Effect of grain refinement and TiB₂ reinforcement on machining behavior of Al-4.5%Cu-3%TiB₂ composites: A hybrid experimental-ML approach

The machining of particle-reinforced aluminum matrix composites is governed by a complex interplay between microstructure driven strengthening mechanisms and thermo-mechanical interactions at the tool-workpiece interface. In this study, the machining behavior of an in-situ synthesized Al–4.5%Cu–3%TiB₂ composite is systematically investigated by integrating experimental characterization with machine learning based predictive modeling. Detailed microstructural analysis using optical microscopy, SEM, TEM, and XRD confirms uniform dispersion of TiB₂ particulates, refined equiaxed grains, and the presence of strengthening Al₂Cu precipitates. TEM observations reveal particle stimulated nanograin formation via dynamic discontinuous recrystallization, contributing to enhanced thermal stability and resistance to machining induced softening. Turning experiments were conducted under varying cutting speed, feed rate, and depth of cut, and the corresponding cutting, radial, and feed forces were measured using a piezoelectric dynamometer. The results demonstrate that, unlike conventional aluminum alloys, machining forces increase with cutting speed due to the retention of hardness and thermal stability imparted by TiB₂ reinforcement. To quantitatively correlate microstructure induced strengthening with machining responses, multiple machine learning models (ANN, SVR, RF, GBR, XGBoost, and KNN) were developed and evaluated. Among them, XGBoost exhibited superior predictive accuracy (R² > 0.999, MAPE < 1.5%). Quantitatively, the composite exhibited refined grains of approximately 18 μm, and machining forces increased by nearly 15–25% with increasing cutting speed within the investigated range, highlighting the strong influence of thermally stable TiB₂ reinforcement on machining resistance. Feature importance analysis confirms cutting speed as the dominant parameter governing force evolution. The strong agreement between experimental results, microstructural observations, and ML predictions establishes a robust microstructure-machining-modeling framework for TiB₂ reinforced aluminum composites.