Application of artificial neural network in predicting the wear rate of copper surface composites produced using friction stir processing

The application of artificial neural network (ANN) to predict the wear rate of the surface composites produced using a solid-state technique called friction stir processing (FSP) is presented in this work. The copper surface composites were prepared by incorporating different sort of ceramic particles such as SiC, TiC, Al₂O₃, WC and B₄C. The design of experiments (DOE) strategy was utilised to direct the experimental work. The considered operating parameters were sort of ceramic particle, traverse speed, tool rotational speed and groove width, whereas wear rate is the response. An approximation mechanism having an arbitrary function, the ANN was consequently used for simulating the wear rate of the surface composites. The feedforward back propagation technique was employed to alter the weights of the network to minimise the mean squared error for the development of ANN models. The predicted trends were explained and studied the influence of the considered factors with the aid of observed micro structures. The lower wear rate was observed with B₄C-reinforced surface composites.