There is a specific kind of failure that shows up on 3D printed sliding parts about six to eighteen months in, and it almost never gets diagnosed correctly. The part still works. It just stops being right. A drawer slide that used to run with a light push now needs a shove. A hinged clamp that held a hose snug at month one leaks slightly at month twelve despite never being adjusted. A guide block that tracked a rod with no play now rattles.
The owner usually says the plastic "wore out" or "stretched", and then reaches for a stiffer filament. That is where the diagnosis goes wrong. What happened is not one wear process but three, and only one of them responds to a stiffer material. Two of them get worse when you reach for a stiffer material, because the fix for those is geometry and clearance, not hardness.
This article is about abrasive wear specifically — the mechanism that removes material rather than deforming it. It is separate from the adhesion-and-heat problems that cause a printed sliding surface to weld itself shut, and separate from the long-term creep that causes a loaded part to lose preload months later. Those two are covered elsewhere. Abrasion is the third failure path, and on residential, workshop, garden, marine and automotive parts it is frequently the one that actually ends the part's useful life.
Friction is not the enemy. Friction is the mechanism that makes sliding work
Friction gets a bad reputation in 3D printing discourse, and it is undeserved. A sliding joint needs friction to have any feel at all, and a detent, a friction hinge or a press fit is friction doing exactly the job it was hired for. Friction coefficient tells you how much push the joint needs. Whether the joint survives is a completely different question.
Material loss is that different question. Two surfaces in contact under load, sliding past each other, will lose material by one of four routes, and the route determines what you should change:
| Wear mode | What is actually happening | Signature you can see | The fix that works |
|---|---|---|---|
| Adhesive | Microscopic high spots cold-weld, then a lump tears out of the softer side | Scuffing, transfer film, matte patches on a glossy surface | Lower surface pressure, dissimilar materials, lubrication |
| Abrasive | Hard particles cut grooves in the softer surface, like sandpaper | Fine parallel scratches aligned with the direction of travel; polished-looking tracks | Keep particles out, let them escape, or make the surface harder than the particles |
| Fatigue | Repeated load cycles crack the surface, flaking off small plates | Pitting, shallow spalls, a surface that looks chewed rather than scratched | Lower pressure, thicker walls, more contact area |
| Erosive | Particles or fluid impact a surface at an angle and blast it away | Worn leading edge, rounded corners on the exposed face | Deflect the stream, harden the impact face, add a wear plate |
In 3D printing, abrasive wear dominates on anything used outdoors, near soil, near metalworking or on a machine that sees grit. And it has a feature no other wear mode has: it can be generated by the part itself. That is the part everyone misses.
Two-body wear, three-body wear, and why the distinction decides your design
Two-body wear is a hard surface cutting a soft one: a steel shaft running in a printed bushing. The wear volume is predictable, it is proportional to load and distance, and it is proportional to how much harder one surface is than the other. If your printed part runs against bare steel and there is no grit in the system, two-body wear is the whole story, and the usual fixes — more contact area, lower pressure, a harder filled filament — are the right ones.
Three-body wear is different in kind, not degree. The two surfaces are no longer touching each other directly. A third party is in the gap: sand, sawdust, grit, metal swarf, dried mud, salt crystals, or debris the printed part shed itself. That third party is usually harder than both surfaces, it cannot be squeezed out of the gap, and it rolls or slides between them, cutting both.
The consequence is counterintuitive and it is the single most useful idea in this article: in three-body wear, a bigger clearance wears slower than a tight one. A tight sliding fit is the worst possible geometry for a gritty environment, because a tight fit traps the particle at the point of highest pressure and forces it to cut. A slightly loose fit gives the particle somewhere to go.
That is precisely why the drawer slide got sloppy: the initial fit was correct for a clean bench, and it was doing three-body wear on day two. And it is why the intuitive remedy — tighten the clearance to remove the rattle — makes the wear rate climb.
