Friction coefficients in 3D printing are often presented as if each filament has one fixed number. Real sliding parts do not behave that way. A printed surface carries layer lines, local heat history, void distribution, texture, and print-direction effects, so the number you see in a paper is usually a measured test result, not a permanent catalog constant. For polymer parts that slide, rub, guide, grip, or wear against another surface, it is smarter to read friction data as condition-specific tribology rather than as a single material promise. That is where good material selection starts. [a]
One number is never the full story. Pin-on-disk and reciprocating tests are both common in published tribology work, but they do not create the same contact mechanics, debris flow, or running-in behavior. A material can look calm in one setup and much more grabby in another. [b]
| Material | Representative Reported Dry COF | Typical Sliding Character | What Usually Drives the Result | Where It Often Fits Best |
|---|---|---|---|---|
| PLA | Static ~0.20–0.30; kinetic ~0.14–0.23 | Usually low to medium drag when geometry is stiff and the surface is controlled | Layer height, orientation, contact load, and steel counterpart finish | Guides, light-duty sliders, printed mechanisms that need stiffness more than grip |
| PETG | Static ~0.24–0.29; kinetic ~0.18–0.26 | Often a balanced sliding material with better toughness than PLA | Load sensitivity, stick-slip tendency, coating condition, and surface waviness | Wear-exposed parts, tougher sliding pairs, printed gears in light-duty service |
| ABS | About ~0.18–0.23 in one dry setup; can shift higher in other studies | Commonly medium friction, with surface finish playing a large role | Layer thickness, roughness, test geometry, frictional heating | General machine parts where heat resistance matters more than the lowest drag |
| Nylon / PA | ~0.04–0.20 as printed in one FDM tribology study; polishing can raise it to ~0.40–0.60 | Can be surprisingly slippery as printed, but finish can reverse that | Moisture, polish state, layer thickness, crystallinity, and transfer film behavior | Bushings, guides, dry-running parts, quiet gear-like motion |
| TPU | Roughly ~0.37–1.17 depending on hardness, infill, and surface pattern | Often high-grip rather than low-drag | Bulk deformation, real contact area, hardness, infill density, and counterpart roughness | Traction surfaces, damping parts, belts, anti-slip contact zones |
| Co-Polyester / PCL | Co-polyester ~0.29; PCL ~0.39 in one dry study | Co-polyester trends smoother; PCL trends grippier | Hardness and surface roughness | More specialized material studies than mainstream hobby use |
The table above is best read as a decision aid. If two materials sit close together, print settings and counterface finish can easily move the ranking.
Table of Contents
🧪 What a Friction Coefficient Actually Means
The coefficient of friction, usually written as μ, is the ratio between friction force and normal force. In sliding work, two values matter most: static friction, which resists the start of motion, and kinetic friction, which describes motion once sliding is already happening. That distinction matters a lot in printed parts, because a part that starts with a sticky breakaway feel can still slide smoothly after the first movement. [c]
- Static μ: breakaway resistance
- Kinetic μ: ongoing sliding resistance
- Higher μ: more grip, more drag
- Lower μ: easier sliding, lower drive force
That also explains why a high-friction material is not automatically a bad one. TPU, for example, may be the wrong choice for a dry-running slider but the right choice for a traction pad, belt contact face, or anti-slip interface. Printed tribology is application-specific. Always.
📊 Representative Reported Dry Friction Coefficients by Material
PLA and PETG
In dry contact against a steel counterbody, one published comparison measured PLA with static COF values from about 0.20 to 0.30 and kinetic COF from about 0.14 to 0.23 as load moved from 30 N to 70 N. PETG in the same work ran a bit higher, with static COF around 0.24 to 0.29 and kinetic COF around 0.18 to 0.26. That puts both materials in the usable mid-friction zone, but PETG tended to hold onto a little more drag while still remaining attractive for wear-loaded parts because of its toughness and stable motion profile. [d]
ABS
ABS is one of the easier materials to misread. In one dry-sliding study, printed ABS showed friction coefficients in the range of 0.18 to 0.23. That sounds very tame. In practice, though, ABS can swing more than many users expect because its result depends strongly on surface state, heat buildup, and test geometry. It is often chosen for temperature tolerance first and friction second. [e]
Nylon / PA
Nylon is where the story gets more interesting. A 2024 FDM polyamide study reported that unpolished printed PA could reach a COF of about 0.05, which is very low for an as-printed polymer sliding dry. The same paper also reported that polishing pushed the COF up into the 0.40 to 0.60 range because the smooth, hard skin produced by extrusion was removed. That is a useful reminder: a part can become more attractive to the eye and worse in service at the same time. [f]
TPU
TPU behaves like an elastomer, not like a stiff engineering slider. In recent dry-sliding FFF results, dense TPU samples were reported around 0.371 to 0.465 average COF, while softer or less structurally supported configurations reached about 0.647, 0.729, and even 1.174. That spread is wide, but the logic is simple: more deformation creates more real contact area, and more real contact area often means more drag. For parts that must hold, brake, damp, or grip, that can be exactly what you want. [g]
ASA and HIPS
Public friction-coefficient datasets for ASA are still thinner than those for PLA, PETG, ABS, or nylon. What is available already shows a useful pattern: in a six-material wear study, ASA and HIPS landed near ABS in average friction force, while PETG and PA were lower and PLA was higher. So for outdoor-stable ASA parts, it is reasonable to think about weather resistance and heat stability first, then validate tribology directly on the final print settings. [h]
Less Common but Useful Reference Materials
Less common filaments can still help frame the landscape. In one dry study of printed polymers, co-polyester was reported around 0.29 mean COF, PLA around 0.30, PCL around 0.39, and ABS around 0.40. The pattern is not that every polyester is low-friction by default. It is that harder, smoother surfaces often slide with less drag when the counterface and loading stay controlled. [i]
What This Means in Practice
- PLA is often better for low-load sliding than people assume.
