Tribological behavior of natural rubber: a multivariate analysis using friction and mechanical properties

Abstract

This study presents a comprehensive mathematical modeling framework for evaluating the friction and wear behavior of rubber materials used in engineering applications. Owing to rubber's nonlinear viscoelastic nature and its sensitivity to temperature, pressure, and interfacial phenomena, tribological modeling remains complex. This work develops an empirical multivariate regression model that relates the rubber friction coefficient (μ) to Shore A hardness (H) and ultimate tensile strength (σₜ), using a power-law relationship. Experimental data from standardized IRHD hardness tests and ASTM D412-compliant tensile tests were employed. Data analysis was performed using logarithmic transformations and ordinary least squares (OLS) regression. The model was validated through residual analysis and performance on new datasets. Results demonstrate high predictive accuracy (R² > 0.95), with model parameters yielding physical insights into rubber's surface deformation and internal strength mechanisms. The study emphasizes the model’s practical utility in tire design, seal optimization, and predictive maintenance planning, and paves the way for further development incorporating surface energy, temperature dependency, and wear rate parameters.