Optimization and statistical modeling of the thermal conductivity of a pumice powder and carbonated coal particle hybrid reinforced aluminum metal matrix composite for brake disc application: a Taguchi approach

Abstract

Abstract Aluminum metal matrix composites have been gaining traction in recent years due to their good mechanical properties and low weight. Particulate reinforcements for the improvement of its properties have been explored. This research aimed to determine the optimal composition of the reinforcement content (pumice powder and carbonated coal particles) and processing parameters (stirring speed, processing temperature, and stirring time) on the thermal conductivity of the developed material and also to characterize the constituents using x-ray fluorescence, x-ray diffraction, and scanning electron microscope/energy dispersive x-ray. The Taguchi optimization approach and regression analysis were used for the optimization and statistical analysis, respectively. The Taguchi optimization results gave an optimum thermal conductivity of 111.5, 112.5, 111.7, 112.9, and 112.4 W m−1 °C for pumice, carbonated coal, stirring speed, processing temperature, and stirring time respectively. The optimization also revealed the optimum setting for reinforcements and stir casting process factors as regards thermal conductivity to be 2.5%, 5.0%, 300 rpm, 850 °C, and 5 min for pumice powder, carbonated coal particles, stirring speed, temperature, and time, respectively. The optimal thermal conductivity of 120.40 W m−1 °C was obtained for the hybrid composite which gives a 131.54% improvement over the conventional grey cast iron brake disc. The particulate reinforcements (pumice powder and carbonated coal particles) and the processing factors all had significant effects on the thermal conductivity of the material, with the carbonated coal particles having the highest percentage contribution of 16.51%, as established by the analysis of variance. A model for predicting the thermal conductivity was developed using regression analysis, and high prediction accuracy was established with R-Square, R-Square (adj), and R-Square (pred) values of 94.68%, 88.60%, and 79.94%, respectively. The results of the characterization show the presence of hard compounds such as silica, iron oxide, and alumina in pumice powder and carbonated coal particles.