The wear you generate yourself: printed debris is abrasive too
Here is the part that catches experienced designers. A printed sliding surface starts shedding particles almost immediately, and those particles are not soft. The shed material includes partially melted strings, oozed blobs, poorly bonded crust from the top surface of the print, and — if you used a filled filament — exposed fiber and mineral particles that are harder than the polymer matrix around them.
Consider what is actually embedded in a carbon fiber nylon or a glass fiber PET. The fiber is much harder than the plastic. When the surface abrades, the plastic matrix wears back first and leaves the fiber ends protruding, and those ends are now a microscopic rasp working against whatever is sliding on them. On a filled filament, a smooth-looking surface is not a smooth surface at 50 µm scale.
This is also why stone- and rock-effect filaments should never be used for a sliding interface. The filler that gives the print its stone look is mineral particulate, and mineral particulate is what sandpaper is made of. It is fine for a decorative panel. It is the wrong part of the shop for a wear surface, and using it there means you built the abrasive into the joint.
Two practical corollaries follow. First, break in a printed sliding joint before you judge it: run it dry through a few hundred cycles, disassemble, wipe out the dust, and only then evaluate clearance. Second, on any part where you care about the sliding interface, use a filament without hard filler, and use a tungsten or hardened nozzle only where you actually need abrasion resistance in the nozzle, not as a substitute for the right material choice.
Hard particles do not just cut. They embed.
The second overlooked consequence of a third party in the gap is embedding. A hard particle pressed into a surface under load can stop cutting and become part of that surface. Now the printed part has effectively become sandpaper, and it is grinding the shaft it was supposed to be sliding on.
Embedding depends on how soft the surface is and how much pressure it sees. A soft, low-modulus printed surface under high pressure in a gritty environment is an embedding machine. The damage then appears on the other component — a scored steel shaft, a chewed aluminium rail — and the printer owner blames the metal.
The design lesson is that the softer member in a gritty sliding pair must be either replaceable or protected, because once it embeds grit, it attacks the expensive part. Making it replaceable is usually cheaper than making it harder: a printed sacrificial sleeve you swap every season costs a few grams of filament, and it saves the shaft.
Pressure × velocity, and the number that matters when grit is present
Wear rate rises faster than linearly with both contact pressure and sliding speed, and when grit is present the exponent gets worse. The practical consequence is that the same joint which is comfortable at a slow hand-driven speed becomes a wear problem when someone motors it, and the same motorised joint is fine at a low pressure until someone adds a load.
The useful way to think about it is a product, not a single number: pressure times velocity, with pressure measured over the real contact area rather than the nominal one. Two printed surfaces nominally in contact over 200 mm² are usually touching over a fraction of that, because both are slightly curved and both carry the layer texture. All the load goes through a few ridges. That is where the wear starts, and it is why following where the load actually travels tells you more than looking at the bounding box.
| Duty | Real pressure at the interface | Sliding speed | What to expect |
|---|---|---|---|
| Occasional manual, indoor | Under 1 MPa | Under 0.05 m/s | Long life; two-body wear only; nearly any filament is fine |
| Repeated manual, indoor, dry | 1–3 MPa | 0.05–0.2 m/s | Slow polish-out; check clearance at six months |
| Repeated manual, outdoor or gritty | 1–3 MPa | 0.05–0.2 m/s | Three-body wear dominates; clearance and drainage become critical |
| Powered, indoor | 3–8 MPa | 0.2–1 m/s | Heat joins the problem; lubricant or a plain bearing is now mandatory |
| Powered, gritty | Any | Any | Do not run printed-on-printed; use a sealed bearing or a replaceable sleeve |
Note what the last row says. There is a category of duty where the correct engineering answer is to stop trying to make plastic wear well and instead to isolate the sliding interface from the grit entirely. That is not a defeat. It is the same decision as choosing a real bushing over a printed pivot once the duty cycle passes what the material can take.
