Rubber Friction Coefficient Versus Load Relationship Testing: Principles, Methods, and Applications

The rubber friction coefficient versus load relationship experiment measures how the coefficient of friction (COF) of rubber materials changes as the normal load increases, delivering essential design data for tires, seals, belts, and anti-slip products. Unlike rigid solids, rubber rarely obeys Amontons' classical law, and its friction coefficient typically decreases at higher loads due to sublinear growth of the real contact area. This article explains the underlying theory, standard test methods, sample requirements, data interpretation, and industrial applications of load-dependent rubber friction testing.

Technical Overview

Rubber friction differs fundamentally from the friction of rigid solids. The total friction force of an elastomer sliding against a counter-surface is generally decomposed into an adhesion component, generated by molecular bonding and shearing at the real contact spots, and a hysteresis component, produced by viscoelastic energy dissipation as surface asperities deform the rubber during sliding. Because rubber is soft and compliant, the real contact area grows with the applied normal load, but it grows sublinearly rather than proportionally.

This sublinear contact growth has a direct mathematical consequence. The measured friction force F typically follows an empirical power law of the form F = k · Nⁿ (rendered conceptually here), where N is the normal load, k is a constant, and the exponent n usually lies between 0.6 and 0.95, that is, below unity. The coefficient of friction is therefore μ = F / N = k · N^(n-1), which means μ decreases as the normal load increases. In practical terms, doubling the load does not double the friction force, and the friction coefficient read at one load cannot be extrapolated linearly to another.

Several additional phenomena make load-dependent behavior even more pronounced. At low loads and high adhesion, sliding may occur through Schallamach waves, folds of rubber that propagate through the contact zone and produce an apparently low friction force. At intermediate loads, stick-slip oscillations may appear, especially at low sliding speeds. Because rubber is viscoelastic, the measured friction is also a function of sliding speed, temperature, and counter-surface roughness, so a rigorous experiment must control or record all of these variables while sweeping the normal load. Standardized guidance for such measurements is provided by ISO 15113, ASTM D1894, ISO 8295, ASTM D2047, and ASTM G115, and many organizations rely on accredited third-party laboratories to obtain traceable, reproducible data.

Testing Samples

Elastomer Families

Virtually all technically relevant elastomers can be evaluated, including natural rubber (NR), styrene-butadiene rubber (SBR) and butadiene rubber (BR) blends used in tire treads, nitrile rubber (NBR) and hydrogenated nitrile (HNBR) for oil-resistant seals, ethylene-propylene rubber (EPDM) for automotive weatherstrips, fluorocarbon rubber (FKM) for chemical service, silicone rubber (VMQ) for medical and food contact, chloroprene rubber (CR), and thermoplastic polyurethanes (TPU). Filled compounds, covering a hardness range of roughly Shore A 30 to 90, are the most common subjects because carbon black and silica loading strongly influence both adhesion and hysteresis friction.

Specimen Geometry

The most widely used configuration is a flat sheet coupon, typically 2 to 8 mm thick, bonded to a rigid sled backing to prevent wrinkling and edge curling. A common reference footprint is the 63.5 mm square sled specified in ASTM D1894. For research-grade load sweeps, cylindrical button or pin specimens of 5 to 10 mm diameter are also used on rotary or reciprocating tribometers, because the smaller, well-defined contact zone simplifies contact pressure calculation. Specimens must be free of mold release agents, bloom, and surface contamination.

Counter-Surfaces

The mating surface must be defined and documented. Typical choices include smooth float glass (the ASTM D1894 reference surface), polished steel with a measured roughness of Ra 0.05 to 0.8 μm, anodized aluminum, ceramic tile, abrasive paper, or concrete. Surface roughness and preparation method (grinding, polishing, lapping) must be measured and reported, since roughness shifts the balance between the adhesion and hysteresis friction components and can change the load exponent n.

Conditioning

Prior to testing, specimens and counter-surfaces are conditioned in a standard laboratory atmosphere of 23 ± 2 °C and 50 ± 5 % relative humidity for a minimum of 40 hours, following the conditioning practice of ISO 23529 or ASTM D1349. Surfaces are cleaned with isopropanol or a specified solvent and allowed to dry fully, and tests should begin within a defined time window after cleaning to limit recontamination and re-blooming.

Testing Items

A complete load-relationship study characterizes the following quantities:

• Static coefficient of friction (breakaway) at each normal load level
• Kinetic coefficient of friction during steady sliding at each load level
• Raw friction force versus normal load dataset across the full load sweep
• Coefficient of friction versus normal load curve, with regression of the power-law exponent n and coefficient k
• Average contact pressure for each load step, calculated from the measured contact geometry
• Stick-slip amplitude and frequency where oscillatory sliding occurs
• Optional cross-matrices: sliding speed sweep, temperature sweep, and counter-surface roughness sweep at fixed loads
• Drift of friction force with sliding distance, indicating running-in or wear effects within the test

Testing Methods

Test Apparatus

Three instrument types dominate this field. The horizontal sled-type tribometer, in the style of ASTM D1894, drags a loaded sled across a flat counter-surface while a force transducer records the drag force; it is well suited to flat rubber sheets at moderate loads. The rotary pin-on-disc tribometer presses a rubber pin against a rotating disc and is preferred for high contact pressures and continuous long-duration sliding. Linear reciprocating tribometers reproduce the back-and-forth motion found in seals and wipers. Normal load is applied either by calibrated dead weights or by a servo-controlled actuator, and friction force is measured with a load cell accurate to 0.5 % of reading or better. For static-only measurements, inclined-plane (ramp) instruments described in ISO 8295 can be used, although they are limited to the breakaway condition.

