Print a hinge and it works immediately. Print a shaft that spins inside a printed hole and it works for about a week. Then it squeaks, then it tightens, then it welds itself solid — and when you finally pull the parts apart, the bore looks polished in places and chewed up in others.
This is the least discussed corner of FDM design, and it is genuinely a corner: across hundreds of guides on this blog the words bearing, bushing, friction and lubrication have barely appeared at all. That is not because sliding contact is rare — it is everywhere, from a filament spool holder to a drawer runner to a rotating display stand. It is because sliding contact is where printed plastic behaves least like the metal it replaces.
The good news is that a printed sliding pair can run for years. It just has to be designed as a tribological system rather than as a hole with a shaft in it.
The Mechanism Everyone Blames on the Wrong Thing
When a printed pivot seizes, the assumption is usually that the parts are too tight, so the fix is more clearance. More clearance often makes it worse. The real sequence runs like this.
Two printed surfaces touch at a handful of high points, because FDM surfaces are a field of layer ridges and extrusion seams, not a smooth cylinder. Contact pressure at those points is therefore enormous — far higher than the average pressure you would compute by dividing load by area. Under that concentrated pressure, the polymer at the contact point heats up through friction, and as it warms it moves toward its glass transition temperature. Above roughly that point, it stops behaving like a solid and starts behaving like a very stiff paste.
Once that happens the high point smears, tears, or transfers material to the other surface. The torn material becomes a third body trapped in the gap, and a trapped hard particle does what an abrasive always does: it gouges. The surface gets rougher, friction rises, friction generates more heat, and the whole loop accelerates. That final stage is often called frictional welding, and it is why a joint that was smooth in March can be immovable in June.
The important consequence: you cannot fix a tribological problem with clearance alone. Clearance changes the geometry of the loop, not its existence.
Five Variables That Actually Decide Whether a Sliding Pair Survives
Sliding performance in printed plastic comes down to five things, and they matter in roughly this order. Change any one and the expected life can move by an order of magnitude.
| Variable | Why it dominates | Best available option |
|---|---|---|
| Pressure × velocity (PV) | Friction heat generated per unit area is proportional to pressure times sliding speed. This is the single strongest predictor of whether a pair survives. | Keep PV low: bigger contact area, slower motion, or shorter duty cycle. |
| Contact geometry | Ridges and seams concentrate load onto a few points, which is what starts the failure loop. | Low-stress fit with a designed interface, not line-on-line contact. |
| Stiffness match | Two stiff materials cannot conform to each other; the load stays on the original high points. | At least one side compliant enough to spread load (a soft polymer). |
| Material pair | Some polymer pairs have intrinsically low friction against each other; some are effectively adhesives under pressure. | Dissimilar polymers, or a smooth rigid counterface. |
| Debris management | A third body turns a mild wear situation into an abrasive one, and it never leaves on its own. | Grooves or relief pockets to collect particles out of the contact zone. |
Note that temperature is not a separate row. It is the multiplier on all five, which is why a design that survives indoors can fail outdoors, and why an enclosed printer's heated chamber changes what counts as a safe bearing material.
Material Selection for Sliding Contact
Stiffness and friction resistance pull in opposite directions here, so the choice is usually about which compromise fits the application. The table below is a design guide for printed sliding surfaces, not a scientific ranking.
