Experimental investigation and optimization of machining performance characteristics during WEDM of inconel 718: On evaluation of wire electrodes and advanced parameter techniques

Abstract

Wire Electrical-Discharge-Machining (WEDM) is a well-known unconventional machining process to produce intricate shapes. However, obtaining a satisfactory WEDM cutting performance is indeed a challenging task during precision cutting. Hence, this investigation aims to attempt a favorable machining parameter setting in order to corner-cutting during WEDM for In-718. Here, machining performance characteristics have been considered based on corner deviation (CD) along with Material Removal Rate (MRR) and surface roughness (SR). Taguchi’s experiment design technique (L 16 ) has been considered to run the experiments. The controllable process parameters are considered as Spark-on-time (S on ), flushing-pressure (F p ), wire-tension (T w ), and discharge-current (I d ). The aforesaid machining performance characteristics have been achieved through the two most popular wire electrodes, i.e., Zinc-coated brass electrode (Zn-BE) and Brass Wire Electrode (BWE), and compared the results. The comparison of performances by the wire electrodes on CD, MRR, and SR varied from 0.0286 mm to 0.0844 mm, 0.0045 g/min to 0.0214 g/min and 3.12 µm to 4.80 µm for BWE and 0.0218 mm to 0.0783 mm, 0.0090 g/min to 0.0342 g/min and 2.58 µm to 4.40 µm for Zn-BE respectively. However, machining with Zn-WE yields reduced CD, SR, and increased MRR value and shows less defect on the WEDMed surfaces than its counterpart. The present study developed the mathematical model based on non-linear regression for correlating the machining parameters with the machining responses. The next step of this study is that a unique optimization strategy, namely grey relation analysis (GRA) integrated with Teaching Learning-Based Optimization (TLBO), has been implemented for achieving optimal parametric setting. The satisfactory machining setting obtained from GRA-TLBO has been compared with GRA-JAYA and GRA-genetic algorithm (GA). The proposed methodology appears more fruitful in terms of computational time and effort.