Determination of the friction coefficient as a function of load is a fundamental tribological test that maps how static and kinetic friction vary with applied normal force for a given sliding material pair. This article explains the underlying contact mechanics, standard test configurations such as pin-on-disk and reciprocating sliding, load-sweep protocols, sample requirements, and typical industrial applications, giving engineers and quality managers a practical guide to obtaining reliable friction–load curve data from accredited third-party laboratories.
Technical Overview
The coefficient of friction (COF, μ) is defined as the ratio of the tangential friction force (F) resisting relative sliding to the applied normal load (N), that is, μ = F / N. Classical Amontons–Coulomb theory treats μ as a constant property of a material pair, independent of load and apparent contact area. In reality, most engineering contacts exhibit a systematic dependence of μ on the normal load, and characterizing this μ–N relationship is essential for accurate design, simulation, and failure prevention.
The load dependence arises from the mechanics of real contact. According to the Bowden–Tabor adhesion model, friction force is proportional to the true contact area, not the nominal area. For a smooth spherical Hertzian contact under elastic loading, the true contact area grows roughly as N^(2/3), so the friction force scales as F ∝ N^(2/3) and μ decreases with load approximately as N^(-1/3). For fully conforming or plastically deforming contacts, the real area grows nearly linearly with N and μ remains comparatively constant. Superimposed on these mechanisms are ploughing and deformation terms, protective oxide and third-body films, and thermally activated softening at high loads, all of which shape the measured μ–N curve.
A load-sweep friction test therefore does more than report a single number. It reveals run-in behavior, steady-state friction levels, the onset of lubricant or coating breakdown, scuffing and galling thresholds, and stick–slip tendencies as a function of preload. The resulting curves provide direct inputs for analytical design calculations, finite element and multibody dynamic simulations, PV (pressure–velocity) limit definitions, and supplier qualification. Reliable, traceable measurement of these relationships is a core competence of accredited third-party laboratories operating under ISO/IEC 17025 quality systems.
Testing Samples
Metals and Alloys
Self-mated and dissimilar metal couples such as carbon steel on steel, stainless steel, aluminum alloys, copper alloys, and cast iron are commonly tested as machined pins, balls, blocks, or flat coupons. Surface finish (typically Ra 0.05–0.8 μm) must be documented because roughness strongly influences the low-load portion of the μ–N curve.
Polymers and Composites
Engineering plastics including PTFE, PEEK, POM, PA, UHMWPE, and their glass- or carbon-fiber reinforced grades are tested against steel or ceramic counterfaces. Polymer specimens require conditioning to a standard atmosphere (e.g., 23 ± 2 °C and 50 ± 5% relative humidity per ISO 291 or ASTM D618) before testing, since moisture content alters friction behavior.
Flexible Films and Sheet Materials
Packaging films such as PE, PP, PET, and laminated structures are evaluated using horizontal-plane sled methods in accordance with ASTM D1894 or ISO 8295. Load functions are generated by varying the sled mass, which corresponds directly to varying wrap tensions on converting and filling lines.
Coatings and Surface Treatments
PVD/CVD hard coatings, DLC layers, anodized films, nitrided cases, electroplated zinc and zinc-alloy coatings, phosphated conversion layers, and dry-film lubricants are tested on coated coupons against standardized counterbodies (typically AISI 52100 balls or hardened pins) to determine how μ evolves as coating integrity responds to increasing load.
Lubricated Couples
Metal pairs running under oils, greases, forming lubricants, or thread compounds are tested in dip, drop, or bath-fed configurations to determine the load at which the lubricant film transitions from hydrodynamic or mixed lubrication to boundary lubrication and eventual breakdown.
General specimen requirements include defined counterface hardness and roughness, solvent cleaning immediately before testing, avoidance of fingerprint contamination, and a minimum of three replicate specimens per load condition to support statistical evaluation.
Testing Items
A comprehensive load-function friction study typically delivers the following measured and derived quantities:
- Static coefficient of friction (μs) at each discrete normal load level
- Kinetic (dynamic) coefficient of friction (μk) as a function of normal load
- Breakaway force or breakaway torque versus preload, relevant to fasteners and actuation mechanisms
- Run-in transients and time to steady-state friction at each load step
- Stick–slip amplitude and frequency as a function of load and sliding speed
- Critical transition loads marking surface film breakdown, scuffing, or galling onset
- Wear volume and specific wear rate versus load, analyzed through the Archard wear law (V = k · W · L / H)
- Contact temperature rise and its correlation with friction level
- Fitted μ–N model parameters, for example power-law fits μ = a + b · N^(-m) or μ = μ0 · N^(-n)
- Statistical summaries (mean, standard deviation, coefficient of variation) for each load condition
Testing Methods
Pin-on-Disk Testing (ASTM G99)
A stationary pin or ball of defined diameter is pressed against a rotating disk under a controlled normal load while the tangential force is continuously measured. Typical configurations use a 5–10 mm pin, track radii of 10–30 mm, sliding speeds of 0.01–1 m/s, and normal loads from 1 N up to several hundred newtons. Performing a stepped load ladder on a single track allows efficient generation of a μ–N curve, while separate tracks at constant load provide data free of sequential wear effects.
Linear Reciprocating Testing (ASTM G133)
A ball or flat-ended pin oscillates against a flat specimen with a typical stroke of 5–20 mm at frequencies of 1–50 Hz. Reciprocating motion reproduces the reversal and boundary-lubrication conditions found in slider bearings, seals, and biomedical joints, and permits compact load sweeps on small specimens.
