Prediction of scour depth around monopiles in combined waves and current: A probabilistic assessment of M5’-MCS

Abstract

Addressing the overlooked uncertainties and stochastic elements in prior studies on scour prediction, this research introduces a probabilistic prediction model for the scour depth around monopile foundations. To enhance the accuracy of the model, the M5’ model tree method was employed to construct a deterministic prediction formula, which was then evaluated using statistical indicators for performance. To address the issue of discontinuities within the deterministic formula, a continuity treatment was applied to improve the credibility of the formula. Expanding on the deterministic formula, a probabilistic model for estimating the local scour depth around a monopile subjected to combined wave and current conditions was developed using Monte Carlo simulations. These simulations integrated specific random parameters into the deterministic model, allowing for the assessment of how these parameters influence the failure probability. The results indicate that the M5’ model tree algorithm can effectively predict the equilibrium scour depth of a monopile under the influence of waves and currents, and the formula, post-continuity treatment, demonstrates enhanced credibility and applicability. Furthermore, the study indicates that the failure probability of a monopile escalates in relation to the increase in near-bed current velocity and the rise in maximum bed surface orbital velocity. It was also discovered that within a specific Froude number range, a consistently low failure probability is maintained, a conclusion that provides a reference for the design of monopile foundations.