Combined effects of ZrSiO 4 and graphite reinforcements on the mechanical and tribological behavior of stir-cast LM25 alloy composites

The research work investigates the development and characterization of LM25-based hybrid composites reinforced with varying amounts of graphite (Gr) and a constant 5% zirconium silicate (ZrSiO 4 ). The composites were developed using the stir casting process, and their mechanical and tribological properties were evaluated. Hardness, tensile strength, compressive strength, and wear analysis tests were conducted to assess the influence of graphite content on the characteristics of the composites. The microstructural features of the composites were analyzed using optical microscopy and scanning electron microscopy (SEM), revealing the influence of graphite on the grain structure and reinforcement distribution. The results indicated that the addition of graphite up to 9 wt.% improved the hardness, tensile strength, and compressive strength, while also reducing the wear rate and coefficient of friction. However, further increases in graphite content (12 wt.%) led to particle agglomeration, weak interfacial bonding, and a decrease in mechanical properties. Response surface methodology (RSM) was employed to optimize the tribological performance of the composites, identifying the optimal conditions for minimum wear rate and friction. The BAT algorithm was applied to optimize graphite content, load, and sliding distance to minimize wear rate and coefficient of friction (COF). The optimal parameters (9 wt.% graphite, 20 N load, and 1 km sliding distance) yielded model-predicted wear and COF values of 0.0079 g/km and 0.3107, respectively, demonstrating effective multi-objective optimization. To accurately predict the tribological behavior, a Deep Regression Network (DeepRegNet) is developed.