Wear Performance and Optimization of hBN-Modified AZ91/TiB₂ Magnesium Composites Under Dry Sliding Conditions

The present study focuses on the development and tribological evaluation of AZ91 magnesium matrix
composites reinforced with TiB₂ and hexagonal boron nitride (hBN) nanoparticles. The TiB₂ content was
fixed at 6 wt%, while hBN was varied at 0.5, 1, 1.5, and 2 wt.% to examine its influence on wear
performance. The composites were fabricated using the stir casting technique, followed by tribological
testing under dry sliding conditions. The wear experiments were designed based on an L16 Taguchi
orthogonal array, incorporating variations in normal load, sliding speed, and sliding distance. To establish
predictive insights, Gradient Boosting Regression (GBR) was employed to estimate wear rate and friction
coefficient, effectively capturing nonlinear dependencies between input parameters. Further, Multi
Objective Particle Swarm Optimization (MOPSO) was applied to determine the optimal combination of
reinforcement content and test conditions for minimizing wear and friction while maintaining mechanical
integrity. The results indicate that the hybrid composite with optimized hBN content exhibits enhanced
wear resistance, attributed to the combined effects of solid lubrication and microstructural reinforcement.
The self-lubricating nature of hBN reduces frictional heating and surface damage, whereas TiB₂
contributes to load-bearing capability and matrix strengthening. Experimental validation confirmed a
strong correlation between predicted and observed values, demonstrating the reliability of the proposed
modeling approach. This study highlights the potential of machine learning-driven optimization in
designing advanced Mg-based composites for applications in aerospace, biomedical, and automotive
industries, where high wear resistance and lightweight materials are critical.