Development of forecast models on electrical discharge machined graphene nanoplatelets reinforced aluminum composite fabricated via stir casting route

Abstract

AbstractHigh specific strength and good fatigue limit are the key properties to watch out during development of aerospace components. Aluminum composites are proven high specific strength materials especially graphene embedded composites worth a mention in this context. While, surface finish is an eminent parameter affecting fatigue strength of a component showcasing the importance of machining technique employed to transform a fabricated bulk into finished product. Current study, therefore emphasizes on electrical discharge machining (EDM) of aluminum composites embedded with graphene nanoplatelets (1.5 wt.%) fabricated through a hybrid approach of blending solid and liquid metallurgical routes. Further, mathematical and neurological forecast models are developed to individually predict the machining response variables namely surface roughness (SR), material removal rate (MRR) and tool wear rate (TWR). Among machining parameters current (I), pulse on-time (Ton), pulse off-time (Toff) and flushing pressure (P) considered; Ton is noted to greatly influence the surface quality of composite while TWR and MRR are affected by current during ANOVA analysis. On comparative understanding of forecast models, neurological models outperform quadratic non-linear mathematical models where accuracy of prediction achieved by developed artificial neural network (ANN) model is 96% for surface roughness. The error performance plots, error histograms and overall fit plots depict a marginal over-fit neurological model. However, significantly high coefficient of correlation (R) of 99% possessed by ANN model illustrates their potential in forecasting response parameters.

Keywords

• Aluminum nanocomposite
• artificial neural networks
• electrical discharge machining
• graphene nanoplatelets
• material removal rate
• surface roughness