Optimisation of spark erosion machining process parameters using hybrid grey relational analysis and artificial neural network model

Abstract

Hastealloy C276 is hard to machine superalloy and extensively used in various engineering applications. It possess good strength and lower thermal conductivity which results in decreased tool life and poor machinability by conventional machining. Advanced machining processes have developed to overcome these difficulties and claimed as an alternative methods. Electrical Discharge Machining (EDM) is one of the advanced method used for machining of hard materials. This article details an investigation on EDM process and development of hybrid Grey ANN model. Taguchi method and ANOVA are used for designing the experiments and statistical analysis respectively. Grey Relational Analysis is adopted for determining the Grey Relational Grade (GRG) to represent the multi aspect optimization model and a neural network has been evolved to predict GRG by feeding the Grey Relational Co-efficient (GRC) values as input to developed neural network model. A comparison has been done between the experimental values and predicted values.