Applied Chemical Engineering

Predictive Modelling of Steel Corrosion: A Comparative Study of Statistical and Machine Learning Methods with a Focus on Copper Behavior

Mohanad S. Hasan

College of Engineering, University of Diyala, Diyala, Iraq

Amjad M. Bader

College of Engineering, University of Diyala, Diyala, Iraq

Saad Sh. Sammen

College of Engineering, University of Diyala, Diyala, Iraq

DOI: https://doi.org/10.59429/ace.v9i3.5955

Keywords: Atmospheric corrosion; Low-alloy steel; Copper threshold effect; Machine learning; XGBoost; Predictive modelling; Corrosion rate prediction; Environmental exposure

---

Abstract

Accurately predicting corrosion rates in low-alloy steels is a significant challengein materials engineering due to the intricate and nonlinear interaction between environmental exposure conditions and alloying elements. Conventional statistical corrosion models are by and large based on linear assumptions, and thereby most of them fail to capture such interactions, especially the dual and threshold-controlled behavior of Cu in forming protective rust layers. This study presents a comprehensive comparative analysis of statistical and machine learning approaches for the prediction of atmospheric corrosion rates of low-alloy steels. Several predictive models were developed and validated for this task, namely, Decision Trees, Support Vector Machines with several kernel functions, and the gradient-boosting algorithm XGBoost, by using experimentally derived environmental and compositional datasets. Model performance is assessed based on the coefficient of determination (R²), root-mean-square error (RMSE), and mean absolute error (MAE).
The results demonstrate that non-linear models outperform conventional linear models of non-linear models in comparison to the conventional linear models. Among various models, the best predictive model was XGBoost with an R² value of about 0.88, validating the best ability to exploit threshold values and interactive coefficients. The conclusion is drawn from the result that copper is only effective in imparting corrosion resistance in a critical range, beyond which the protective property becomes saturated or decreases, an imperative that cannot be accounted for in the linear regression models. The obtained results strongly suggest that the corrosion process in low-alloy steels is governed by nonlinear and threshold-dependent behavior. Therefore, the paper establishes the application of gradient - boost-based machine learning algorithms as a strong and authentic method for predicting the development of corrosion models in low-alloy steels. Furthermore, the comparative assessment also emphasizes the importance of kernel-based learning approaches, such as the radial basis function kernel-based SVM, where the models also forecast well with slightly lower accuracy compared to the gradient-booster models. Additionally, the better performance of the non-linear models clearly depicts the model flexibility to cope with the Abrupt transitions of the corrosion process influenced by varying concentrations of alloys. From the scientific aspect, the obtained results demonstrating the threshold effects of the copper concentrations clearly support the formation and saturation of inner Cu-enriched rust layers. Therefore, this analysis attempt to connect the machine learning algorithm results with the scientific concepts of the corrosion process and hence establish the importance of machine learning models as a supplementary aspect to the other scientific methods used in the fields of materials science.

---

References

[1]. J. Tidblad, K. Kreislová, M. Faller, D. De la Fuente, T. Yates, A. Verney-Carron, T. Grøntoft, A. Gordon, and U. Hans, “ICP Materials Trends in Corrosion, Soiling and Air Pollution (1987–2014),” Materials, vol. 10, no. 8, p. 969, 2017.

[2]. H. Y. Chang and Y. S. Wei, "Comparison of machine learning and statistical models for atmospheric corrosion prediction of carbon steel," Materials, vol. 18, no. 4, pp. 888, 2025.

[3]. T. Yamashita, H. Miyuki, Y. Matsuda, H. Nagano, and T. Misawa, "The long-term growth of the protective rust layer formed on weathering steel by atmospheric corrosion during a quarter of a century," Corrosion Science, vol. 36, no. 2, pp. 283–299, 1994.

[4]. Q. Wang, C. Han, and S. Wang, "Insight into the Influence of Cu on the Corrosion Mechanism of 1%Ni Weathering Steel in Simulated Marine Atmosphere," Metals, vol. 14, no. 12, p. 1401, Dec. 2024.

[5]. Y. Wang, X. Zhang, and H. Liu, “Prediction of atmospheric corrosion behavior of low-alloy steels using machine learning approaches,” Corrosion Science, vol. 195, p. 109977, 2022.

[6]. Z. Y. Chen, S. Zakipour, D. Persson, and C. Leygraf, "Effect of copper and chromium on the formation of protective rust layers on weathering steels," Corrosion Science, vol. 47, no. 9, pp. 2268-2278, 2005.

[7]. Z. Pei, et al., "Data Mining to Atmospheric Corrosion Process Based on Evidence Fusion," Materials, vol. 14, no. 21, p. 6754, 2021.

[8]. Y. Li, et al., "Machine Learning-Based Prediction of Atmospheric Corrosion Rate for Low-Alloy Steel using Gradient Boosting," Coatings, vol. 15, no. 8, p. 888, 2025.

[9]. Y. W. Jang, J. H. Hong, and J. G. Kim, "Effects of copper on the corrosion properties of low-alloy steel in an acid-chloride environment," Corrosion Science, vol. 52, no. 11, pp. 3768–3775, 2010.

[10]. J. Tidblad, A. A. Mikhailov, and V. Kucera, “Dose–response functions for the corrosion of low-alloy steels in atmospheric environments,” Corrosion Science, vol. 170, p. 108687, 2020.

[11]. C. Leygraf and T. Graedel, “Atmospheric corrosion: Progress and challenges in modeling and prediction,” npj Materials Degradation, vol. 4, no. 1, pp. 1–10, 2020.

