Application Of A Neural Network Model In Estimation Of Frictional Features Of Tribofilms Derived From Multiple Lubricant Additives

This article discusses the use of machine learning in evaluating the relationship between friction coefficient and elemental distribution of tribofilms formed from multiple lubricant additives. The study proposes the use of convolutional neural networks (CNN) and gradient-weighted class activation mapping (Grad-CAM) for classification and visualization of these tribofilms. The results suggest that these methods are useful for evaluating frictional features and reducing carbon dioxide emissions in machinery.