Optimization of cryo-treated EDM variables using TOPSIS-based TLBO algorithm

Abstract

In order to machine hard and high-strength-to-weight ratio materials, electrical discharge machining (EDM) process is extensively used in aerospace, automobile and other industrial applications. However, high erosion of tool and improper selection of machining variables have emerged as a major obstruction to achieve productivity in this direction. High erosion of tool not only enhances the cost of machining but also increases the machining time by causing interruption during machining. Therefore, proper selection of machining variables and tool material life are the two vital aspects for the tool engineers working in EDM. In view of this, the present work proposes an extensive experimental investigation and optimization of machining variables of cryogenically treated brass tool materials on machining competences of Inconel 718 workpiece. The study primarily highlights the outcome of cryogenically treated soaking duration of tools along with other important process variables, viz. discharge current, open-circuit voltage, pulse-on time, duty factor and flushing pressure, on the performance measures such as electrode wear ratio (EWR), surface roughness and radial over-cut. The study revealed that soaking duration in deep cryo-treatment of the electrode is a significant variable to achieve improved machining characteristics. The performance measures are converted into equivalent single performance measure by calculating the relative closeness coefficient by the techniques for order preferences by similarity to ideal solution (TOPSIS) approach. Finally, a novel teaching–learning-based optimization (TLBO) algorithm has been proposed to find the optimal level of machining variables for the performance measures. The optimal levels of cutting variables obtained through the algorithm are validated through confirmation test, predicting an error of 2.171 percentages between the computational and experimental results. The predicted result suggests that the proposed model can be used to select the ideal process states to achieve productivity for the cryo-treated EDM.