Abstract
Monopiles are a widely-used foundation system for offshore wind turbine support structures. In current practice, design calculations typically employ one-dimensional (1D) models in which the monopile is represented as an embedded beam. The current study presents a data-driven 1D design model for the analysis of offshore monopiles subjected to monotonic lateral load and moment loading. The method is based on the PISA design model framework; enhancements are incorporated in the model to improve its accuracy, scalability and to facilitate applications to a wide range of geotechnical conditions. The data-driven model incorporates a spline-based parametrisation of the soil reaction curves combined with machine learning techniques. The model is calibrated using a database of previously-published three-dimensional finite element calibration analyses. The method described in the current paper is concerned with:•Modifications to the PISA design model framework to develop a data-driven 1D design model.•Calibration of the data-driven 1D model for ground conditions comprising: (i) offshore glacial tills with varying strength-stiffness properties, and (ii) sands with a wide range of relative densities.•Validation of the proposed method by comparing 1D model predictions for monopiles in homogeneous and layered soils with detailed 3D finite element analyses.