Dynamic reliability analysis of a floating offshore wind turbine under wind-wave joint excitations via probability density evolution method

Yupeng Song, Biswajit Basu, Zili Zhang, John Dalsgaard Sørensen, Jie Li, Jianbing Chen*

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

37 Citations (Scopus)

Abstract

Floating offshore wind turbine (FOWT) towers are dynamically sensitive to wind and wave excitations. Since the wave tends to be harsher with the increase of wind speed, FOWT towers are likely to experience the most severe vibration operating at the cut-out wind speed. In the present study, the short-term dynamic reliability of a spar-type FOWT is evaluated based on the probability density evolution method (PDEM). For this purpose, an integrated coupled dynamics model for the FOWT is firstly established by incorporating the multibody dynamics with the finite element (FE) method. Next, the conditional joint probability distribution of the significant wave height and peak spectral wave period at the cut-out wind speed is constructed based on the copula model. Then, the stochastic dynamic response and reliability of the FOWT can be analyzed via PDEM. The numerical example of reliability analysis of a 5-MW spar-type FOWT operating at the cut-out wind speed is carried out, in which the long-term met-ocean data at a South China Sea site is utilized. Simulation results show that the reliability of the FOWT for normal operation is less than 0.2 when the acceleration at the tower top is adopted as the failure criterion.

Original languageEnglish
JournalRenewable Energy
Volume168
Pages (from-to)991-1014
ISSN0960-1481
DOIs
Publication statusPublished - May 2021

Keywords

  • Dynamics modeling
  • Floating offshore wind turbine
  • Probability density evolution method
  • Reliability analysis
  • Wind-wave joint distribution

Fingerprint

Dive into the research topics of 'Dynamic reliability analysis of a floating offshore wind turbine under wind-wave joint excitations via probability density evolution method'. Together they form a unique fingerprint.

Cite this