Modelling the effect of particle size and iron content on forming of Al–Fe composite preforms using neural network

Abstract Cold upsetting experiments were carried out on sintered Al–Fe composite preforms inorder to model and analyse the formability by simulation using neural network (NN). A model has been developed with a radial basis NN algorithm. The data used were collected by the experimental set up in the laboratory for the sintered Al–Fe composites with the various preform densities, the particle sizes and the aspect ratios. The network is trained to predict the forming characteristics such as the axial stress, the hoop stress, the hydrostatic stress and the Poisson ratio. In addition to that, the value of strain hardening coefficients such as instantaneous strength coefficient ( k i ) and instantaneous strain hardening exponent ( n i ) is also simulated to find the effect of particle size and the percent of iron content on formability. Regression analysis has confirmed a good agreement between the predicted and the experimental data with least error and hence this approach helps to facilitate a knowledge base in order to generate advice for the designer at the earlier stages of design so that concurrent engineering practices can be made.