Performance Prediction of Aluminum Oxide, Silicon Oxide, and Copper Oxide as Nanoadditives Across Conventional, Semisynthetic, and Synthetic Lubricating Oils Using ANN

Abstract

In the realm of lubrication, nanoparticles play a pivotal role in enhancing the tribological efficacy of lubricating oils. Unveiling a critical need, the research underscores the necessity for a predictive model capable of anticipating these performance characteristics. This research endeavors to fill this gap by introducing an artificial neural network (ANN) tailored specifically for predicting the behavior of nanolubricants. The optimized neural network structure, at 5 × 8 × 2, attains a remarkable minimum mean square error of 0.00046667, with R-values hovering at impressive proximity to unity (0.99828). During the confirmation phase, the neural network's predictions demonstrate a deviation of 7.51% (negative) and 2.87% (negative) for COF, alongside 0.50% and 1.80% for WSD, further affirming its predictive capacity in assessing lubricant performance characteristics.