Tribological and Predictive Modelling Analysis of Copper - CNT - Titanium Hybrid Metal Matrix Composites

This paper investigates the wear behaviour of copper-based metal matrix composites (MMCs) reinforced with carbon nanotubes (CNTs) and micro titanium (Ti) particles. Composites containing 0.5–1.5 Wt.% CNT and 1–5 Wt.% Ti were fabricated using the stir casting technique to ensure uniform reinforcement dispersion. Abrasive wear tests were conducted under a 1 N load at sliding speeds of 200–500 RPM using a pin-on-disc tribometer. Results revealed that the wear rate increased with speed but decreased notably with higher reinforcement levels. The C9 composition (1.5Wt.% CNT and 5Wt.% Ti) exhibited the lowest wear rate and superior wear resistance. Regression modelling using Linear, Polynomial, Support Vector Regression (SVR), and Random Forest algorithms was applied to predict wear behaviour. Among these, the Polynomial model demonstrated the highest accuracy with an R ² value close to 0.99 and low mean absolute error, while SVR showed consistent but less interpretable results. ANOVA analysis confirmed that both composite composition and speed significantly influenced wear rate. The paper concludes that hybrid reinforcement of CNT and Ti effectively enhances the tribological performance of copper composites, and regression models offer reliable predictive insight into wear mechanisms.