Optimisation of Turning Parameters for Inconel 718 Superalloy Using Taguchi–Grey Relational Analysis and Response Surface Methodology Under Dry, Flood, and Minimum Quantity Lubrication Conditions

Inconel 718, a precipitation-hardened nickel-based superalloy, is extensively employed in gas turbine discs, combustion chambers, aerospace fasteners, and nuclear reactor components owing to its exceptional retention of mechanical properties at elevated temperatures (up to 700°C), outstanding fatigue resistance, and superior corrosion behaviour. These attributes, however, impose severe machinability challenges: high cutting zone temperatures (often exceeding 600°C at practical cutting speeds), rapid tool wear through thermally activated diffusion and adhesion mechanisms, built-up edge (BUE) formation, and pronounced work-hardening of the machined surface that accelerates subsequent tool wear in adjacent cuts. The selection of optimal machining parameters — cutting speed (Vc), feed rate (f), and depth of cut (ap) — and an appropriate cooling-lubrication strategy are therefore critical determinants of process economy and surface quality in Inconel 718 turning. This study presents a comprehensive experimental investigation of Inconel 718 turning using TiAlN-PVD-coated carbide inserts (ISO CNMG 120408) under three lubrication conditions: dry, flood coolant, and Minimum Quantity Lubrication (MQL, 50 mL/hr soybean-based ester oil at 6 bar nozzle pressure). A Taguchi L27 orthogonal array was employed with three levels each of Vc (80, 120, 160 m/min), f (0.05, 0.15, 0.25 mm/rev), and ap (0.5, 1.0, 1.5 mm). Response variables include cutting force Fc, surface roughness Ra, flank wear VBmax, and cutting zone temperature T measured by embedded K-type thermocouple. Grey Relational Analysis (GRA) converts the multi-response optimisation problem to a single grey relational grade. Response Surface Methodology (RSM) using a Box-Behnken design validates the optimal parameter settings and develops predictive regression models for Ra and VBmax. Results demonstrate that MQL reduces Ra by 18.4% and VBmax by 22.1% relative to dry cutting at optimal conditions (Vc = 120 m/min, f = 0.10 mm/rev, ap = 0.5 mm). The Taguchi-GRA optimal combination (Vc = 120 m/min, f = 0.05 mm/rev, ap = 0.5 mm, MQL) simultaneously minimises all four responses, with ANOVA confirming feed rate as the dominant factor for Ra (P-contribution 58.4%) and cutting speed as dominant for VBmax (P-contribution 46.8%).