Abstract
In this paper the potential of using artificial neural networks (ANNs) for the prediction of sliding friction and wear properties of polymer composites was explored using a newly measured dataset of 124 independent pin-on-disk sliding wear tests of polyphenylene sulfide (PPS) matrix composites. The ANN prediction profiles for the characteristic tribological properties exhibited very good agreement with the measured results demonstrating that a well trained network had been created. The data from an independent validation test series indicated that the trained neural network possessed enough generalization capability to predict input data that were different from the original training dataset.
► Artificial neural network (ANN) was applied for modeling sliding friction and wear. ► Experimental data from polyphenylene sulfide composites were used to train the ANN. ► The trained ANN predicted the friction and wear properties with very good accuracy. ► The trained ANN correctly evaluated the impact of filler content on friction and wear.