Geometry that lets grit out instead of grinding it in
If you cannot keep particles out — and in a workshop, a garden or a boat you usually cannot — then the design job is to make sure the particles leave quickly instead of being trapped and re-ground. Six rules do most of the work.
1. Slope the contact face so it self-cleans
A horizontal sliding surface collects grit. A surface inclined even a few degrees lets it migrate to the low edge and drop off. Where the geometry allows it, put the sliding face at an angle and give the low edge a lip that does not trap material.
2. Cut relief grooves and debris channels
A single shallow groove in the middle of a wide contact face gives particles somewhere to fall out of the load zone. Two grooves flanking the contact band work even better. The grooves should be wider than the particle size you expect — 1.0–1.5 mm is a sensible floor for outdoor use — and they should break out of the part at the ends rather than dead-ending.
3. Do not bottom out a sliding face
A face that is machined flat and seats fully along its whole length cannot shed anything. Leave a small relief in the middle so contact happens near the edges or near the load line, and the middle stays open as a debris reservoir.
4. Lead-in chamfers at both ends of travel
A chamfer at the entry edge of a sliding track pushes a particle aside instead of trapping it under the leading edge. On a printed part this costs one line of code and it removes the single most common starting point for a gouge.
5. Face the sliding interface away from the grit source
If a part lives on the underside of a machine or near a wheel, the sliding face will collect everything that falls. Turn the interface sideways, or better, downward. An interface facing down sheds material by gravity and is the cheapest possible seal.
6. Make the wear surface replaceable, and expect to replace it
Print the sliding surface as a separate insert held by two screws, or as a snap-in pad. It is not a failure of the design that this part gets consumed. It is an acknowledgement that in an abrasive environment, a printed wear part is a consumable, and a consumable that takes four minutes to swap is a better design than an integrated one that lasts 20% longer and then means rebuilding the whole assembly.
Let the steel take the wear, not the plastic
There is a simple design inversion that solves most abrasive wear problems outright: put a hard, cheap, replaceable steel item in the joint and let the plastic be the soft, generous, low-pressure member. A plain steel washer behind a printed flange, a stainless rod in a printed guide, a strip of stainless shim on the wear face. The steel is nearly immune to the grit, and the plastic's job becomes locating and cushioning rather than resisting abrasion.
That also settles the filler question. Filled filament exists to raise stiffness and reduce creep, not to make a plastic part resist sand. If your real problem is sand, a washer beats a carbon-fiber spool every time, and it costs less. Use filled material where it earns its place — stiffness, dimensional stability, heat resistance — and solve abrasion with metal at the contact face.
Material reality check: what actually helps, measured in the right direction
Filament data sheets will not answer this question, so here is what the mechanisms imply. The important column is not hardness — it is whether the material sheds its own abrasive debris and whether it can be embedded with grit.
