/diy/ - Do It Yourself

>>2758026
i found that this was usually due to the inconsistency in line widths. Either people assumed you knew what size nozzle they were using, or had really shit accuracy on their widths.
Looking into this i found that the default 0.45mm width for a 0.4 nozzle was a pain in the ass to model for accurately. Partly due to the calculations required. So I switched to specifying 0.4 width for all 0.4 nozzle lines, in all my profiles. It's the one thing i religiously specify all the time.
Now if I have a 10mm diameter cylindrical rod IRL that is to fit into a hole in a printed object, I make the hole 10.4mm in diameter.
Basically adding 0.2 on each side of any object.
10x10mm square object, 10.4x10.4mm boolean cuts a 10.4x10.4mm hole.
This includes a tiny bit of slop to account for the roughness of layering, but still works when the hole exists in any combination of X,Y,Z planes. Sometimes horizontal planes can collapse a bit on the last bridging layer they put on top of the hole, but shoving a same-sized drill or file into the hole makes it trivial to clear them.
This applies to making printed parts fit into printed parts too, except in addition to making the hole bigger you MAY need to also make the inserted part 0.4mm smaller, but if your line width is accurate in the first place, and your layers are all smoothly parallel with no globbing it should work fine by simply adding 0.4 to the hole alone.
Do some test fit prints to see. It's a very close match for me, and if i need a tighter fit its easy to reduce the hole to 0.3 or 0.2... but thats usually pretty damn tight, and i cant recall when i last need to do that.

Picrel the same principle applied to an M3 bolt, the nut is square because hex nuts suck to embed, and it's slot is extra long because in that axis it's advantageous to have a 'big' gap.