Multi-objective optimization of the dry electric discharge machining process

TL;DR

This paper presents a multi-objective optimization of dry electric discharge machining using NSGA II, modeling process parameters to optimize material removal rate and surface roughness, and identifying process regions for finish and rough machining.

Contribution

It introduces a novel application of NSGA II for dry EDM, developing polynomial models and region-specific process insights for improved optimization.

Findings

Optimal finishing conditions: low current, high pulse-on time, low duty factor.

Optimal roughing conditions: high current, low pulse-on time, high duty factor.

Identified process constraints related to flushing efficiency.

Abstract

Dry Electric Discharge Machining (EDM) is an environment-friendly modification of the conventional EDM process, which is obtained by replacing the liquid dielectric by a gaseous medium. In this study, multi-objective optimization of dry EDM process has been done using the non dominated sorting genetic algorithm (NSGA II), with material removal rate (MRR) and surface roughness (Ra) as the objective functions. Experiments were conducted with air as dielectric to develop polynomial models of MRR and Ra in terms of the six input parameters: gap voltage, discharge current, pulse-on time, duty factor, air pressure and spindle speed. A Pareto-optimal front was then obtained using NSGA II. Analysis of the front was done to identify separate regions for finish and rough machining. Designed experiments were then conducted in these focused regions to verify the optimization results and to identify the region-specific characteristics of the process. Finishing conditions were obtained at low current, high pulse-on time and low duty factor, where as rough conditions were obtained at high current, low pulse-on time and high duty factor. Focused experiments revealed an additional process constraint based on the flushing efficiency.