Wear characterization of Al–SiC–graphite hybrid composite and its prediction using artificial neural network method

The purpose of this paper is to produce AW6061 alloy hybrid composite reinforced with graphite and silicon carbide particles and study their wear rate and also to develop ANN model to forecast the wear rate of hybrid composite. Hybrid composites were produced by using liquid-stir casting setup. Experiments were accompanied by using L 9 orthogonal array on pin-on-disc wear testing machine. Experimental data-set were used to train ANN model with two-hidden layer feedforward network by Levenberg-Marquardt algorithm and predicted the experimental values. Further, 3D response surface on wear rate, worn-out surface study and quantitative wear damage mapping were studied using SEM images. The results reveals that the mean percentage error for predicted ANN model is 1.64% and overall R value found to be 95.6%. Sliding distance and wear rate shows a high correlation of 0.85. Delaminated layer, peeled off graphite particle, undecorated plastic deformation was seen on the worn-out surface of hybrid composite. Computable wear surface characterization metrics are also discussed in detail.