Prediction of Dry Sliding Wear Response of AlMg1SiCu/Silicon Carbide/Molybdenum Disulphide Hybrid Composites Using Adaptive Neuro-Fuzzy Inference System (ANFIS) and Response Surface Methodology (RSM)

Abstract

In this research work, an effort was made to predict the dry sliding wear response of AlMg1SiCu alloy hybrid composites which were reinforced with 10% Silicon carbide particles (SiC) together with weight fractions of 3, 6 and 9% of self-lubricant molybdenum disulphide particles (MoS2) through melt stir casting. The wear behaviour of the hybrid composite samples was evaluated based on Box-Behnken design on pin-on-disc tribometer without lubrication. The output response weight loss was employed to train the neural network model in ANFIS back-propagation algorithm. The weight loss of 9% MoS2 hybrid composite reduced at low sliding speeds, due to the development of shallow sliding grooves and MoS2-lubricated tribolayer. Scanning electron micrographs and EDS of the AlMg1SiCu alloy hybrid composites revealed a uniform distribution of SiC and MoS2 particles. The tensile strength of the as-cast hybrid composites increases as the wt.% of MoS2 particles increases, according to the tests. However, the addition of MoS2 improved the hardness of the hybrid composites until it reached 6 wt.%, after which it decreased slightly. Weight loss and coefficient of friction decreased by addition of self-lubricant MoS2 in the matrix material. Worn-out surface of the hybrid composite shows the controlling wear mechanisms of the composites, and well-trained ANFIS model could accurately predict the responses better when compared with the response surface methodology model.