- PETG is usually the safer all-rounder when wear and toughness matter together.
- ABS can work well, but you usually choose it for heat tolerance before friction.
- Nylon / PA is the natural candidate for dry-running motion parts.
- TPU belongs on the grip side of the spectrum unless a tested formulation proves otherwise.
⚙️ What Changes the Number Most
Surface Roughness
Surface roughness is one of the biggest reasons published COF values do not travel cleanly from one paper to another. A broad review of as-printed polymer surfaces shows that AM process choice and print settings have a direct effect on roughness, and roughness changes the contact mechanics before the material chemistry even gets a chance to speak. On printed polymers, topography is part of the material system. [j]
Layer Height and Infill Density
Layer height and infill density do not move friction in a perfectly straight line, but they matter a lot. One parameter study on PLA and ABS found that lower layer thickness generally reduced roughness, and that the minimum COF points appeared at different infill and layer-height combinations for the two materials. That is why “100% infill is always best” or “0.2 mm is always fine” is too blunt for tribology work. [k]
Print Orientation
Orientation can change friction even when the filament stays the same. In a printed bronze/PLA tribology study, vertically oriented test pieces showed the highest friction tendency but the lowest wear depth. That is a neat example of why friction and wear should never be collapsed into one idea. The contact plane sees the layer architecture directly. [l]
Counterface, Load, Speed, and Lubrication
Most papers test printed polymers against steel, but even then the steel finish, normal load, speed, and lubrication state can move the result hard. Dry PLA and PETG against steel may sit in one mid-range band at 30 N and shift again at 70 N. Add oil, and both static and kinetic values usually drop. Change the motion from rotating to reciprocating, and the running-in path changes too. That is why a friction table without the test conditions is only half a table.
🧩 Low-Friction Parts and High-Grip Parts Need Different Filaments
People often ask for the “best” friction coefficient. That question is missing the application.
- For bushings, guides, and dry-running sliders, you usually want lower drag, calmer running-in, and stable wear. Nylon / PA is the standout starting point. PETG is often the safer fallback when moisture sensitivity or print difficulty pushes nylon out.
- For gears and motion parts, low friction helps, but wear resistance and heat handling matter just as much. PETG often punches above its weight here because it stays tougher than PLA and can outlast ABS in some printed gear datasets.
- For grip surfaces, feed rollers, anti-slip contact, seals, or damping interfaces, a low coefficient is not the goal. TPU can be the right answer precisely because it is not slippery.
- For weather-exposed motion parts, ASA may be the better structural choice even if public COF data is thinner, because UV stability and temperature behavior may dominate the design window.
- Low-Drag Bias
- Nylon / PA, some PETG setups, tribo-filled engineering filaments.
- Middle Ground
- PLA and ABS when print quality is controlled and duty is light to moderate.
- High-Grip Bias
- TPU, especially soft grades or low-density structures that deform more under load.
🔄 Wear and Friction Are Not the Same Thing
A low-friction part can still wear out early. A higher-friction part can still last longer. Printed gears are a good example: one gear study comparing ABS, PLA, and PETG reported that PETG had the best wear resistance, with wear depths lower than ABS and PLA, and also the longest service life in the tested set. That does not mean PETG is the lowest-μ material in every rig. It means the material made a better overall tribological system for that use case. [m]
This is where many friction charts fail the reader. They rank materials by one number and ignore transfer films, heat softening, ductility, debris, and contact geometry. Real motion parts do not ignore those things.
📐 How to Read a COF Table Without Getting Misled
- Check whether the value is static or kinetic. Breakaway feel and sliding feel are often different.
- Check the counterface. Steel, aluminum, polymer-on-polymer, and coated counterparts do not produce the same ranking.
