Optimization of hybrid agro-waste reinforced aluminium composites using response surface methodology and artificial neural networks

Aluminium matrix composites (AMC) reinforced with agro-waste fillers, as promising materials, have been increasingly demanded due to their sustainability and economic requirements. This research examines the processing, characterization, and optimization of AMC produced through stir casting, utilising a novel combination of agro-waste fillers including coconut shell ash (CSA) and eggshell powder (ESP). The composites are evaluated for mechanical properties and microstructural integrity through Scanning Electron Microscopy (SEM). The process parameters (filler content, stirring speed and duration) are varied and optimized using Response Surface Methodology (RSM). A model utilizing an Artificial Neural Network (ANN) is developed to predict composites’ properties accurately. The developed composite revealed 58.21% higher hardness and notable increments in tensile (41.93%), compressive (10.53%), and impact strength (22.08%) over the base alloy. The optimized process parameters (CSA = 5 wt.%, ESP = 3.5 wt.%, stirring duration =12 min and speed = 650 rpm) derived from RSM markedly improved mechanical properties and the ANN model exhibited a good agreement with experimental values by a mean absolute error of around 5%. Findings suggest the potential for agro-waste-reinforced AMC for low-weight structural and automotive applications and provide a path to sustainable material development and waste utilization. This research presents a new and effective framework for the processing and optimization of AMC incorporating agro-waste fillers.