Misbehaving bool

Argh! I seem to keep running into issues making the boolean operator work.

Normally it’s something obvious like bad normals. But every so often it just seems to break for no obvious reason.

Here’s another example (See attached file - B-2.72). I’m trying to union together a model and a sprue (for 3D print).

One part (out of 40 - the only other part in this sample file) is refusing to ‘boolean’.

To see what I mean, open file, Select sprue on layer 1, Add a boolean, UNION with ‘FootBallFL’ (the only other
object - it’s on layer2).

No union.

You’ll see something if you select INTERSECT !? But you’ll also lose the top and bottom
sections of the sprue.

Does anyone have any idea what’s going on?

Ross.

Attachments

BoolGllitch.blend (741 KB)

To prevent issues with ngons, you can add a Triangulate modifier on the top of each modifier stack. After you apply the boolean modifier, remove the remaining triangulate modifier



Yep That’ll do it.

Just found another way too. Just triangulate the problem part (which has no ngons!?)

Any idea why? The nGons on the sprue were just a quick way to fill the open
ends. I was planning on triangulating as I exported the STL. After all the boolean
‘generates’ ngons, it shouldn’t be averse to them.

Thanks Richard. I can at least get moving again on this.

Well. For the record, removing the nGons (manually) didn’t work. So I fell back on triangulating the problem part(s) - turns
out I’d lost two.

Both parts, and the sprue were 100% quad (checked with meshlint). Triangulating the parts (but not the sprue) got me the
correct boolean result.

Sure glad I spotted this problem before getting a print done.

I don’t know if this might help others wrestling with the boolean modifier. But I persisted,
trying various things to aviod triangulation.

Something else that did work was to modify the topology of the problem part.

I’d left some rather tacky quad fans at the top of the part. So I cleaned them up like so…


…and the boolean started working.

But it’s not the quad fan (at least not directly). I think it was just the change in
topology or face count. Other parts had (and still have) the same structure and
caused no problems. Go figure.

R.