Enhanced mechanical and tribological performance of AA332 aluminium alloy reinforced with ZrB2 using integrated experimental and machine learning approaches
摘要
This scientific research systematically investigates the mechanical and tribological behavior of AA332 aluminium alloy composites reinforced with 0-7.5 wt% zirconium diboride (ZrB2) particles, synthesized via stir casting. Qualitative microstructural analysis revealed a uniform dispersion of ZrB2 particles and strong interfacial bonding, which quantitatively resulted in a significant grain refinement of 46.6%. Mechanical characterization demonstrated substantial quantitative improvements: Brinell hardness increased by 33.0% (from 92.0 to 122.4 BHN), ultimate tensile strength by 32.9% (from 155 to 206 MPa), and yield strength by 67.4% (from 92 to 154 MPa). A characteristic strength-toughness trade-off was observed, characterized by an impact energy reduction from 3.8 to 2.0 J and a qualitative transition from ductile to mixed-mode fracture mechanisms. Tribological performance was evaluated and optimized using Response Surface Methodology (RSM), identifying that a combination of 7.5 wt% ZrB2, 200 rpm sliding speed, and 25 N load yields a minimal specific wear rate of 0.843 × 10⁻⁴ mm³/Nm. This represents an 80.4% reduction in wear rate compared to the maximum wear condition observed in the 2.5 wt% ZrB2 composite. Predictive modeling using Artificial Neural Networks (ANN) validated the RSM findings, achieving superior predictive accuracy (Testing R² = 0.9947), with feature importance analysis quantitatively establishing ZrB2 content as the dominant predictor (68.3%) of wear behavior. This integrated experimental and data-driven approach highlights the potential of ZrB2-reinforced AA332 composites for structural applications requiring elevated strength and wear resistance.