- Check whether the surface was left as printed, sanded, polished, coated, or textured. That detail can flip the result.
- Check the load and speed. Soft materials and viscoelastic materials respond hard to both.
- Check moisture and lubrication state. Nylon and elastomeric materials are especially sensitive.
- Check whether the part was built for grip or for slip. TPU and tribo-filaments live on opposite ends of that spectrum.
One Useful Rule
If a paper does not tell you the material pair, surface state, and test mode, treat the coefficient as only a rough hint.
❓ FAQ
Is There One Universal Friction Coefficient for PLA?
No. PLA can move across a range depending on whether the value is static or kinetic, which counterface it slides against, and how the printed surface was finished. Load and print orientation also change the result.
Which Common Filament Is Usually Best for Low-Friction Motion Parts?
Nylon / PA is the strongest starting candidate among common filaments for dry-running motion parts. PETG is often the easier second choice when you want better printability and good wear behavior together.
Why Can Polishing Nylon Make Friction Worse?
Because the as-printed extruded skin can act as a smooth, hard surface layer. Removing that layer can change contact mechanics and raise the coefficient of friction instead of lowering it.
Does More Infill Always Lower Friction?
No. Infill changes stiffness and deformation, but the direction of the effect depends on the material. Dense TPU often slides with less drag than sparse TPU, while PLA and ABS each show their own best combinations of infill and layer height.
Is Lower Friction the Same as Better Wear Resistance?
No. A part can slide easily and still wear fast, or run with more drag and still last longer. Wear depends on more than friction alone, including heat, debris, toughness, hardness, and transfer-film behavior.
Is TPU a Low-Friction 3D Printing Material?
Usually no. TPU tends to be a higher-grip material, and its coefficient of friction can rise a lot when the structure is soft or under-supported. That makes it useful for traction and damping, not for every slider.
References
- ASTM G99 — Standard Test Method for Wear and Friction Testing with a Pin-on-Disk or Ball-on-Disk Apparatus (used here for the lab basis behind many published COF values; reliable because ASTM is a long-established standards body).
- ASTM G133 — Standard Test Method for Linearly Reciprocating Ball-on-Flat Sliding Wear (used here to explain reciprocating wear and friction testing; reliable because ASTM publishes formal test methods used across labs).
- Stoimenov et al., “Static and Kinetic Friction of 3D Printed Polymers and Composites” (used for the static/kinetic distinction and load-linked dry vs lubricated friction behavior; reliable because it is a peer-reviewed journal paper in Tribology in Industry).
- Stoimenov et al., “Static and Kinetic Friction of 3D Printed Polymers and Composites” (used for PLA and PETG dry static and kinetic COF values against steel; reliable because it reports direct measured tribology data in a peer-reviewed journal).
- Nomani et al., “Effect of Layer Thickness and Cross-Section Geometry on the Tensile and Compression Properties of 3D Printed ABS” (used for the reported ABS dry sliding COF range; reliable because it is indexed on ScienceDirect/Elsevier and reports experimental data).
- Wang et al., “Mechanical and Tribological Properties of FDM-Printed Polyamide” (used for nylon / PA as-printed vs polished COF behavior; reliable because it is a Tribology International article on ScienceDirect).
- Brăileanu et al., “Structure—Property—Performance Relationships in Thermoplastic Polyurethane: Influence of Infill Density and Surface Texture” (used for TPU average COF ranges across infill and hardness states; reliable because it is an open biomedical repository copy of a peer-reviewed article).
- Struž et al., “Wear and Dynamic Mechanical Analysis (DMA) of Samples Produced via Fused Deposition Modelling (FDM) 3D Printing Method” (used for the six-material comparison including ASA, HIPS, ABS, PETG, PLA, and PA; reliable because it is an open-access journal article with measured test data).
- Ramadan et al., “Tribological Properties of 3D Printed Polymers: PCL, ABS, PLA, Co-Polyester” (used for less common-material reference values and hardness/wear linkage; reliable because it is a peer-reviewed tribology paper with direct measurements).
- Golhin et al., “Surface Roughness of As-Printed Polymers: a Review” (used for the effect of AM process and settings on roughness; reliable because it is a review article in a Springer journal).
- Portoacă et al., “Optimization of 3D Printing Parameters for Enhanced Surface Quality and Wear Resistance” (used for layer height and infill effects on PLA and ABS friction behavior; reliable because it is a peer-reviewed experimental paper).
- Hanon et al., “Effect of Print Orientation and Bronze Existence on Tribological and Mechanical Properties of 3D-Printed Bronze/PLA Composite” (used for orientation-linked friction and wear trends; reliable because it is a Springer journal article with experimental tribology data).
- Tunalioglu et al., “Wear and Service Life of 3-D Printed Polymeric Gears” (used for the PETG vs PLA vs ABS gear wear comparison; reliable because it is a full open-access research article archived in PubMed Central).