Enhancing EDM performance of Inconel 825 alloy using the rotary EDM process

Abstract

Electrical Discharge Machining (EDM) is a pivotal non-traditional machining technique used for processing hard, electrically conductive materials. Rotary Tool EDM is a novel improvement that incorporates electrode rotation for enhanced performance and efficiency. Since Inconel 825 alloys have such excellent thermo-mechanical characteristics, they have been widely applicable in the aerospace, electronics, and other industrial fields. Inconel 825 is exceedingly challenging to machine using the traditional drilling process. However, non-traditional machining techniques like EDM can be used to machining these alloys to the required level of precision. A detailed experimental investigation has been performed with the copper tool by RT-EDM setup on the Inconel 825 specimen. Parametric analysis has been done by varying the tool rotation speed (TRS), input current (Ip), duty factor (Tau), pulse on time (Ton), and voltage (V) through experiment based on response surface methodology (RSM) with central composite design (CCD). The impact of tool rotation and other parameters has been explored on material removal rate (MRR), tool wear rate (TWR), and surface roughness (SR). To model and examine the impacts of different RTEDM parameters on the machining performance, the artificial neuro-fuzzy inference system (ANFIS) and RSM were utilized to describe the intricate relationships between factors and predicted machining performance. When tool rotation is increased from 0 to 900 rpm, the MRR rises by around 39.53% and the SR falls by about 40.02%. Comparing machining at 900 rpm to a stationary tool, MRR increases quickly by 40.04% at 3A, significantly by 37.04% at 6A, 30.63% at 9A, and nominally by 31.35% at 12A. With average percentage errors of 1.6%, 2.4%, 4.59%, 3.42%, and 3.89%, respectively, the ANFIS model predictions closely match the experimental trial data and correctly predicted all five responses. The experimental findings showed a significant connection with observed data, validating the RSM and ANFIS models’ prediction accuracy.