| Material | Component wear in gritty sliding | Debris & embedding behaviour | Where it belongs |
|---|---|---|---|
| PA12-CF | Good resistance, low moisture sensitivity in the base resin | Exposed fiber ends rasp the counterface | Load-bearing slides that stay dry; not for bare steel shafts |
| PA6-CF | Good, but the base resin absorbs moisture and properties shift | Same fiber-rasp issue; more swelling-driven clearance change | Structural wear parts when stiffness matters more than precision |
| PA6-GF | Better abrasion resistance than unfilled nylon | Glass is hard and tends to break into sharp fragments | Sliding where the counterface is a hard steel or a replaceable sleeve |
| PPS-CF | Best of the group for hot, chemically aggressive and abrasive environments | Low debris; expensive and demanding to print | Service parts on industrial equipment; not a first choice for a hobby joint |
| PET-CF | High stiffness and low moisture pickup, decent against grit | Fiber ends present, but the matrix is hard and sheds less crust | Precision slides where dimensional stability is the driving concern |
| ASA | Moderate; abrasion resistance is ordinary | Soft enough to embed grit | Outdoor housings and enclosures rather than wear faces |
| PETG | Moderate; smooth and somewhat self-lubricating in dry slide | Softer, so grit embeds easily and then attacks the partner | Light indoor slides, low pressure, no grit |
| ABS / ASA-CF | Ordinary to good, with better heat tolerance than PETG | The filled version has fiber ends; the plain version embeds | Structural brackets where sliding is incidental |
| PLA | Low; brittle debris and poor heat tolerance | Produces hard, sharp chips as it wears | Prototypes and indoor light-duty guides only |
| TPU 85A / TPU 95A | High in bulk wear, but deforms around particles rather than being cut | Embeds grit readily and holds it | Wipers, seals and grit scrapers ahead of the real sliding surface |
That last row is the most useful idea in the table and the least obvious. A soft elastomer is a terrible wear surface and an excellent grit scraper. Put a TPU lip ahead of a sliding interface and it wipes the shaft clean before the bearing sees it. You are not trying to make the seal wear well; you are using its willingness to deform to keep the abrasive away from the part that must hold tolerance. A 95A TPU is the usual starting point for a scraper lip: stiff enough to stand up, soft enough to follow a slightly imperfect surface.
Softness also matters for a second job — the one covered in the article on parts your hand touches. A compliant pad distributes load over a large area, which drops the real pressure at the interface, and lower real pressure is the cheapest way to slow every one of the four wear modes at once. On a hand-operated mechanism, that is often all the "fix" a worn joint needs.
Printing choices that change the wear rate before the part is ever used
Orientation is the first decision and it is usually made for the wrong reason. On a sliding face, the layer lines should run parallel to the direction of travel, not across it. Layers running across the direction of travel present their edges to the grit, and a layer edge chips away much faster than a continuous bead. That is also why a printed rail wears at both ends of travel first: the direction reverses and the exposed bead ends take the hit.
Surface finish matters more than it looks. A rough top surface is a field of ridges that trap particles. If the sliding face can be printed against the bed or against a smooth plate, do it — a bed-side surface on a cold PEI plate comes off far smoother than a top surface and gives grit far less to hold onto. If it has to be a top surface, ironing helps more than most settings, and it helps specifically because it flattens the ridges rather than polishing them.
Wall count is the second lever. A wear surface is a surface, and in a printed part the surface is the shell. Three or four perimeters give the wear zone real material to lose; a two-perimeter wall exposes infill within the first few tenths of a millimeter, and once infill is exposed the surface collapses into a cavity rather than wearing smoothly. This is one more case where adding shells does more than adding infill.
Dimensional control is the third lever and it is the one that makes the difference between a joint that wears predictably and one that seizes. Clearance on a printed slide comes from the slicer, the material's shrinkage and the printer's calibration, and every change to those moves the fit. Check the actual fit as printed, using the method in measuring printed parts instead of guessing, before you decide the joint is "too tight" and start filing.
Finally, print the sliding part dry. Moisture in a filament turns to steam in the melt zone, which produces micro-voids, and micro-voids are pits in the wear face from day one. A dry box matters for wear parts at least as much as for appearance parts, and nylon especially will not behave consistently without it.
Dust is not the only third body: the cases people forget
The three-body wear conversation usually stops at sand and sawdust. There are several others worth planning for, and each one has a specific countermeasure:
- Metal swarf and grinding dust. Hard, sharp, and often magnetic. Facing a slide downward plus a magnetic pick-up nearby handles more of it than any surface treatment.
- Salt crystals. Salt is hard, and on a boat it is everywhere. It also pulls moisture in, so the joint runs wet and abrasive at the same time. Rinse the mechanism, and design the interface to drain.
- Dried mud and clay. Behaves like a lapping paste when damp and like a grinding powder when dry. It is the main reason garden and farm equipment joints die, and the fix is a large clearance with a drain path rather than a tight seal.
