Development of predictive models for wire electrical discharge machining of aluminium metal matrix composites by using regression analysis and neural network

• Published in: Materials today : proceedings Vol. 68; pp. 1581 - 1587
• Main Authors: Surendranatha, G.M., Vishnu Vardhana Naidu, B., Gangaraju, Moopuri, Sarapure, Sadanand, Hemanth, KR
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 2022
• Subjects: Aluminium; MMC; Neural Network; Prediction; Regression Analysis; WEDM; Regression Analysis; MMC; Neural Network; Aluminium; Prediction; WEDM
• ISSN: 2214-7853; 2214-7853
• DOI: 10.1016/j.matpr.2022.07.258

Metal Matrix Composites (MMC) play a major role in numerous engineering uses such as aerospace, automobile and structural industries due to their exceptional characteristics such as lightweight, improved strength and resistance to wear. Amongst numerous contemporary approaches of machining, Wire Electrical Discharge Machining (WEDM) attracts attention due to its performance, especially in the machining of hard materials with complex shapes. In this present exploration, an endeavour was done to evolve predictive models for WEDM of aluminium-based metal matrix composites reinforced with Boron Nitride (BN). Pulse on time, pulse off time and peak current are considered independent parameters. Taguchi’s approach has been adopted for planning the experiments and to analyze the influence of process variables on desirable performance measures such as dimensional deviation and form/orientation tolerance errors. Neural network and multiple regression approaches were adopted for forecasting the aforementioned process variables. A comparison has been made among the predicted values from regression and neural network methods. The outcome of the comparative analysis proves that the neural network precisely predicts the desired performance.