کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
561033 | 1451857 | 2016 | 16 صفحه PDF | دانلود رایگان |
![عکس صفحه اول مقاله: Rotor anisotropy as a blade damage indicator for wind turbine structural health monitoring systems Rotor anisotropy as a blade damage indicator for wind turbine structural health monitoring systems](/preview/png/561033.png)
• We propose a novel damage indicator for a wind turbine structural health monitoring system.
• The indicator is based on loss of rotor isotropy due to a blade damage.
• The experimental technique for extracting the damage indicator from measured vibration data is proposed.
• We demonstrate the damage indicator using a simple rotor model, on data from simulated wind turbine and finally, on experimentally obtained data.
Structural damage of a rotor blade causes structural anisotropy of the rotor. In rotor dynamic, the anisotropy affects the symmetry of the rotor mode shapes, and the latter can be utilized to detect the blade damage. The mode shape symmetry can be characterized by relative blades’ magnitude and phase. The study examines the potential use of these parameters as rotor damage indicators.Firstly the indicators are studied analytically using a simple 6 degrees-of-freedom model of a rotating rotor. Floquet analysis is used due to the time periodic nature of the considered system. Floquet analysis allows one to perform analytical modal decomposition of the system and study the sensitivity of the damage indicators to the amount of damage. Secondly, operational modal analysis (OMA) is involved to extract the same damage indicators from simulated experimental data, which was synthesized via numerical simulations.Finally, the same procedure was applied to operating Vestas V27 wind turbine, first using the simulated experimental data obtained by using aeroelastic simulation code HAWC2 and then using the data acquired during the measurement campaign on a real wind turbine.The study demonstrates that the proposed damage indicators are significantly more sensitive than the commonly used changes in natural frequency, and in contrast to the latter, can also pinpoint the faulty blade. It is also demonstrated that these indicators can be derived from blades vibration data obtained from real life experiment.
Journal: Mechanical Systems and Signal Processing - Volume 74, 1 June 2016, Pages 183–198