Comparative investigations of cryo-treated and untreated inserts on machinability of AISI 1050 by using response surface methodology,ANOVA and Taguchi design

Abstract

This study aims to focus on the machinability of the AISI 1050 workpieces with cutting inserts, treated under deep cryogenic heat (−146 °C), and with untreated ones, and to investigate the optimization of cutting parameters and heat treatment conditions for surface roughness and cutting force by using Taguchi mixed design and Response Surface Methodology (RSM). The machining experiment was performed on a CNC lathe with machining parameters such as three feed rates, three cutting speeds and a constant depth of cut under dry condition and with heat treatment condition. As is known, Taguchi design L18 (3² 2¹) consists of three factors; cutting parameters with each one of three levels and heat treatment condition with two levels. The results of machining tests were evaluated considering surface roughness, vibrations and cutting force. Furthermore, chip morphology and wear led by cryo-treated and untreated inserts were detected with the aid of a scanning electron microscope. The results demonstrated that cryo-treated (CTI) insert had lower tool wear, vibration, and cutting force than untreated insert (UI) in all conditions. In aspect of chip morphology, untreated inserts had bigger and larger serrations than the treated inserts. In addition, according to Taguchi S/N ratio, optimal cutting parameters and heat treatment conditions were obtained from CHT1, V3, and f1 for the Fc and from CHT1, V1, and f1 for the Ra, respectively. Also, the most significant control factors on surface roughness and cutting force were feed rate depending on ANOVA results and RSM. Validation test results demonstrated that RSM and Taguchi mixed design calculated the cutting force (R²_{RSM (CTI and UI)} = 99.99% and R²_Tag. = 99.95%) and surface roughness (R²_RSM (CTI) = 99.76%, R²_RSM (UI) = 99.59% and R²_Tag. = 99.12%). Therefore, RSM and Taguchi mixed design predicts highly well match experimental data with prediction data.