Load Schedule

To resolve the load relationship reliably, a minimum of five to seven load levels should be tested, geometrically spaced so that they span at least one decade of normal load or contact pressure. A typical practical range covers 0.5 N to 200 N, corresponding to nominal contact pressures of roughly 5 kPa to 1 MPa for common sled geometries. Each load level should be repeated at least three times, and the test sequence (ascending, descending, or randomized loads) should be documented because friction memory and warm-up effects can bias results.

Test Procedure

After conditioning and cleaning, the specimen is mounted and aligned so that the sliding interface is parallel to the drive axis. The lowest normal load is applied and the system is allowed to settle. The static friction force is captured as the peak breakaway force at the start of motion. Sliding then proceeds at a fixed speed, commonly 100 to 150 mm/min for sled-type tests in line with ASTM D1894, or 1 to 100 mm/s in research configurations, and the kinetic friction force is averaged over a steady-state window of 50 to 200 mm of travel. The normal load is then incremented to the next level, a rest interval of several minutes is observed to allow thermal recovery of the interface, and the measurement is repeated until the full load matrix is complete. Temperature, humidity, and surface condition are logged continuously.

Data Processing and Curve Fitting

At every load step, the coefficient of friction is computed as μ = F / N for both the static peak and the steady-state kinetic plateau. A logarithmic regression of friction force against normal load yields the power-law constants k and n, with the coefficient of determination reported as an indicator of fit quality. The final test report presents the μ versus load curve, the fitted exponents, the individual μ values at each reference load, and an uncertainty statement. Well-run programs calibrate the load cell against certified class weights, verify sled parallelism, and periodically verify the instrument with a reference material, consistent with the quality practices expected of accredited third-party laboratories operating under ISO/IEC 17025 management systems.

Applications

Load-dependent friction data are used across a broad range of industries:

• Tire engineering: μ versus contact pressure curves feed semi-empirical tire models and guide tread compound development for wet grip and braking performance under realistic wheel loads
• Seals and O-rings: breakout and running friction as a function of squeeze load determines actuation forces in hydraulic and pneumatic systems and helps designers avoid stick-slip
• Conveyor belts: the friction between the belt and drive pulleys must be characterized across the tension and load range to prevent slippage and to size drive systems correctly
• Footwear and flooring: slip resistance is evaluated under loads representative of body weight, with ramp and sled methods aligned to EN 16165 and related national procedures
• Anti-vibration mounts and bushings: frictional damping under varying preloads is a key input for vibration isolation design
• Soft grippers and robotics: grasping force calculations for elastomer pads require accurate COF values at low contact pressures
• Simulation and FEA: the fitted power-law constants k and n are inserted directly into contact models to predict friction behavior across operating load ranges

FAQ

Why does the friction coefficient of rubber decrease as the load increases?

The real contact area between rubber and a rough counter-surface grows with load but grows less than proportionally, because the material deforms and progressively conforms to the surface. Since the adhesion component of friction scales roughly with real contact area, friction force follows F = k · Nⁿ with n below 1, so the ratio μ = F / N declines as N rises. Hysteresis friction also saturates at high loads once the asperity deformation field fully develops.

Which standards govern this type of testing?

The most relevant references are ISO 15113 for the determination of frictional properties of rubber, ASTM D1894 for sled-type static and kinetic friction measurement, ISO 8295 for the inclined-plane method, ASTM D2047 for sole and heel materials, and ASTM G115 as a general guide for measuring and reporting friction coefficients. Corresponding GB/T and EN methods exist for specific product sectors.

How many load levels are needed to characterize the relationship?

At least five geometrically spaced load levels spanning one decade of load or contact pressure are recommended, each with a minimum of three replicates. This provides sufficient degrees of freedom for a stable log-log regression of the exponent n and a defensible uncertainty estimate.

What is the difference between static and kinetic coefficients in these tests?

The static coefficient corresponds to the breakaway peak force recorded at the instant sliding begins, while the kinetic coefficient is calculated from the averaged force during steady sliding. In rubber, the static value is usually higher, and the gap between them widens at low speeds, which is the root cause of stick-slip in seals and slow-moving mechanisms.

How do sliding speed and temperature affect the results?

Because rubber is viscoelastic, both adhesion and hysteresis friction depend on the deformation frequency, which is set by sliding speed, and on temperature. Friction typically passes through a maximum near the glass transition region. Speed and temperature must therefore be fixed, reported, and held constant during a load sweep, or deliberately mapped using time-temperature superposition when a full characterization is required.

What sample dimensions and preparation are required?

A flat coupon at least 2 mm thick, backed by a rigid sled with a 63.5 mm square face, is the standard configuration for sled testing; smaller pins of 5 to 10 mm diameter suit high-pressure tribometry. Specimens must be conditioned at 23 ± 2 °C and 50 ± 5 % relative humidity for at least 40 hours, cleaned of bloom and release agents, and tested with a documented counter-surface roughness.