Artificial neural network prediction of the wear rate of powder metallurgy Al/Al_2O_3 metal matrix composites

摘要

In this study, the artificial neural network (ANN) approach is used to predict the wear rate of A1/Al_2O_3 metal matrix composites (MMCs). The Al/Al_2O_3 MMCs were fabricated using the conventional powder metallurgy route. Different ANN structures were used to develop models for prediction and optimization of the wear rates of the A1/Al_2O_3 composites under dry sliding conditions. The effects of the volume fraction of Al_2O_3 particulates, density, and hardness of the Al/ Al_2O_3 composites, as well as the applied pressure, sliding speed, and test temperature on the wear rates were evaluated using the ANN models. The results showed that the developed multilayer perceptron ANN model can be effectively used to predict the wear rates of Al_/Al_2O_3 MMCs. The generalized radial basis function showed lower prediction performance.