Tribological Performance and Machine Learning-Assisted Prediction of Wear and Friction in D-Gun Sprayed WC-12Co and Al2O3-13TiO2 Coatings

Abstract

Thermal spray coatings are widely employed to enhance the wear resistance and frictional performance of engineering components operating under severe tribological conditions. In this study, WC-12Co and Al₂O₃-13TiO₂ coatings were deposited on AISI 6151 steel using the detonation gun (D-gun) spraying technique, and their wear and friction behavior were systematically investigated. A Taguchi L27 orthogonal array, combined with response surface methodology (RSM), was used to evaluate the effects of coating type, load, sliding speed, and test duration on the wear rate and coefficient of friction (CoF) under dry-sliding conditions. Microstructural and phase analyses were performed using SEM, EDS, and XRD to elucidate dominant wear mechanisms. In addition, machine learning (ML) models, including random forest (RF), decision tree (DT), K-nearest neighbors (KNN), and extreme gradient boosting (XGBoost), were developed to predict the wear rate and CoF from experimental data. Model performance was assessed using multiple error metrics and statistically validated through variance analysis and confidence intervals. The results indicate that WC-12Co exhibits superior wear resistance due to high hardness and low porosity, while Al₂O₃-13TiO₂ provides a balanced combination of moderate wear resistance and stable friction behavior. The ML models demonstrated high predictive accuracy within the investigated parameter space; however, given the structured and limited dataset, the models are primarily intended for interpolation rather than extrapolation beyond the experimental domain. This integrated experimental-statistical-ML framework provides a reliable decision-support tool for optimizing thermal spray coatings in practical engineering applications.