[12]. Z. Pei, J. Zhang, and Y. Liu, “Data-driven modeling of atmospheric corrosion using ensemble learning methods,” Corrosion Science, vol. 189, p. 109604, 2021.

[13]. Y. Li, H. Zhang, and J. Wang, “Prediction of corrosion rate of weathering steel using machine learning approaches,” Construction and Building Materials, vol. 272, p. 121942, 2021.

[14]. M. Melchers and R. Jeffrey, “Probabilistic modeling of long-term atmospheric corrosion processes,” Corrosion Science, vol. 185, p. 109456, 2021.

[15]. S. Zakipour, D. Persson, and C. Leygraf, “Role of alloying elements in rust layer formation on weathering steels exposed to marine atmospheres,” Corrosion Science, vol. 191, p. 109725, 2021.

[16]. A. K. Singh, S. K. Sahu, and P. Banerjee, “Application of machine learning techniques for corrosion rate prediction: A review,” Materials Today: Proceedings, vol. 46, pp. 10218–10225, 2021.

[17]. Y. Zhang, X. Liu, and Q. Wang, “Nonlinear prediction of atmospheric corrosion of low-alloy steel using support vector regression,” Materials, vol. 14, no. 7, p. 1794, 2021.

[18]. L. Guo, Y. Zhao, and J. Chen, “Effect of copper content on corrosion resistance of weathering steel under chloride-rich environments,” Materials Chemistry and Physics, vol. 267, p. 124645, 2021.

[19]. H. Wang, Z. Liu, and M. Zhu, “Machine learning-based evaluation of corrosion behavior in complex atmospheric environments,” Coatings, vol. 12, no. 3, p. 415, 2022.

[20]. S. Zhang, X. Ma, and Y. Sun, “Threshold effects of alloying elements on atmospheric corrosion of weathering steels,” Corrosion Science, vol. 198, p. 110100, 2022.

[21]. P. R. Roberge and R. Klassen, “Modeling atmospheric corrosion using hybrid physical–data-driven approaches,” Corrosion Engineering, Science and Technology, vol. 57, no. 6, pp. 529–540, 2022.

[22]. M. Kermani and D. Harrop, “The role of copper in corrosion protection of low-alloy steels: A modern perspective,” Materials Performance, vol. 61, no. 9, pp. 34–41, 2022.

[23]. Y. Liu, J. Zhou, and H. Chen, “Gradient boosting-based prediction of corrosion depth in steel structures,” Structural Health Monitoring, vol. 22, no. 1, pp. 512–525, 2023.

[24]. F. Zhang, Q. Han, and S. Wang, “Explainable machine learning for corrosion rate prediction of steels,” npj Materials Degradation, vol. 7, no. 1, pp. 1–11, 2023.

[25]. R. Kumar, A. Sharma, and P. Singh, “Comparative assessment of linear and nonlinear models for corrosion prediction,” Engineering Failure Analysis, vol. 146, p. 107084, 2023.

[26]. H. Chen, Y. Wang, and J. Li, “Atmospheric corrosion modeling using XGBoost and SHAP interpretation,” Corrosion Science, vol. 216, p. 111023, 2024.

[27]. Q. Wang, C. Han, and Y. Liu, “Synergistic effects of Cu and Ni on corrosion resistance of weathering steel in marine atmospheres,” Metals, vol. 14, no. 3, p. 312, 2024.

[28]. S. Ahmad, M. A. Khan, and T. Ali, “Data-driven prediction of corrosion rates using ensemble machine learning algorithms,” Coatings, vol. 14, no. 5, p. 612, 2024.

[29]. Y. Li, Z. Chen, and C. Leygraf, “Machine learning-assisted corrosion modeling of weathering steels under real atmospheric exposure,” Corrosion Science, vol. 225, p. 111312, 2025.

[30]. Y. Bai, B. Liu, Y. Zhang, S. Zhang, P. Wu, X. Ren, and Y. Wei, “Accurate Long-Term Atmospheric Corrosion Prediction of Weathering Steel via SHapley Additive exPlanations (SHAP) Enhanced Neural Network,” Engineered Science, vol. 39, p. 1921, 2025.

[31]. P. Zhang, C. Wang, W. Liu, and H. Su, “A hybrid deep learning model driven by physical mechanisms and data for predicting corrosion in natural gas pipelines,” Journal of Loss Prevention in the Process Industries, p. 105852, 2025.

[32]. L. Cai, A. Kim, and S. Kainuma, “Surface texture evolution of weathering steel in corrosive marine atmospheres: Geospatial variability and explainable machine learning models,” Corrosion Science, vol. 243, p. 112602, 2025.

[33]. R. Li, H. Wang, Y. Zhu, X. Mu, and X. Wang, “Machine learning–assisted in situ corrosion monitoring: a review,” Corrosion Reviews, vol. 44, no. 1, p. 20250014, 2026.

[34]. A. Karkadakattil, “Physics-informed machine learning framework for grain size prediction in laser-processed 316L stainless steel: integration of experimental microstructure and surrogate data,” Canadian Metallurgical Quarterly, pp. 1–14, 2025. doi:10.1080/00084433.2025.2593051.

[35]. A. Karkadakattil, “A physics-informed intelligent digital twin using multi-task CNN–LSTM for acoustic emission-based fault prognostics in safety-critical systems,” Journal of Intelligent Manufacturing and Special Equipment, vol. 7, no. 1, pp. 33–58, 2026. doi:10.1108/JIMSE-10-2025-0023.

[36]. A. Karkadakattil, “Physics-guided neural network framework for surface roughness prediction in additively manufactured metallic alloys,” Journal of Laser Applications, vol. 38, no. 2, 2026.