Best filament for 3D printed bolts and threads

You printed a big knob-bolt to clamp a shelf bracket. Cranked it down tight. Felt great.

A week later the shelf's drooping and the bolt spins with two fingers.

You didn't strip it. Nothing broke. The plastic just… slowly gave up. And it's almost always the same plastic.

PLA.

Here's the short answer for the best filament for 3D printed bolts and threads:

• PETG for most things. Tougher than PLA, handles more heat, cheap and easy.
• Nylon for threads you turn a lot, or that take a beating.
• PC for anything hot or under constant load.
• ASA for outdoors.
• PLA for prototypes, test fits, and light stuff that doesn't stay clamped.

Now the why, with actual numbers instead of vibes.

The thing that kills printed threads: creep

A bolt holding something tight is under load all the time. Plastic under constant load slowly stretches. That's creep, and it's why a printed bolt loosens even when nothing's touching it.

A 2022 study (Dogan, Strojniški vestnik) loaded six printed materials and watched them stretch at room temp, 40°C and 60°C. The results:

• PLA had the worst creep resistance. The paper's advice is to use it at room temperature with no load or very low loads.
• PC was the best. It was the only material that never broke in testing.
• PLA and Tough PLA both broke at 60°C.
• Nylon stretched the most, but never snapped.

CNC Kitchen ran a scrappier test that's closer to real life: bolts clamped through printed parts, re-tightened every day for a week. Un-annealed carbon-fiber PA6 nylon needed re-tightening "almost every day." Annealed, it was way better. PA12 nylon only needed it once — about the same as PLA and ASA in that test.

So creep isn't just a PLA thing. It's a load + heat + time thing. PLA just hits it first.

If it's clamped tight and it'll stay that way for months, don't use PLA.

Heat: when your threads go soft

Every plastic has a temperature where it starts sagging under load. Here's that number (heat deflection temperature) from real datasheets. These vary by brand, so I'm naming the brand:

| Filament | Heat deflection temp | Source | |---|---|---| | PLA | 55°C | Prusament PLA | | PETG | 68°C | Prusament PETG | | ASA | 86–93°C | Prusament ASA | | PC blend | 93–113°C | Prusament PC Blend | | CF nylon | 170–194°C | Bambu PAHT-CF |

CNC Kitchen also heat-tested PLA, PETG and ASA side by side: PLA softened around 60°C and failed at 65. PETG made it to 80/85. ASA held to 110/120.

I live in Phoenix. A PLA part in my car in July is a puddle with ambitions. If your threaded thing lives in a car, near a motor, next to a printer's hotend, or out in the sun, PLA's out.

Brittle vs tough

PLA is actually strong. It's stiff, and in CNC Kitchen's layer-adhesion test, PLA hooks held 40 kg versus 25 for PETG and 17 for ASA.

But it's brittle. Impact strength from the same comparison:

• PLA: 5 kJ/m²
• PETG: 8.6 kJ/m²
• ASA: 18 kJ/m²

For a thread, brittle is bad. Thread crests are tiny ridges, and a brittle tiny ridge chips off the first time you cross-thread it or yank it sideways.

Also — "PLA" isn't one material. Prusament PLA lists 51 MPa tensile strength. Bambu's PLA Basic lists 35 MPa. Same name, very different plastic. Read the datasheet on your spool.

Orientation beats filament

Okay, this one's important, and it's a little counterintuitive.

Printed parts are way weaker between layers than along them. Prusament PLA's datasheet: 51 MPa in-plane, 17 MPa between layers. Three times weaker.

A bolt printed standing up has every layer stacked along its length. Pull on it, and it snaps cleanly at a layer line. Shit, that's annoying.

So which way do you print it?

• Standing up gives the roundest, cleanest threads. That's why the basic printed-threads guide tells you to print bolts vertically — it's the right call for fit.
• Lying down makes a much stronger bolt, because the layers run the length of the shank. MyTechFun's tests found horizontally printed bolts were much stronger. The cost: the underside of the threads gets flattened by the bed.

