A data-driven model for combined current-wave induced boundary layers

• A novel data-driven framework for predicting the velocity profiles of combined current-wave induced boundary layers is proposed. • A database for the velocity profiles of combined current-wave induced boundary layers under different flow conditions is built. • A deficit function is used to model the wave part, and a multilayer expression is used to model the current part. This study introduces a novel data-driven framework for efficiently predicting the time-varying velocity profiles of combined current-wave induced boundary layers, which are crucial for analyzing the on-bottom stability of near-seabed pipelines and cables. The characteristics of the combined wave-current boundary layer are strongly dependent on various parameters such as the wave period, the wave semi-excursion and the seabed roughness. Furthermore, there are also interactions between the current part and the wave part of the boundary layer, which influence the behaviors of their velocity profiles. To address these challenges, a data-driven model integrating one-dimensional Computational Fluid Dynamics (CFD) simulations and parametric mathematical expressions is proposed to predict the velocity profiles of the current and the first harmonic of the wave part for the boundary layer. CFD simulations are performed to generate the velocity profiles under different flow conditions and the database for the variables in the mathematical expression is built. An interpolation methodology based on the database is used to obtain the corresponding variables under a new given flow parameter set. This framework is then validated against numerical simulations and experimental data.