PREDICTION OF ABRASIVE WEAR BEHAVIOR OF CARBON FABRIC REINFORCED EPOXY COMPOSITES USING TAGUCHI ANALYSIS AND NEURAL NETWORK

An approach to the prediction of three body abrasive wear behavior of unfilled and graphite filled carbon fabric reinforced epoxy composite using two modeling techniques namely Taguchi analysis and artificial neural network are presented in this study. A plan of experiments using orthogonal array based on techniques of Taguchi was arrived at and carried out to acquire data in a controlled way. From the experimental results, it is noted that, addition of graphite particulate into carbon epoxy composite deteriorated the abrasive wear resistance of the composites, also wear loss increased with the increase in abrading distance and loads. Analysis of variance was performed on the acquired data, and S/N ratio was established to investigate the influence of control parameters on the wear behavior of these composites. Among the control parameters, normal load showed the highest physical as well as statistical influence on the abrasive wear of the composites (77%) followed by abrading distance (18.2%) and filler content (1.3%). The 3-[5]1-1 neural network architecture with Levenberg Marquardt (LM) training algorithm was used to predict the wear properties of composites as a function of testing conditions. The correlations were obtained by Taguchi regression analysis and artificial neural network and compared with the experimental results. The results indicated that artificial neural network predicts the wear rate better than regression analysis. Hence, a well-trained artificial neural network system is expected to be very helpful for estimating the weight loss in the complex three-body abrasive wear situation of polymer composites.