| Material | Sliding behaviour | Use it when | Watch out for |
|---|---|---|---|
| PLA / PLA+ | Stiff and dimensionally consistent, low creep, good surface hardness. But low glass transition temperature, so friction heat softens it quickly. | Low-speed, intermittent, indoor motion where dimensional stability matters most. | Heat. A warm room or a fast-moving part will destroy it. Poor for continuous rotation. |
| PETG | Lower friction and better toughness than PLA, and noticeably more forgiving of slight misalignment. Softer, so it conforms and spreads load. | The default printed bearing surface. Good for moderate speed and continuous rotation. | It is the material most prone to cold flow under sustained load, so bores can ovalise over time. |
| ABS / ASA | Higher heat resistance than either, so it tolerates friction heating far better. Can be vapour smoothed to remove layer ridges from the contact face. | Warm environments, higher sliding speeds, or anywhere the joint runs continuously. | Warps and shrinks, so bores come out undersized and out of round unless compensated. |
| Nylon (PA) | The classic low-friction engineering polymer: tough, self-lubricating, excellent sliding behaviour, and used for real gears and bushings. | Loaded, continuous, demanding motion where a plastic bearing has to genuinely work. | Hygroscopic — it absorbs moisture and grows, which changes fits. Needs drying and dimensional allowance. |
| TPU | Very high friction against almost everything, but highly conformable, so it spreads load and tolerates misalignment. | As a sacrificial liner or gripper — deliberately the soft side of a pairing. | Never as a sliding surface against a rigid part if low friction is the goal. |
| Steel-nozzle-only composites (CF / GF filled) | Stiffer and more creep resistant, with improved wear resistance in some pairings, but abrasive to whatever they run against. | Loaded, slow, high-pressure pivots where creep resistance is the priority. | The filler also abrades the nozzle during printing. Requires a hardened or tungsten carbide nozzle. |
Two practical rules fall out of this table. First, pair dissimilar materials: a PETG bore running on an ABS or nylon shaft, or a printed bore on a smooth steel pin, is far better than PETG on PETG. Identical polymers in sliding contact have a strong tendency to bond at the high points. Second, if the joint is continuous and loaded, stop treating it as a printed-to-printed interface at all and put a real bearing or a metal sleeve in the load path.
Geometry: Designing the Contact Instead of Avoiding It
Since high spots cause the failure, the design goal is to control where contact happens and at what pressure. Four patterns cover most printed sliding joints.
- Plain bore and pin. Simplest and worst. Contact is a line along the bore, so pressure is concentrated and the whole load lands on a few ridges. Acceptable only for slow, light, occasional motion.
- Printed bushing pressed into a bore. Separate the wear part from the structure. The bushing is printed in a low-friction, load-conforming material and can be replaced without remaking the whole assembly. This is the single highest-value change for most projects.
- Flatted or crowned shaft. A full cylinder touches only along one line. Reducing the shaft to a flat or a crown against a matching pocket spreads the load over a defined area and lets you tune the contact pressure by changing that area.
- Real rolling element. Where genuine durability is required — a filament spool holder that turns for a thousand hours, a drive wheel, a load-bearing pivot — use a proper ball or roller bearing in a printed housing. Printing the housing and buying the bearing is almost always cheaper than reprinting the joint every few months.
Clearance is for motion, not for fit forgiveness
Printed bores come out undersized and slightly oval because the extrusion bead is squeezed into the corners of a circular toolpath. A nominal 8.00 mm bore typically measures somewhere between 7.7 and 7.9 mm depending on machine, material and flow calibration. If you model an 8.00 mm bore for an 8.00 mm shaft, the parts will not turn no matter how much you ream them, because the shape error is not uniform.
Two habits fix this reliably. First, calibrate with a test coupon rather than trusting a single global offset: print a strip of bores at 8.0, 8.1, 8.2 and 8.3 mm, measure each, and pick the one that gives the running fit you want. Second, bore compensation is direction dependent: a vertical bore printed in the XY plane and a horizontal bore printed in the Z direction will come out different sizes. Coupon first, then adjust the model. A design that specifies tolerances rather than exact dimensions is a design that can be printed on more than one machine.
Debris relief
Since a trapped particle turns mild wear into abrasion, give wear debris somewhere to go. A shallow circumferential groove at the ends of the bore, or a few axial relief slots, will collect particles out of the contact zone instead of letting them circulate. This costs nothing in print time and is the cheapest life extension available to a sliding joint.
Lubrication: What Works and What Backfires
Printed PLA and PETG are generally not rated for petroleum-based greases, and some formulations will swell, soften or craze when exposed to them. But there is a wider set of options that behave well.
- Dry lubricants first. PTFE powder, graphite powder, or a dry-film PTFE spray applied to the contact face add a low-shear boundary layer without collecting dust. This is the safest starting point and works remarkably well on PLA and PETG pivots.
- Silicone grease. Compatible with most printed polymers, stays put, and does not attack the material. Good for slow, high-pressure joints. Keep it away from anything that must later be glued or painted.
- Lithium or PTFE grease. Fine on ABS, ASA and nylon. Verify on PLA and PETG with a scrap test piece before committing, because compatibility varies by formulation.
