Taguchi-Based Artificial Neural Network Modeling of Friction Process on Aluminum Alloy Reinforced with SiC Nanoparticles

Abstract

Nowadays, the aluminum metal matrix composites (AMCs) play an empirical role to improve the mechanical performance by various applications. Therefore, the secondary processes need to enhance the surface morphology of intermetallic phases with the appropriate reinforcing particles. In this research, Al7075- and ceramic-based nanosilicon carbide (SiC) were utilized to compose the metal matrix composites. These composites were subjected to friction welding for intermetallic surface modification with the various forging pressure and rotating speed. Initially, the AMCs were prepared with three (8–12) kinds of SiC weight proportions by the design of Taguchi L9 orthogonal array. As per the weight proportions, nine samples were prepared and then conducted the friction welding with 10–20 MPa of forge pressure and 1,650–2,050 rpm of rotational speed of spindle. Then, the entire nine specimens were allowed to conduct the tensile and microhardness test. During the mechanical test, the overall welded zone had higher mechanical properties than the base metal. Then, the artificial neural network was utilized to predict the output responses as per the designed concept of trial and error method. The overall predicted responses are nearly closed to the experimental values.