OBRABOTKAMETALLOV Vol. 28 No. 2 2026 332 MATERIAL SCIENCE References 1. Almomani O., Venkatesh B., Chaudhary S.P., Mishra A., Sujai S., Juneja S., Pradhan P., Venkatesan S.P., Bhowmik A., Tamene Y. Intelligent tool wear monitoring using XGBoost, SVR, and DNN models in NMQL environment. Scientifi c Reports, 2026, vol. 16 (1), p. 10030. DOI: 10.1038/s41598-026-40968-8. 2. Cherguy O., Chmielowski R., Hachem E., Lahouij I. Deep learning prediction of dry friction in DLC coatings using literature-derived data. Tribology Letters, 2025, vol. 73 (4), p. 125. DOI: 10.1007/s11249-025-02056-2. 3. Daghbouch A., Louhichi B., Terres M.A. Optimization and prediction of mass loss during adhesive wear of nitrided AISI 4140 steel parts. Crystals, 2025, vol. 15 (10), p. 875. DOI: 10.3390/cryst15100875. 4. Bergmann B., Denkena B., Junge N., Kalscheuer C., Bobzin K., Liu X. Infl uence of the interlayer and the substrate on the wear behavior of PVD tool coatings during turning of C60+ N. Wear, 2026, vol. 584–585, p. 206396. DOI: 10.1016/j.wear.2025.206396. Wear rate prediction of PVD hard coatings using a hybrid Taguchi–RSM–machine learning framework with SHAP-based interpretability Umesh Subhash Patharkar 1, a, Sunil Apparao Patil 1, b, Nitin Ambhore 2, c, * 1 Government Engineering College, Chh. Sambhajinagar, 431005, Maharashtra, India 2 Vishwakarma Institute of Technology, SPPU, Pune 411037, Maharashtra, India a https://orcid.org/0009-0005-9303-641X, patharkarumeshsubhash@gmail.com; b https://orcid.org/0009-0000-4804-0720, sakpatil967@gmail.com; c https://orcid.org/0000-0001-8468-8057, nitin.ambhore@vit.edu Obrabotka metallov - Metal Working and Material Science Journal homepage: http://journals.nstu.ru/obrabotka_metallov Obrabotka metallov (tekhnologiya, oborudovanie, instrumenty) = Metal Working and Material Science. 2026 vol. 28 no. 2 pp. 318–334 ISSN: 1994-6309 (print) / 2541-819X (online) DOI: 10.17212/1994-6309-2026-28.2-318-334 ART I CLE I NFO Article history: Received: 06 April 2026 Revised: 11 April 2026 Accepted: 30 April 2026 Available online: 15 June 2026 Keywords: PVD coatings Wear Response surface methodology Random Forest XGBoost SHAP analysis ABSTRACT Introduction. The wear behaviour of physical vapour deposition (PVD) hard coatings is diffi cult to predict because of the nonlinear relationship between coating chemistry, deposition thickness, operating temperature, and applied contact load. Multi-dimensional parameter mapping through exhaustive experimentation is both timeconsuming and costly. The purpose of the work is to implement a low-data hybrid architecture that combines Taguchi design, response surface methodology (RSM), and machine learning (ML) to predict and optimize wear performance based on a small dataset (n = 28). Methods. Three coating types – AlTiN, CrN, and TiC – were tested at temperature range of 40–50°C, contact load range of 5–15 N, and coating thickness range of 2–4 μm. Analysis of variance (ANOVA) was performed to identify the most infl uential parameter aff ecting wear. A predictive wear model was developed using response surface methodology. Results and discussion. ANOVA revealed that load and coating chemistry are the most infl uential factors aff ecting wear. Temperature and thickness were not found to be signifi cant within the studied range. A RSM model was found statistically signifi cant for predicting contact load and coating chemistry with R2 = 0.901 (Adj-R2 = 0.834, RMSE = 0.0076). Random forest (RF) had highest generalisation performance (CV R2 = 0.755) and gradient boosting (GB) had the best foverall fi t (R2 = 0.913) among the ML models tested using fi ve-fold cross-validation. SHAP analysis indicated that coating chemistry formed the major contribution, then contact load, and little temperature contribution. Gradient boosting optimisation indicated that AlTiN at 4 μm thickness, 15 N load and 50 °C are the preferred settings and the expected wear rate is 0.010823 (mm3/Nm). The proposed framework demonstrates credible, interpretable wear forecasting using limitted experimental data. For citation: Patharkar U., Patil S., Ambhore N. Wear rate prediction of PVD hard coatings using a hybrid Taguchi–RSM–machine learning framework with SHAP-based interpretability. Obrabotka metallov (tekhnologiya, oborudovanie, instrumenty) = Metal Working and Material Science, 2026, vol. 28, no. 2, pp. 318–334. DOI: 10.17212/1994-6309-2026-28.2-318-334. (In Russian). ______ * Corresponding author Nitin Ambhore, Ph.D. (Engineering), Associate Professor Vishwakarma Institute of Technology, Pune - 411037, Maharashtra, India Tel.: +91-2026950441, e-mail: nitin.ambhore@vit.edu
RkJQdWJsaXNoZXIy MTk0ODM1