Condition monitoring can be used to detect faults and failures at an early stage. Thus it decreases the overall maintenance expenses. This report gives an example of condition monitoring with focus on early crack detection in the main shaft of an offshore wind turbine. This article discusses the applicability of different condition monitoring techniques like performance monitoring, strain gauge results and vibration analysis for crack detection on the low speed shaft. Different signal processing methods like descriptive statistics, Fourier Transforms, Wavelet transforms, Modal Assurance Criteria and Time Synchronous Averaging are investigated and tested for applicability in order to detect cracks on the low speed shaft of a wind turbine. The results of this study can be used to define alarm thresholds as well as detection characteristics for main shaft cracks leading to failure of the main shaft. The example in this paper showed that the crack on the main shafts needs to be relatively large (around 20 % of the main shaft diameter) until it can be detected.
展开▼
