Analysis of the effect of reinforcement particles on the compressibility of Al–SiC composite powders using a neural network model

A neural network (ANN) model was developed to predict the densification of composite powders in a rigid die under uniaxial compaction. Al–SiC powder mixtures with various reinforcement volume fractions (0–30%) and particle sizes (50 nm to 40 μm) were prepared and their compressibility was studied in a wide range of compaction pressure up to 400 MPa. The experimental results were used to train a back propagation (BP) learning algorithm with two hidden layers. A sigmoid transfer function was developed and found to be suitable for analyzing the compressibility of composite powders with the least error. The trained model was used to study the effect of reinforcement particle size and volume fraction on the densification of Al–SiC composite powders. The outcomes of the ANN model are analyzed based on the mechanisms of densification, i.e., particle rearrangement and plastic deformation. The proper condition of compaction for achieving the highest density by tailoring the reinforcement particle size and volume fraction dependent on the compacting pressure is presented.