Predictive Modeling of Tool Life in Turning Using ANN-Taguchi Hybridization

Abstract

A tool that will last is crucial for refining machining processes, influencing the quality of products, and reducing the expenses of making them. Previous research has demonstrated that several factors influence tool longevity, including the cutting depth, speed, the feed rate, the properties of the tool’s material, and those of the workpiece being machined. Understanding precisely how each of these factors impacts tool life is essential for refining processes and choosing the right tool. There are established mathematical models for estimating the tool’s lifespan, particularly for CBN tools when performing turning operations. Nonetheless, understanding the link between tool lifespan and cutting speed is challenging, given that it does not follow a linear pattern. Traditional methods for determining the equation for a tool lifespan using cutting speed often necessitate conducting tests at various speeds, which may not provide the statistical foundation required by the design of experiment (DoE) techniques. In this research, we delve into the complex relationship between tool lifespan and cutting speed through experiments guided by the Taguchi method and artificial neural network (ANN) models. Several case studies have been conducted to test the practicality and effectiveness of this method in representing complex tool lifespan-cutting speed relationships.