Why do my printed threads only catch halfway before they jam?

Nov 20, 2026 Last reply: 20 hours ago 4 Replies

First reply

PilotForge

Yeah, hand-modeled threads binding after a turn or two is almost always the thread profile itself, not just the clearance number. The usual culprit is that your thread crest and root are both sharp, or your flank angle is too steep, so the threads are actually interfering…

Read the full reply

Just finished my first model with actual moving parts and I'm stuck. It's a little hanging lamp cover, teardrop shaped, that screws onto a base with a loop up top so you can hang it from a hook. Looks decent on the shelf but when I try to thread the two halves togather, they engage fine for about one or two turns and then it just binds up no matter how gently I go.



I modeled the threads myself and I'm guessing that's the problem. I used the same profile all the way around but I didn't relly think about tolerance until it was printed. The deeper threads expecially seem to grab and grind rather than glide. Printed the whole thing twice now with slightly diffrent clearances and got the same result both times.



Is there a rule of thumb for how much clearance to leave on threads, or is this a sanding-and-hope situation? Also curious whether it's better to model them slightly oversize or just run a bolt into the hole while the plastic is still warm. It's petg if that matters. The lamp part of it works great, I just want the lid to actually screw on without me forcing it.


Yeah, hand-modeled threads binding after a turn or two is almost always the thread profile itself, not just the clearance number. The usual culprit is that your thread crest and root are both sharp, or your flank angle is too steep, so the threads are actually interfering instead of gliding. Easiest fix without redesigning everything: cut a chamfer on the inside corner of the nut side and round the thread tips slightly, so only the flanks touch, not the peaks. That alone often turns a grinder into a smooth screw-on. For clearance numbers, on a print this size I'd give the threads about 0.3 to 0.4 mm of total slack, meaning shrink the male thread by that much on the diameter. PETG is forgiving because it's a bit flexible, but it also oozes and droops, so printed threads come out fatter than modeled, especially near the top of a curved wall where overhangs sag. Your 'deeper threads grab' comment fits that perfectly, the last turns are prolly the most droopy ones. If you can, orient so the thread helix prints as a gentle overhang rather than a steep one, and slow the outer perimeter down.

As for oversize modeling vs running a bolt in while warm, I'd skip the warm-bolt trick for PETG. It works okay on PLA, but PETG is stretchy and tends to just deform the threads into a sloppy fit instead of cutting them clean. Model it right and test-fit as you go. Also try screwing the halves together while both are still on the printer bed and warm, before they cool, as a diagnostic: if it threads fine hot and binds cold, it's a shrinkage/clearance issue; if it binds hot too, it's geometry. One more thing: first two or three turns matter most. Give those a slightly looser fit or add a lead-in taper at the start of the thread so it picks up cleanly instead of cross-threading. Sanding can rescue a mild bind but if it grinds after two turns, sanding won't get you there, the profile needs the fix.

I ran into the same grabby spiral thing on a pill bottle style lid I made. What partly helped me was printing the cap sideways instead of flat, so the bumpy layer lines don't sit right where the parts rub. Still had to file it a bit by hand after. Also dumb question, is there a way to have the drawing program draw the spiral for you instead of building it by hand? Feels like somthing the computer should be able to do.

Printing a tiny test ring of just the threads saved me a bunch of filament on my own threaded lid project. Instead of burning a whole print to find out the fit is off, I model one short section, maybe a quarter turn of each half, and print those flat on the bed. Takes twenty minutes and I can try the two pieces together right there. I also measure the actual printed threads with my calipers first, because they always come out fatter than what's in the model, and that tells me how much to shave off before I commit. Once the test pieces screw together smoothly, the full part usally just works.

Update: did the little test ring thing from the third reply and wow, that saved me a whole print. Turns out the crest on my deeper turns was basially square, not rounded like the rest. Filed the ring down and it threads smooth now. One question though, the first reply mentioned flank angles being too steep. How do you even check that, just eyeball it in the model or is there a trick?

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required