Neural network model for predicting strain hardening and densification constants of sintered aluminium preforms

Abstract

A neural network model has been developed for the prediction of strain hardening and densification constants of sintered aluminium preforms. The model is based on a three layer neural network with a back propagation learning algorithm. The training data were collected by the experimental setup in the laboratory for sintered aluminium and with various preform densities with different aspect ratios by using MoS₂ as a lubricant. The network is trained to predict the values of strain hardening exponent index n _i, strength coefficient k _i, density power law exponent B _i and density constant C _i. Regression analysis between experimental and values predicted by the neural network shows the least error. This approach helps in the reduction of the experimentation required to determine these constants.