- Build-plate adhesion aids. Do not use them here. Their entire job is to make polymer stick to a surface, which is precisely the failure you are trying to prevent.
- Nothing at all. For a light-duty dry pivot, an unlubricated dissimilar-material pair with debris relief often outlasts a badly chosen grease. Grease that traps dust becomes an abrasive slurry.
Whatever you choose, apply it to a cleaned surface. Oil and handling residue on a freshly printed part attracts dust, and dust in a sliding joint is an abrasive. Wipe, dry, then lubricate.
When to Stop Designing and Buy the Bearing
Printed sliding surfaces have a real envelope. Inside it they are excellent; outside it they fail predictably. Move to a purchased bearing or a metal sleeve when any of the following is true.
| Condition | Why printing stops being the answer |
|---|---|
| Continuous rotation for thousands of hours | Wear rate per hour is small but never zero, and total wear is what matters over a device's life. |
| Load above roughly a few kilograms on a small bore | Contact pressure rises past the point where friction heat softens the polymer. |
| Any precision requirement on runout or wobble | Printed bores are oval and machine dependent; a bearing delivers a round, hard, repeatable raceway. |
| Ambient temperature near the material's glass transition | The safety margin against friction heating disappears before the joint even starts moving. |
| Dust, grit, or abrasive environment | Sealed bearings handle contamination; an open printed clearance does not. |
The pragmatic pattern that experienced makers converge on is print the housing, buy the bearing. A printed bracket with a press-fit pocket for a standard bearing costs a few cents of filament and gives you metal-on-metal rolling contact, while keeping the freedom of printed geometry for everything around it. When you do press a bearing into a printed pocket, size the pocket for a light interference fit in a material with low creep — a PETG or ABS pocket will hold it far better than a PLA one in a warm environment — and print the pocket walls thick enough that the hoop stress does not split them.
Commissioning Checklist
Before you declare a sliding joint finished, run these eight checks. Each one maps to a failure mode above.
- Material pair is dissimilar — not the same polymer on both sides of the interface.
- Fit coupon printed and measured — you know your actual bore size, not the nominal one.
- Axial and radial clearance checked hot, not cold — if the joint runs warm, clearances shrink.
- Load path does not depend on printed sliding surfaces if the joint is continuous and loaded.
- Debris relief present — grooves, slots or end relief so particles can escape.
- Lubricant chosen for the material and applied to a clean surface.
- Nozzle is appropriate if any filled or abrasive composite is in the build.
- Filament is dry — wet filament prints a rough, porous surface, and a rough surface is a high-friction surface.
Getting the Material Side Right
Almost every decision above lands on the filament and the hardware around it. The bore material sets the friction and creep behaviour, the nozzle has to survive whatever filler you choose, the bed surface controls the dimensional accuracy of the first layers that define your fit, and dry filament is what keeps the contact surfaces smooth enough not to chew themselves up. None of those are optional details on a sliding joint; they are the joint.
For a general-purpose printed bearing surface, PETG filament is the practical default: lower friction and more forgiving of misalignment than PLA, and it conforms to spread load. Where the joint runs warm or continuously, move up to ABS filament or ASA filament for the higher heat resistance and the option of vapour smoothing the contact face. If the pivot has to carry real load, nylon filament is in a different class for sliding wear. And when creep resistance matters more than friction, PETG-CF filament and ABS-CF filament hold their dimensions under sustained pressure far better than their unfilled versions.
On the hardware side, filled and abrasive filaments will destroy a brass nozzle, so pair them with a tungsten carbide nozzle. Dimensional accuracy starts at the first layer, which is what a PEI build plate is for. And because a rough surface is a high-friction surface, keeping filament dry with a filament dryer cabinet is part of the tribology, not a separate maintenance chore.
If you want to go deeper on the two design problems that sit closest to this one, the fit and tolerance side is covered in the guide to tuning dimensional accuracy and tolerances, and the assembly side — joints that move without fasteners — in the piece on print-in-place design for moving parts.
Sliding contact is the most honest test a printed part faces. There is nowhere to hide: friction finds the high points, heat softens them, and the debris does the rest. Design for the contact, choose dissimilar materials, give the debris a way out, and pick your material for the temperature the joint will actually reach — and a printed bearing becomes one of the quietest, longest-lived things on the machine.