Performance Prediction of Electric Discharge Machining of Inconel-718 using Artificial Neural Network. (2018)

• Record Type: Journal Article
• Authors: Vishnu, P.; Santhosh Kumar, N.; Manohar, M.
• Is Part Of: Materials today. Volume 5:Number 2(2018)Part 1
• Page Start: 3770
• Page End: 3780
• Publication Date: 2018
• Subjects: EDM -- Inconel-718 -- ANN -- optimisation; Materials science -- Congresses -- Periodicals; 620.1
• DOI: 10.1016/j.matpr.2017.11.630
• Languages: English
• ISSNs: 2214-7853

Abstract

Abstract: Inconel-718 is a nickel based alloy which displays very high yield, tensile and creep rupture properties at temperatures of up to 978 K. This alloy has extensive applications in jet engines, high speed air frame parts, power plant turbine components, automobile engine components and high temperature fasteners. Inconel-718 is classified under 'difficult to machine' material using traditional techniques. This paper deals with a non-traditional approach of machining Inconel-718 using Electric Discharge Machining (EDM). Experiments were designed and conducted according to Taguchi's L18 orthogonal array. Experiments were carried out under different cutting conditions of polarity, pulse on time, pulse off time and peak current. Electrolytic copper was used as tool electrodes. The researchers propose analytical models that simulate the machining conditions to understand the role of contributing factors and their interaction effects and establish cause and effect relationships among various factors and desired product quality requirements. An Artificial Neural Network (ANN) is a branch of Artificial Intelligence (AI) which represents a human brain that tries to simulate its learning process. To predict the performance characteristics namely Material Removal Rate (MRR), Surface Roughness (SR) and Tool Wear Rate (TWR), ANN models were developed using back-propagation algorithms. Sufficient level of fitness was observed for the trained model. A comparison was made between experimental response values and the predicted values. The prediction accuracies were found to be sufficiently high which indicates the effectiveness of the model.