Block-on-Ring Testing (ASTM G77)
A rectangular block is loaded against a rotating ring, generating a nominal line contact under severe sliding conditions. This method is well suited to higher load ranges and to ranking materials, coatings, and lubricants under conditions approaching scuffing failure, making the load-at-failure a key deliverable.
Sled Methods for Films (ASTM D1894 / ISO 8295)
For flexible films, a sled of defined mass is dragged across the film surface at a constant speed, and the force trace yields both static and kinetic COF. Load functions are built by stacking calibrated masses on the sled, with the standard 200 g configuration serving as the reference point for packaging-industry specifications.
Additional Standards and National Methods
ISO 6601 governs friction and wear testing of plastics against counterfaces, GB/T 3960 covers sliding friction and wear of plastic materials in Chinese national practice, and GB/T 10006 mirrors ISO 8295 for plastic film COF. ASTM G115 provides an authoritative guide for measuring and reporting friction coefficients, including the documentation practices that all credible laboratories follow.
Load-Sweep Protocols and Data Reduction
Two complementary protocols are used. In the stepped ladder method, the normal load is increased in discrete increments (for example 1, 2, 5, 10, 20, and 50 N), each step held for a fixed sliding distance or time until friction stabilizes, and the steady-state portion of the force trace averaged to compute μ at that load. A subsequent descending sequence verifies whether wear-induced hysteresis has biased the ascending results. In the continuous ramp method, closed-loop actuators increase the load at a controlled rate while instantaneous μ = F / N is computed point by point, exposing transition loads with high resolution. Results are plotted as μ versus N with error bars from replicate runs, and regression fits quantify the load-sensitivity exponent for design use.
Instrumentation, Calibration, and Quality Control
Modern tribometers apply the normal load through dead weights, pneumatic cylinders, or servo-electromagnetic actuators with closed-loop force control, and measure tangential force with calibrated transducers at sampling rates high enough to resolve stick–slip. Environmental chambers extend testing from -40 °C to 300 °C with controlled humidity. Accredited third-party laboratories maintain force and temperature traceability to national metrology institutes, include pin and counterface material certificates in the test dossier, and report measurement uncertainty in accordance with their ISO/IEC 17025 scope.
Applications
Load-dependent friction data directly support engineering decisions across many industries:
- Automotive engineering: brake pad compressibility and fade analysis, piston ring–liner and valvetrain friction maps, transmission synchronizer and clutch calibration, and body fastener torque–tension characterization through the nut factor relationship
- Fastener and threaded joint engineering: determination of how thread coatings and lubricants change the torque–tension coefficient across the preload range encountered in production tightening
- Packaging and converting: film COF verification against line-speed and wrap-tension windows to prevent slip faults and web breaks on filling machines
- Metal forming: selection of drawing, stamping, and rolling lubricants based on friction stability as die pressure rises, including galling thresholds for coated tooling
- Bearings and bushings: construction of PV limit diagrams for polymer bushings and dry-film lubricated joints from friction–load–speed matrices
- Medical devices: UHMWPE-on-cobalt-chromium bearing couples for implant wear and friction qualification
- Railway and heavy machinery: wheel–rail and coupler friction management, including friction modifier evaluation under axle-load-representative contact pressures
- Simulation and R&D: supply of validated μ–N input curves for finite element contact models, multibody dynamics, and analytical joint designs
- Quality control and dispute resolution: independent, traceable verification of supplier COF claims and root-cause support for field friction failures
FAQ
Why is the friction coefficient not constant as load increases?
Because the true contact area does not grow in exact proportion to load. In elastic Hertzian contacts the real area grows as N^(2/3), causing μ to fall with load, while plastically conforming contacts keep μ nearly constant. Additional effects such as oxide film behavior, ploughing, flash temperature, and lubricant film collapse further reshape the curve at the extremes of the load range.
What normal load range can be tested?
Typical laboratory tribometers cover 0.1 N to 500 N, and specialized block-on-ring or servo-hydraulic rigs reach several kilonewtons. The load ladder is selected to bracket the application pressure range, usually with at least five points spanning an order of magnitude.
How are static and kinetic coefficients distinguished?
The static coefficient is computed from the peak breakaway force at the onset of sliding, while the kinetic coefficient is the average of the tangential force divided by normal load over the steady sliding interval. Both are reported at every load step so that the full μs(N) and μk(N) functions are documented.
How long is each load step held?
Each step is maintained until the friction trace reaches steady state, commonly 5 to 30 minutes or a defined sliding distance, and the average is taken over the final portion of the step. Shorter steps risk capturing run-in transients that inflate apparent load sensitivity.
Can tests be run with lubricants or at elevated temperature?
Yes. Lubricated fixtures, heated stages, and environmental chambers allow load sweeps under boundary, mixed, or hydrodynamic lubrication and at temperatures relevant to service, enabling construction of Stribeck-type maps with load as an explicit variable.
What does a complete test report include?
Reports conforming to ASTM G115 guidance include material and counterface specifications with hardness and roughness, cleaning and conditioning procedures, full test parameters per step, raw friction traces, μ–N curves with replicate scatter, fitted model parameters, wear measurements, observed transition loads, and a stated measurement uncertainty.
How many specimens are needed for valid statistics?
A minimum of three replicate tests per material pair and load sequence is standard practice, with additional replicates when scatter exceeds roughly 10% relative standard deviation, which is common for dry sliding and boundary-lubricated conditions.