A DATA-DRIVEN SURROGATE MODEL FOR ESTIMATING RESIDUAL AXIAL RESISTANCE OF CORRODED OFFSHORE TUBULAR MEMBERS FROM INSPECTION DATA
DOI:
https://doi.org/10.61591/jslhu.26.1089Keywords:
Offshore jacket structures; Corrosion degradation; Axial resistance; Machine learning; Inspection data.Abstract
Offshore jacket tubular members are continuously exposed to harsh marine environments, where corrosion significantly affects structural integrity. Therefore, accurate assessment of the residual axial resistance of corroded members is essential for ensuring safe operation and effective maintenance planning of offshore structures. This study proposes a data-driven analytical framework to examine the relationship between corrosion parameters obtained from ultrasonic thickness gauge (UTG) inspections and the residual axial resistance of offshore tubular members. Corrosion-related geometric descriptors extracted from inspection data are used as input features for several machine learning models, including Linear Regression, Random Forest, Support Vector Regression, and XGBoost. The axial resistance is calculated using the ISO 19902 based on the measured geometric properties. The results demonstrate strong predictive performance across the models (R² > 0.99) and confirm that effective thickness and geometric parameters dominate the residual load-carrying capacity. This study highlights the potential of machine learning as a complementary tool for interpreting field inspection data in the structural integrity assessment of offshore structures
References
American Petroleum Institute, Recommended practice for planning, designing and constructing fixed offshore platforms—Working stress design (API RP 2A-WSD), American Petroleum Institute, 2014.
DNV AS, Submarine pipeline systems (DNV-ST-F101), DNV AS, 2021.
International Organization for Standardization, Petroleum and natural gas industries — Fixed steel offshore structures (ISO 19902), 2007.
T. Chen, C. Guestrin, XGBoost: A scalable tree boosting system, Proc. 22nd ACM SIGKDD Int. Conf. Knowl. Discov. Data Min., 2016, 785–794.
M. Komp, Atmospheric corrosion ratings of weathering steels—Calculation and significance, Mater. Perform., 1987, 26 (7), 42–44.
Y. Wang, S. Xu, H. Wang, A. Li, Predicting the residual strength and deformability of corroded steel plate based on the corrosion morphology, Constr. Build. Mater., 2017, 152, 777–793.
X. Liu et al., Physics-informed machine learning for corrosion fatigue life prediction of offshore wind turbine foundations, Reliab. Eng. Syst. Saf., 2022, 218, 108151.
S. M. Lundberg, S. I. Lee, A unified approach to interpreting model predictions, Adv. Neural Inf. Process. Syst., 2017, 30.
R. E. Melchers, Corrosion uncertainty modeling for offshore structures, J. Constr. Steel Res., 1999, 52 (1), 3–19.
J. K. Paik, Ultimate limit state analysis and design of steel structures, John Wiley & Sons, 2018.
Y. Zhu, C. Guedes Soares, Data-driven models for the ultimate strength of corroded stiffened plates, Mar. Struct., 2023, 88, 103345.
Z. Zhao, J. Liu, B. Zhao, Z. Jin, X. Jian, N. Zhang, Shear capacity of corrugated steel plates with a random corrosion damage, Thin-Walled Struct., 2023, 193, 111264.
T. Nakai, H. Matsushita, N. Yamamoto, Effect of pitting corrosion on the ultimate strength of steel plates subjected to in-plane compression and bending, J. Mar. Sci. Technol., 2006, 11, 52–64.
Y. Wang, J. A. Wharton, R. A. Shenoi, Ultimate strength assessment of steel stiffened plate structures with grooving corrosion damage, Eng. Struct., 2015, 94, 29–42.
D.-D. Nguyen, T.-H. Nguyen, Reliability evaluation of 2D semi-rigid steel frames accounting for corrosion effects, J. Mater. Eng. Struct., 2022, 9 (3), 339–353.
S.-M. Nguyen, V.-L. Phan, N.-L. Tran, X.-H. Nguyen, T.-H. Nguyen, Time-dependent reliability assessment of a continuous I-shaped steel beam considering corrosion effects, Eng. Technol. Appl. Sci. Res., 2022, 12 (6), 9523–9526.
L. Breiman, Random forests, Mach. Learn., 2001, 45 (1), 5–32.
T. Chen, C. Guestrin, XGBoost: A scalable tree boosting system, Proc. 22nd ACM SIGKDD Int. Conf. Knowl. Discov. Data Min., 2016.
S. M. Lundberg, S. I. Lee, A unified approach to interpreting model predictions, Adv. Neural Inf. Process. Syst. (NIPS), 2017.
S. H. Xu, C. Guedes Soares, Uncertainty in the ultimate strength of steel plates with pitting corrosion, Struct. Saf., 2012, 34 (1), 211–218.
H. Ahmadi, M. A. Lotfollahi-Yaghin, Stress concentration factors of central brace in statically loaded multi-planar tubular XT-joints, Ocean Eng., 2012, 47, 130–148.
S. Mangalathu, J. S. Jeon, Classification of failure modes of RC bridge columns using machine learning techniques, Eng. Struct., 2018, 160, 226–234.
Y. Zhu, C. Guedes Soares, Reliability analysis of stiffened panels with pitting corrosion using machine learning, Mar. Struct., 2022, 82, 103137.
D.-D. Nguyen, N.-G. Tran, V.-C. Ho, Tr.-H. Nguyen, Prediction of remaining design resistance and bending stiffness of the steel column base plate considering metal corrosion using GA-ANN model, Case Stud. Constr. Mater., 2024, 21, e03930.
