Machine Learning and Statistical Approach to Predict and Analyze Wear Rates in Copper Surface Composites

Abstract

This research demonstrates the application of machine learning models and statistics methods in predicting and analyzing dry sliding wear rates on novel copper-based surface composites. Boron nitride particles of varying fractions was deposited experimentally over the copper surface through friction stir processing. Experimental and statistical analysis proved that the presence of BN particles can reduce wear rate considerably. Analysis of worn-out surface revealed a mild adhesive wear during low load condition and an abrasive mode of wear during higher load conditions. Artificial neural network based feed forward back propagation model with topology 4-7-1 was modeled and prediction profiles displayed good agreement with experimental outcomes.