Rice husk ash reinforced aluminium matrix composites: fabrication, characterization, statistical analysis and artificial neural network modelling

This paper explores the microstructural, mechanical and the tribological behaviour of rice husk ash (RHA, 5, 10 and 15 wt%) reinforced aluminium based composites fabricated using the powder metallurgy (PM) route. The main advantage of this composite is utilization of RHA (an agricultural waste), with its improved mechanical and wear properties. Powder mixtures are cold pressed uniaxially and later the green compacts are sintered under argon gas atmosphere in electric furnace. For the investigation of microstructural features, Scanning Electron Microscopy (SEM) and x-ray diffraction (XRD) analysis has been performed. Tribological behaviour was evaluated on pin-on-disc wear tester machine using Taguchi and ANOVA techniques. Addition of RHA increased the composite's hardness by 20%-25% and wear behaviour got improved by 15%-40%. Based on the micrographic images of worn out surfaces and wear debris, wear mechanism is also discussed. In addition to this, artificial neural network model is also proposed and wear behaviour of the composite is also predicted. By comparing the experimental results with predicted results, it can be said that a well-trained ANN model is an efficient tool for predicting tribological behaviour.