integer multiples of rotational frequency (P components) the other turbine modes which have relatively weaker participations in the overall response and which can also exhibit significant damping cannot be identified easily from PSD plots only. However, several system identification methods enable weaker frequencies to be detected. A more detailed investigation (22) of the measured turbine response carried out by utilizing these identification methods showed that the second edgewise, second and third flapwise modes of the blades can also be identified. However, further analyses and measurements are needed for validation of observability of these high frequency modes from photogrammetric measurements (23). The in-operation response data was also recorded by the strain gauges installed in the blades. Unfortunately, symmetric, FW (forward whirling) or BW (backward whirling) flapwise modes could not be identified from the short measurement durations available. A possible explanation can be that the correlation function based identification methods need very long measurement durations to detect high damping modes. However, it is believed that measurement and analysis of longer data series will enable these high damping modes to be detected too. On the other hand, the 2 tower modes having relatively lower damping ratios could be detected by strain gauge data. Figure 12 Change of tower damping ratios wrt. wind speed 273
RkJQdWJsaXNoZXIy MTMzNzEzMQ==