The kinetic coefficient of friction (COF) is one of the most influential tribological properties in engineering design, and surface roughness is among its primary physical drivers. This article presents a systematic experiment for characterizing the relationship between the kinetic friction coefficient and surface roughness parameters such as Ra, Rq, Rz and Sa, following recognized methods from ISO, ASTM and GB standards. It covers test principles, sample preparation, instrumentation, data analysis, typical findings and industrial applications for metals, polymer films, coatings and elastomers.
Technical Overview
The kinetic (dynamic) coefficient of friction, μk, is defined as the ratio of the tangential force required to maintain steady relative sliding between two surfaces to the normal force pressing them together. Unlike the static coefficient of friction, which describes the onset of motion, μk governs continuous sliding behavior and is therefore critical for machinery speed stability, film transport, bearing performance and slip resistance.
Surface roughness quantifies the micro-geometry of a surface using profile parameters such as Ra (arithmetic mean deviation), Rq (root-mean-square deviation), Rz (maximum height of profile), Rsk (skewness) and Rku (kurtosis) as defined in ISO 4287 and GB/T 3505, or areal parameters such as Sa and Sq as defined in ISO 25178. Because friction arises from the interaction of surface asperities, roughness directly controls the real contact area, the adhesive junction strength and the deformation resistance at the interface.
According to the classical two-term friction model, the total friction force consists of an adhesion shearing component and a plowing or asperity deformation component. As a result, the relationship between μk and roughness is generally non-monotonic and material dependent. For many hard metal pairs in dry sliding, friction increases with roughness because asperity interlocking and plowing dominate. For soft polymers, films and elastomers, extremely smooth surfaces can paradoxically produce higher friction because the real contact area and adhesive junctions grow. The objective of this experiment is therefore to measure μk across graded roughness levels under controlled load, speed, temperature and humidity, to establish a quantitative regression relationship, and to identify the optimum roughness window for a given material pairing.
Testing Samples
Metallic Specimens
Typical metallic samples include 304 stainless steel, bearing steel (for example GCr15/52100), 6061 aluminum alloy and brass, prepared by polishing, fine grinding, turning or milling to obtain target roughness grades spanning Ra 0.05 μm to 6.3 μm. A fixed counterface, such as a hardened steel ball or lapped disc, is used throughout the matrix so that roughness remains the single variable.
Polymer Films and Sheets
For flexible packaging studies, polyethylene (PE), polypropylene (PP), polyester (PET) and polyamide (PA) films with thicknesses of 30 μm to 200 μm are tested per ASTM D1894, ISO 8295 and GB/T 10006. Samples with polished chill-roll, matt and embossed surfaces provide a natural roughness spectrum without changing the base resin.
Rubber, Elastomers and Coatings
Sealing and safety applications use NBR, FKM, EPDM and thermoplastic polyurethane samples, while surface engineering studies include anodized aluminum, electroplated zinc-nickel, PVD coatings and PTFE-impregnated dry-film lubricants, each characterized before and after any run-in treatment.
Preparation and Conditioning
All specimens are cleaned with isopropanol or an equivalent mild solvent and dried to remove residual oils and release agents. Samples are conditioned for at least 40 hours at 23 ± 2 °C and 50 ± 5 % relative humidity in accordance with ISO 291 or ASTM D618. Specimen geometry follows the applicable standard, for example a 63 mm × 63 mm sled face sliding on a plane specimen at least 250 mm wide for ASTM D1894 sled testing.
Testing Items
The experiment is designed to capture both the tribological response and the complete surface texture signature. The measured items include:
- Kinetic coefficient of friction μk (primary output)
- Static coefficient of friction μs (comparative reference for stick-slip tendency)
- Profile roughness parameters Ra, Rq, Rz, Rsk, Rku and Rmr per ISO 4287 and GB/T 3505
- Areal roughness parameters Sa, Sq, Ssk and Sku per ISO 25178
- Normal load, sliding speed, sliding distance and ambient temperature/humidity logs
- Optional wear scar dimensions or mass loss per ASTM G99 for metal sliding pairs
- Statistical outputs: regression coefficients, Pearson correlation coefficient r, coefficient of determination R² and ANOVA significance levels
Testing Methods
Surface Roughness Characterization
Roughness is measured with a contact stylus profilometer using a diamond stylus with a 2 μm tip radius, a Gaussian filter cutoff λc of 0.8 mm and an evaluation length of 5 × λc = 4 mm, following the rules of ISO 4288 and GB/T 10610. Instruments are calibrated and verified in accordance with ISO 12179. At least five traces are taken at representative locations on each specimen, oriented along the sliding direction. Areal parameters are obtained by white-light interferometry or laser confocal microscopy per ISO 25178, typically over three 1 mm × 1 mm regions. For plateau-type stratified surfaces such as honed bores, the functional parameters Rk, Rpk and Rvk of ISO 13565 are additionally reported.
Kinetic Friction Measurement
For films and sheets, the horizontal plane sled method of ASTM D1894, ISO 8295 and GB/T 10006 is applied using a 200 g sled pulled at 150 mm/min (ASTM D1894) or 100 mm/min (ISO 8295). The force–displacement curve is recorded continuously: μs is taken from the initial force peak, while μk is calculated as the average tangential force over the steady-state segment of the stroke, typically between 15 % and 85 % of travel, with stick-slip oscillation regions excluded. For bulk metals, a pin-on-disc apparatus per ASTM G99 (for example a 6 mm pin under 5 N normal load at 0.1 m/s dry sliding) or a block-on-ring rig per ASTM G77 is used. Friction data acquisition, reporting formats and uncertainty statements follow the guidance of ASTM G115.
