Meta-heuristic algorithm based optimization studies in cryogenic and conventional milling of magnesium alloy AZ91

Abstract

The surface finish of a machined product is essential for assessing its quality and other attributes. Modeling the surface roughness and hardness of a machined component is challenging for several reasons. The present study examines the effectiveness of four meta-heuristic algorithms in optimizing surface characteristics like roughness (R<inf>a</inf>) and hardness (HV) in the machining of magnesium alloy AZ91. Experiments with uncoated carbide inserts have been conducted under dry and cryogenic conditions. The study's input parameters are the depth of cut, feed rate, and cutting speed. Modeling and prediction studies have been conducted using Multi Layered Perceptron (MLP) Neural Network, and the output of this model has been considered as the objective function for the optimization algorithms. Algorithms, namely Particle Swarm Optimization (PSO), Bat Algorithm (BA), and recently developed algorithms, namely Jaya Algorithm (JAYA) and Fruit Fly Optimization Algorithm (FOA), have been tested. The optimization accuracy of FOA has been found to be superior to that of the other algorithms. As per the knowledge of the authors, this work probably presents a first attempt in applying the JAYA and FOA metaheuristic algorithms in the machining studies of an AZ series magnesium alloy. © 2025 The Authors