Effect of B₄C/Gr on Hardness and Wear Behavior of Al2618 Based Hybrid Composites through Taguchi and Artificial Neural Network Analysis

Abstract

Artificial neural networks (ANNs) have recently gained popularity as useful models for grouping, clustering, and analysis in a wide range of fields. An ANN is a kind of machine learning (ML) model that has become competitive with traditional regression and statistical models in terms of useability. Lightweight composite materials have been acknowledged to be the suitable materials, and they have been widely implemented in various industrial settings due to their adaptability. In this research exploration, hybrid composite materials using Al2618 reinforced with B4C and Gr were prepared and then evaluated for hardness and wear behavior. Reinforced alloys have a higher (approximately 36%) amount of ceramic phases than unreinforced metals. With each B4C and Gr increase, the wear resistance continued to improve. It was found that microscopic structures and an appearance of homogenous particle distribution were observed with an electron microscope, and they revealed a B4C and Gr mixed insulation surface formed as a mechanically mixed layer, and this served as an effective insulation surface that protected the test sample surface from the steel disc. The ANN and Taguchi results confirm that load contributed more to the wear rate of the composites.

Keywords

• B₄C
• Gr
• Al2618
• delamination wear
• MMCs
• ANN