- Printer dust itself. A printed mechanism living inside a printer enclosure collects fine plastic and carbon dust. This is a low-rate wear case, but it is real, and it is why the same part that survives ten years on a bench can fail in three inside a machine.
- Sacrificial debris. If your mechanism has a sacrificial element that is designed to wear, its own debris becomes the third body for the next joint downstream. Either locate the sacrificial element last, or give it a catch pocket.
Testing wear in a week instead of waiting a year
Wear is slow and therefore usually untested. A short, crude test is far better than no test, and it can be done in an afternoon with a printed rig:
- Print three identical test coupons of the joint at three clearances — nominal, nominal +0.15 mm and nominal +0.30 mm.
- Add a measured dose of the actual contaminant — the real sand from the real location, or the real sawdust from the real saw. Estimate it by volume, not by eye.
- Cycle each coupon by hand, counting strokes, to at least two hundred cycles, and stop every fifty to wipe the joint and re-add the same dose so you are not testing a single charge of grit forever.
- Measure the play at the start of each run and at the end, using a dial indicator or a simple lever-and-feeler setup. The number you want is change in play per hundred cycles, not absolute play.
- Photograph the wear track at the same magnification each time. Scratches running the direction of travel confirm abrasive wear; pitting means the pressure is too high for the material; a polished band with transfer smears means the material pair is adhesive-welding, which is a different fix entirely.
The result is almost always counterintuitive in the same way: the loosest coupon wears slowest, and the tightest one develops visible gouging first. If that is not what your test shows, you probably do not have a three-body problem, and you should be looking at adhesion or pressure instead.
When to stop optimising plastic and change the architecture
There is a point where the correct answer is a different design rather than a better material, and it is worth naming the conditions so you do not spend three iterations discovering them:
- The joint is powered rather than hand-driven, and the environment is gritty. Sealed rolling element, or a replaceable sleeve on a steel pin.
- The part must hold a tolerance over years while sliding in contamination. Printed wear faces do not do this; printed locating features with a metal wear face do.
- The counterface is expensive — a machined shaft, a hydraulic rod, a precision rail. Protect it with a sacrificial printed sleeve or a wiper, never by letting printed material rub it with grit in between.
- The consequence of wear is unsafe. A latch that loosens with wear, a load-carrying hook that thins, a brake or stop that relies on a fit. Do not put a printed sliding wear surface there.
- You are already on your third replacement and the interval is getting shorter. Something is embedding grit and grinding, and more of the same will not fix it.
The checklist
- Name the wear mode before choosing a material: adhesive, abrasive, fatigue or erosive. Only abrasive wear calls for hardness.
- Ask whether there is a third body in the gap. If there is, clearance and escape routes outrank every material property.
- Audit your own debris: no stone-effect, wood-effect or mineral-filled filament at a sliding interface, and no exposed fiber ends against a bare shaft.
- Assume the printed part will embed grit and then attack its partner. Make the soft member replaceable.
- Put the hard, cheap, replaceable steel item in the joint and let plastic carry load rather than abrasion.
- Slope the wear face, cut debris channels, chamfer the lead-in, and never let a sliding face seat along its full length.
- Orient layers parallel to travel, use at least three perimeters, and print the wear face on the plate side when possible.
- Print wear parts dry. Micro-voids are pits that start the whole process.
- Add a TPU scraper lip ahead of the tolerance-critical interface. Softness deforms around grit; that is the point.
- Run the three-clearance test with the real contaminant before you commit to a design. Expect the loosest one to win.
None of this is exotic. It is the ordinary tribology of grit, applied to a material that — unlike metal — is soft enough to embed the grit and shed its own abrasive debris. Once that is the frame, the mount that holds the slide and the slide itself stop being one problem, and the design starts solving the problem that is actually killing the part instead of the one that is easiest to name.