problematic yet but in the process of removing the one bad booleaned piece that was left on it, it now seems to be doing ok. Something like it doesn't happen to fire a ray for inside/outside testing through a problematic area anymore. I've attached the result of the boolean so you can keep going with working on it.

- Michael

Hi MA, I didn't find any self-intersections in the illustration area, but there are several areas like this with very very skinny remnants from previous booleans just kind of barely not matching up to other pieces:





Those little surface fragments are skinny enough (like only 0.0003 units across) that they probably appear to the volume ray fire mechanism the same as a self-intersection. So if a rayfire happens to try and cross over that kind of skinny fragment it can get confused.

That's my best guess right now.

One other thing you can try if you run into this situation with boolean difference behaving like union instead is to select your main object and use Edit > Separate and then follow that by an Edit > Join. That's going to make it recalculate the outside/inside directions. If I do that on your original model it seems to be ok with the boolean difference after that.

- Michael

Excellent help! Thank you.