My rule: light-duty, needs to spin nice? Stand it up. Taking real load? Lay it down — a lot of designers cut a small flat along one side of the bolt so it sits flat on the bed and nothing droops.

Torque is where threads actually die

One more number that changes how you design.

CNC Kitchen drove M3 screws straight into printed holes and pulled them out. In Prusament PLA, they held 142 kg before pulling out. But they stripped at about 1 Nm of torque. PETG: 118 kg pull-out, also about 1 Nm. Heat-set inserts in the same tests took 3–4 Nm.

Translation: printed threads almost never pull out. They strip when you crank them.

So if a thread gets tightened hard or often, stop picking filaments and use an insert or a nut.

Should you anneal PLA?

Annealing — baking a print so it recrystallizes — gets pitched as the way to make PLA tough. CNC Kitchen tested it (100°C for 45 minutes):

• Tensile strength went up 7.5%.
• Hook strength went up 16%.
• Layer adhesion didn't change at all.
• And the parts changed size by up to 10%. A 50 mm part came out 45 mm one way and 55 mm the other.

For threads, that's a hard no. Your damn nut won't fit your bolt anymore. CNC Kitchen's own verdict: the gain is overshadowed by the warping.

What each one needs from your printer

• PLA, PETG: any printer. PETG strings, so bump retraction and drop the temp a bit. And don't print PETG straight on smooth PEI — it can stick so well it damages the sheet.
• ABS: enclosure is essential. It shrinks and warps, and the fumes aren't great.
• ASA, PC, nylon: enclosure recommended.
• Anything carbon- or glass-filled: hardened nozzle. It'll chew through brass.
• Nylon: dry it first. Every time. It soaks up water like it's getting paid.
• TPU: works for soft caps, plugs and spacer washers. I wouldn't trust it as a fastener.
• PEI / PEEK: 400°C+ nozzle and a heated chamber. Not your printer. Not mine either.

What I'd actually print

• PLA: test fits, prototypes, display stuff, the knob you'll redesign three times anyway.
• PETG: shop jigs, clamps, brackets, bin hardware. My default.
• ASA: anything outside — garden, garage door, bike.
• PC: brackets near motors and heat, stuff under constant clamp load.
• Nylon: gears, thumbscrews you turn every day, anything that gets dropped.

Whatever you pick, the thread geometry is the same job. The Nasty Bastard bolt maker builds matched printed pairs with a clearance you dial in — and once you've found the temps and clearance that work for each filament, save them as a 3D printer setup in the Machine Setups Repository so you're not re-learning PETG every spring.

So what's the most PLA thing you've printed that really should've been PETG?

Frequently asked questions

What is the best filament for 3D printed bolts and threads?

PETG for most jobs: tougher than PLA, handles more heat, and easy to print. Use nylon for threads that turn often or take impacts, PC for heat or constant clamping load, ASA outdoors, and PLA for prototypes and light-duty parts that won't stay clamped.

Why do 3D printed PLA bolts loosen over time?

Creep. Plastic under constant load slowly stretches, and in a 2022 study PLA had the worst creep resistance of six printed materials, while PC had the best. Heat makes it much worse — PLA's heat deflection temperature is around 55°C.

Should I print bolts vertically or horizontally?

Vertically gives the cleanest, roundest threads, but the bolt can snap along a layer line because printed parts are much weaker between layers (Prusament PLA: 51 MPa in-plane vs 17 MPa between layers). Horizontally printed bolts are much stronger; add a small flat on one side so the threads don't droop.

Does annealing PLA make threads stronger?

Not usefully. In CNC Kitchen's test, annealing raised tensile strength 7.5% but changed part dimensions by up to 10%, shrinking one way and growing the other — enough that printed threads won't fit anymore.