Prediction on wear properties of polymer composites with artificial neural networks

An artificial neural network (ANN) technique is applied to predict the wear properties of polymer-matrix composites. Based on an experimental database for short fiber reinforced polyamide 4.6 composites, the specific wear rate, frictional coefficient and furthermore some mechanical properties, such as compressive strength and modulus, were successfully calculated by a well-trained ANN. 3-D plots for the predicted wear and mechanical characteristics as a function of material compositions and testing conditions were established. The results are in good agreement with measured data. It shows that the prediction accuracy is reasonable, and the network has potential to be improved if the experimental database for network training could be expanded.