Enhanced mechanical and machinability of AZ91D hybrid composites via TOPSIS and machine learning models

Abstract

This study aims to optimize and predict machining outcomes for AZ91D magnesium alloy hybrid composites reinforced with aluminum oxide (Al₂O₃) and graphite (Gr), to enhance both machinability and mechanical properties. TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) was used to optimize turning parameters, including cutting speed, feed rate, and depth of cut. Three hybrid composites (AZ3Al1Gr, AZ2Al2Gr, AZ1Al3Gr) were evaluated using three tool types: untreated (T1), cryogenic-treated for 24 h (T2), and cryogenic-treated for 48 h (T3). The mechanical testing results showed that AZ3Al1Gr exhibited the highest tensile strength (295 MPa) and hardness (124 HV), while AZ1Al3Gr displayed the lowest values (279 MPa, 105 HV), highlighting a trade-off between strength and machinability. TOPSIS optimization identified the AZ2Al2Gr composite with the following parameters as the most optimal configuration: Tool 2 (Cryogenic, 24 h), Cutting Speed: 140 m/min, Feed: 0.24 mm/rev, and Depth of Cut: 0.9 mm. The ensemble machine learning model demonstrated high prediction accuracy, with R² values of 0.9303 for SR, 0.9621 for CF, 0.996 for CT, and 0.9596 for TW. SEM analysis revealed that graphite particles were displaced during machining, leading to increased surface roughness, whereas Al₂O₃ particles remained intact, enhancing hardness, and wear resistance. The surface quality was influenced by the interaction between the reinforcements, confirming their impact on machinability and surface integrity. This work provides new insights into hybrid composites and the use of cryogenically treated tools, offering an innovative approach to machining optimization. Future research will explore the real-time application of machine learning models for dynamic optimization in industrial settings.