Wave Slamming Forces on Offshore Wind Turbine Jacket Substructures

Abstract

Currently, the development of offshore wind energy is mainly in shallow or intermediate water, where bottom-fixed substructures (e.g. monopiles and jackets) are mainly used. In harsh environmental conditions at certain locations, these substructures are exposed to plunging breaking waves, which cause slamming forces. A slamming force features a high impact force within a short time. It can affect the structural integrity and the fatigue life of Offshore Wind Turbines (OWTs). Slamming forces on cylindrical structures have been widely studied, while the investigations regarding these forces on OWT jacket substructures are limited.

This study addresses the statistical characteristics of slamming forces on OWT jacket substructures and develops a global slamming force model, based on the large-scale experimental data from the WaveSlam project.

Three methods are developed to reconstruct slamming forces on the jacket model, including the optimization-based deconvolution (ODC), vertical approach and extended vertical approach. The vertical approach, which is based on linear regression, is more robust and easier to use than the state-of-the-art horizontal approach that uses deconvolution techniques. The vertical approach can be applied to reconstruct the time series of both local and global slamming forces.

A plunging breaking wave impacts different locations on the braces of the jacket model at different instants in a more or less random order. Statistical analysis indicates that both local and global slamming forces exhibit high variability for the given wave condition, which is controlled in the laboratory. This variability contributes to the uncertainties of the slamming forces, so it is important and should be considered.

A global slamming force model is proposed based on the statistical analysis of the experimental data. The force model involves five parameters, including two exponential parameters and three dimensionless coefficients for the expressions of wave-dependent parameters. Given a sea state, this force model provides a deterministic and conservative prediction of global slamming force time series, which inherently have random features.

Three major aspects regarding the application of slamming forces to OWT simulations are discussed, including the detection of slamming events, the calculation of slamming loads and the integration of slamming loads in fully coupled analysis. A supervised machine learning approach is proposed for the detection of plunging breaking waves. A classifier is trained with wave elevation data by using logistic regression algorithm. The classifier has a better performance than the classical McCowan breaking wave criterion.