Research Article

ANN-based predictive modelling of the effect of abrasive water-jet parameters on the surface roughness of AZ31 Mg alloy

¹ Department of Mechatronics, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal 576104, Karnataka, India
² Department of Mechanical and Industrial Engineering, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal 576104, Karnataka, India
³ Department of Artificial Intelligence and Machine Learning, Symbiosis Institute of Technology, Symbiosis International (Deemed University), Lavale, Maharashtra, Pune, India

* e-mail: sk.bhat@manipal.edu

Received: 18 February 2024
Accepted: 10 August 2024

Abstract

In today's world, there is an acute need to increase the usage of ecologically sustainable materials like AZ31 magnesium (Mg) alloy, possessing high strength-to-weight ratio and biocompatibility. However, its machinability through conventional machining techniques remains a challenge due to its high flammability. AWJM of Mg alloys is a promising method in this scenario. The present study investigated the effects of three important operating parameters, viz., stand-off distance (SOD), feed rate, and number of passes on the surface roughness parameters (R_a, R_q and R_z). Experiments were conducted based on Taguchi's L9 orthogonal array, and the effects of parameters on R_a, R_q and R_z were analysed statistically using analysis of variance (ANOVA). The results demonstrated that SOD and number of passes together have significant influence on the surface roughness (between 60% and 80% contribution). The individual and interaction results effects of parameters revealed that, SOD of 1–2 mm, feed rate of 130 mm/min and two cutting passes resulted in the best surface quality with least roughness (R_a, R_q < 3 μm and R_z < 12 μm). Finally, an artificial neural network model was developed with 7 neurons in the hidden layer, which simultaneously predicted R_a, R_q and R_z with high accuracy (R > 0.99).