For some reason, the boolean subtract just doesn't work on this:
Trying to Boolean difference #2 from #1

file: https://altuit.cloud/web/dbpub/3D/booleanProb2.3dm

Any help would be much appreciated!

Hi Chipp, you can often times get a clue of what is going wrong by generating intersection curves by selecting the 2 objects and doing Construct > Curve > isect.

In this case the intersection curves look like this:



Note how the intersection does not form just one cleanly defined closed loop, there are several different loops running into each other - anytime the intersection between 2 objects results in complex situations like this it can make it difficult for the booleans to clearly identify which areas of the model are supposed to be sliced up and kept or discarded.

The problem is that the 2 objects have some overlapping surface area in these spots:




When 2 objects just barely skim over some coincident surface area it can tend to make it difficult to boolean them very well, the intersection can become poorly defined and not easily separated into distinct regions.

Booleans will be more robust when the objects more distinctly push through each other rather than just barely skimming over some coincident shared surface areas. So it tends to be difficult to try to cut objects in this particular way that you are attempting here.

- Michael

Thanks Michael. I was originally trying to fillet that side, but it wouldn't take, so instead I *grabbed* the edge of the armrest, extended the front and back with tangent lines and then swept a perfect aligned 'cutter' -- or so I thought. Why wouldn't the cutter be aligned correctly, especially since it was generated from the originating curves and started from the Z direction?

I kinda hoped, as a last resort, the sweep along a path and boolean subtract would work instead of fillets. Guess I should rethink. Thanks for the tip on Construct > Curve > isect !! :-)

Hi Chipp,

> Why wouldn't the cutter be aligned correctly, especially since it was generated from
> the originating curves and started from the Z direction?

If you turn on the surface control points for both pieces, you should see that they are not 100% identical - sweeps are constructed by an adaptive fitting process that refines the generated surface repeatedly until it is within an acceptable tolerance. The result that you get from a sweep won't be the 100% exact same result as one from an extrusion for example. So in a case like you've got there, you end up with 2 very similar yet still not identical surfaces that overlap right on top of each other. Such things are generally difficult for booleans to deal with.

There are other methods that produce exact geometry from curves, like for example Extrude. But Sweep goes through a fairly different process.

- Michael