Optimization of hybrid aluminum composites wear using Taguchi method and artificial neural network

Abstract

PurposeThis research aims to describe the influence of weight per cent of graphite (Gr), applied load and sliding speed on the wear behavior of aluminum (Al) alloy A356 reinforced with silicon carbide (SiC) (10 Wt.%) and Gr (1 Wt.% and 5 Wt.%) particles. The objective is to analyze the effect of the aforementioned parameters on a specific wear rate.Design/methodology/approachThese hybrid composites are obtained by means of the compo-casting process. Tribological analyses were conducted on block-on-disc tribometer at three different loads (10, 20 and 30 N) and three different sliding speeds (0.25, 0.5 and 1 m/s), at the sliding distance of 900 m, in dry sliding wear conditions. Optimization of the tribological behavior was conducted via the Taguchi method, and ANOVA was used for the analysis of the specific wear rate. Confirmation tests are used to foresee and check the experimental results. Examined samples were analyzed via a scanning electron microscope (SEM). Regression models for predicting specific wear rate were developed with Taguchi and ANN (artificial neural network) methods.FindingsThe biggest impact on value of specific wear rate has the load (43.006%), while the impact of Wt.% Gr (31.514%) was less. After comparison of the results, i.e. regression models, for predicting the specific wear rate, it was observed that ANN was more efficient than the Taguchi method. The specific wear rate of Al alloy A356 with SiC (10 Wt.%) and Gr (1 Wt.% and 5 Wt.%) decreases with a decrease in the load and weight per cent of Gr-reinforcing material, as well as with a decrease in sliding speed.Originality/valueThe results obtained in this paper using the Taguchi method and the ANN method are useful for improving and further investigating the wear behavior of the SiC- and Gr-reinforced Al alloy A356.