Statistical and Intelligent Techniques for Modeling and Optimization of Duplex Turning for Aerospace Material

Abstract

Duplex turning becomes an important metal cutting process due to unique features like higher productivity with better surface finish at lower specific energy and vibration. Such process requires two-cutting tools which are mounted parallelly and fed inward to cut the material from rotating surfaces. Such complex process needs modeling and optimization to analyze the effect of factors and identify the optimal cutting condition. This paper focuses to develop two models related to statistical and intelligent techniques especially responses surface methodology (RSM) and artificial neural network (ANN) for prediction analysis of duplex turning. Based on prediction potential, the ANN model is utilized to analyze the effect of various parameters (cutting speed, feed rate, primary depth-of-cut (DOC) and secondary-DOC on the responses as surface roughness and cutting forces (primary and secondary). Further, the parameters are optimized using Taguchi Methodology (TM) and experimentally validated. The results show that ANN model predicts the data with more precision than RSM model. Further, the optimal data are experimentally validated and significantly agreed with predicted data of ANN model with percentage error as 2.24%, 1.40% and 0.75% for surface roughness, cutting forces (primary and secondary), respectively.