Abstract

Offshore wind turbines (OWT) are slender structures with sensitive dynamics, strongly influenced by soil-structure interaction. These structures are subjected to various dynamic loads with natural frequencies close to the dynamic response of the offshore wind turbine. This study is aimed at the precise evaluation of the initial first natural frequency of a wind turbine. The present paper deals with design of an offshore wind turbine scaled model and experimental determination of its first natural frequency. The results compared with existing analytical methods allow us to conclude on the relevance of this scaled model and on the accuracy of the considered analytical methods.

Introduction

Offshore wind turbines are submitted to various dynamic loads that need to be taken into account for the design of these structures. In order to avoid resonance phenomenon, the response of the structure must not overlap the natural frequencies of the relevant loads. The environmental loads - wind, waves, and currents - are dynamic loads with natural frequencies below 0.1 Hz. Moreover, the rotor frequency (1P) as well as the blade passing frequency (3P) define frequency ranges that need to be considered in the design as their behavior influenced the wind load on the mast and the blades. These loads and respective frequencies for the reference NREL 5-MW offshore wind turbine [Jonkman, 2009] are summed up in Fig. 1. Standard offshore wind turbines are designed as "Soft-Stiff" structure, their first natural frequency is located in the interval between the rotor frequency (1P) and the blade passing frequency (3P).

As the first natural frequency of the structure is lying in a narrow interval, it is essential to evaluate precisely this parameter. In-situ measurements on offshore wind turbines from Walney offshore wind farm presented by [Kallehave et al., 2012] showed that the current standard method defined by the American Petroleum Institute [API, 2000] underestimated the first natural frequency of the structure with an error between 5 and 7 %. This inaccuracy is mostly due to underestimation of the soil stiffness. Hence, different modified API methods presented among others by [Sorensen et al., 2010], [Kallehave et al., 2012] and [Thieken et al., 2015] were developed to model more accurately the soil-structure interaction considering small or extreme loads on the structure.

This content is only available via PDF.
You can access this article if you purchase or spend a download.