Effect of Mg on the Microstructure, Mechanical Properties, and Surface Roughness of Functionally Graded A413 Composite: Machine Learning Approach at Wire-Cut Electric Discharge Machining Zone

Abstract

The present study assists the industrial sectors by implementing low-weight high-strength aluminium composite. The pristine magnesium as a reinforcement in three different weight percentages (3.5, 7, and 10.5) is utilized to fabricate the A 413 composite using centrifugal casting. The density analysis concludes a higher level of Mg inclusion prompts the material to save 3.9% of weight. Microstructural study of centrifugal casting specimens reveal a rich existence of hard phase Mg₂Si in the outer zone, with its richness decreasing toward the inner zone. The tensile test result affirms the functionally graded nature of the material with the increasing trend of tensile strength from the inner to the outer. Fractured surface analysis concludes the existence of cracks and their propagation for the inner region specimen. The influence of magnesium on the surface finish of functionally graded composite at the wire cut electric discharge machining (WEDM) zone is predicted using a machine learning-based random forest regressor (RFR) algorithm. The supervised learning algorithm declares that 10.5 wt% of Mg facilitates the minimal surface roughness of 0.229 µm with the optimal combination of parameters 6 A of current, 115 µs of pulse-on time, 60 µs of pulse-off time, and 8 N of wire tension.