Why printed holes come out undersize
Print the same part on three printers and measure the same hole: three different diameters. Holes print undersize - the extrusion crowds into the bore and the plastic pulls inward as it cools - and curved walls miss worse than flat ones. Each machine is also a little different (hotend and chamber temperature, part cooling, speed), so it’s hard to predict how far off any one of them will be. That spread is the problem for any tight fit: an interference sized on one printer jams on the next and rattles on a third. Every pattern below makes the geometry absorb the spread instead of fighting it.
Chamfer every hole entry
A lead-in chamfer starts the screw, insert or bearing square before the fit engages, and it absorbs the first-layer flare that otherwise makes the entry the tightest spot on the part. It costs nothing to print. If you take one habit from this page, take this one.
Use teardrops and polygons, not circles
Circles are the hardest feature a printer makes. A horizontal-axis hole sags at its unsupported crown - a teardrop or diamond profile gives it a self-supporting roof. A vertical bore carries the full curvature error - where the hole mates with hex hardware, a polygon pocket beats a round one, because flat walls print more predictably than curved ones and an across-flats fit repeats where a bore diameter won’t. For standard hex nuts the sizing is already done - every nut card in the FitCheck app dimensions the pocket in three fits.
Let the screw make its own fit
Where you'd reach for a tapped hole or a precise clearance, consider a pilot hole and let the screw form its own thread in the plastic. Thread engagement is deep and forgiving: a pilot that prints a few hundredths off still grips, while a precision clearance at the same error either binds or rattles.
Crush ribs for press fits
A rigid bore asks the whole circumference to be right. Crush ribs flip that: model the bore with working clearance, then add a few rounded ribs standing proud into the hole. Pressing the part in crushes the rib tips, so the interference lives in small features designed to yield - one geometry tolerates a machine-to-machine spread a plain press bore can’t. Taper the ribs toward the entry so the press starts easily. Expect a slight break-in over the first few insertions and then a stable grip - on the one plate we cycled, ribs after 50 insertions looked much the same as ribs after 4. For bearings, prefer more, smaller ribs - a few hard contact points can distort a thin bearing ring enough to affect how it spins. For common hardware the FitCheck app ships validated rib recipes on the bearing and nut cards - pick a grade and model the numbers it gives you.
Grip fins: the elastic alternative
Thin fins leaning in from the bore wall deflect rather than crush, so the force curve stays flat across a wider band of hole error - the reason to reach for them is spread, not reuse. On the one plate we tested they held harder than crush ribs rather than softer: every rung that seated read as a press fit, with nothing in the snug range, and the inward lean grips on the way out. Crush ribs are the better default for most fits, which is why they are what the app’s validated recipes use.
Relief cuts and split bosses
A slot through a boss wall turns the whole feature into a spring: the bore opens slightly around an oversize shaft instead of cracking. Pair it with a clamping screw when you need real grip.
When you need the exact number
The patterns above are for machines you can’t measure. When the machine is yours and you want the exact number - the drilled clearance, the printed hole, the insert pocket, the housing bore - start from the charts or any size page on this site, and correct it to your own machine’s measured error in the app.