Discriminant analysis-based parametric study of an electrical discharge machining process

Abstract

This paper presents the application of discriminant analysis in an Fleet rical Discharge Machining (EDM) process to determine the comparative contribution of each of its input parameters on the measured responses. It also identifies the most significant RDM process parameters influencing those responses. For this process, voltage, current, pulse-on time, and pulse-off time are treated as the input parameters, whereas, material removal rate, electrode wear rate, and surface roughness are the responses. Based on the past and simulated experimental data, both simultaneous and step-wise estimations are carried out for each of the three responses showing the relationships between the FDM process parameters and the considered responses. It is observed that in both these estimations, pulse-off time, current and pulse-on time respectively evolve out as the most significant parameters for material removal rate, electrode wear rate, and surface roughness. Step-wise estimation identifies voltage as the least significant input parameter for all these responses. The developed discriminant functions, which can also help in predicting the responses, are finally cross-validated.