Test Matrix Design
A single-variable matrix is constructed with five to seven roughness grades, for example Ra values of 0.05, 0.1, 0.2, 0.4, 0.8, 1.6 and 3.2 μm, each paired with an identical fixed counterface. A minimum of five valid replicates per grade is tested, and the run order is randomized to neutralize drift in temperature, humidity and instrument state. Material pair, lubrication condition, normal load and sliding speed are held constant so that any change in μk can be attributed to surface texture.
Data Processing and Correlation Analysis
Results are expressed as mean ± standard deviation, with outliers screened by Grubbs testing. The kinetic friction coefficient is regressed against each roughness parameter using a power-law model of the form μk = μ0 + k · Ra^n, and the fits for Ra, Rq, Rz, Rsk and Sa are compared. Pearson correlation and one-way ANOVA at a significance level of α = 0.05 determine whether the roughness effect is statistically robust. Where the curve is non-monotonic, the minimum-friction roughness window is identified for engineering use.
Typical Findings and Interpretation
For dry steel-on-steel sliding, μk generally increases with Ra because plowing and third-body abrasion of asperities dominate the interface, with a strong positive correlation against Rz and Rq. For polyolefin films, the opposite initial trend is common: μk decreases as roughness increases from a mirror finish, because the real contact area and adhesive junctions shrink, then levels off or rises slightly once asperity interlocking becomes significant. An optimum range of approximately Ra 0.2 μm to 0.5 μm is frequently reported for chill-roll films intended for high-speed converting lines.
- Negatively skewed (plateau-like) surfaces with Rsk below zero retain lubricant in the valleys and typically show lower μk than Gaussian surfaces of equal Ra.
- High kurtosis Rku indicates sharp peaks that flatten during running-in, so post-run-in μk is lower and more representative of service behavior than first-slide values.
- The material ratio parameter Rmr at a given cutting depth correlates well with the load-bearing capacity and steady-state friction of honed or ground functional surfaces.
- Amplitude parameters alone are insufficient; combined amplitude, spacing and shape descriptors give the most reliable friction prediction.
Applications
The experiment supports surface specification, friction tuning and quality control across a wide range of industries:
- Flexible packaging: defining COF windows and chill-roll roughness for reliable vertical and horizontal form-fill-seal machinability.
- Automotive engineering: friction material development for brakes and clutches, door and glass-run seals, and wiper blade lip optimization.
- Precision machinery: linear guides, plain bearings and hydraulic cylinder rods where hard-chrome shaft roughness must balance friction, seal wear and lubricant retention.
- Flooring and pedestrian safety: correlating surface texture with slip resistance measured by horizontal pull or articulated strut methods.
- Medical devices: verifying the lubricity of hydrophilic catheter coatings and low-friction polymer components.
- Metal forming: selecting die surface textures that trap lubricant and reduce galling during stamping and deep drawing.
- Surface treatment verification: confirming that anodizing, plating, PVD or laser-texturing processes deliver the targeted friction-roughness combination in accredited third-party laboratories or in-house metrology centers.
FAQ
Does a rougher surface always produce a higher kinetic friction coefficient?
No. The trend depends on which friction mechanism dominates. For hard metals in dry contact, plowing and asperity interlocking usually make friction rise with roughness. For soft, compliant materials such as polymer films and elastomers, very smooth surfaces can increase friction because the real contact area and adhesive bonding grow. Many pairings exhibit a minimum-friction roughness somewhere between the extremes.
Which roughness parameter correlates best with kinetic COF?
Ra alone is rarely sufficient. Rq and Rz capture amplitude, Rsk and Rku describe peak shape and sharpness, and Rmr or the areal parameter Sa describes load-bearing area. In practice, a multi-parameter regression combining amplitude and shape descriptors predicts friction far better than any single value.
Which standards govern this experiment?
Friction measurement is covered by ASTM D1894, ASTM G99, ASTM G77 and ASTM G115 for metallic and film specimens, and by ISO 8295 with its Chinese equivalent GB/T 10006 for plastic films. Roughness characterization follows ISO 4287, ISO 4288, ISO 13565 and ISO 25178, together with GB/T 3505 and GB/T 10610. Conditioning atmospheres follow ISO 291 and ASTM D618.
How many specimens and replicates are required?
A minimum of five valid replicates per roughness grade is recommended to support statistical analysis, with additional specimens whenever stick-slip or visible wear transfer occurs. Each roughness measurement location should be taken in the actual sliding track so that the texture being correlated is the texture that generated the friction data.
Which test conditions must be strictly controlled?
Temperature and humidity, specimen cleanliness, normal load, sliding speed, counterface identity and the number of prior sliding passes all measurably shift μk. Conditioning at 23 ± 2 °C and 50 ± 5 % relative humidity, solvent cleaning before each run and a documented run-in protocol are essential for reproducible results.
Why do static and kinetic COF differ, and what is stick-slip?
The static COF reflects the force needed to initiate motion from rest and is usually higher than the kinetic value required to sustain sliding. When the difference is large relative to system stiffness, the motion becomes intermittent, alternating between sticking and rapid sliding. Recording the full force–displacement trace and averaging only the steady-state segment yields a robust kinetic COF and quantifies stick-slip severity for machine-speed-stability assessments.