harmonics in the fixed frame. Depending on the operating conditions (wind speed and rotor speed) these frequencies may coincide with the eigenfrequencies of the system and affect the estimated system parameters. Several researchers performing measurements on wind turbines report that they had to discard some of their measurements and analyses since several P harmonics coincide with the real turbine modes (27). Such a problem can be solved by two different ways. Vibration data measured at a different rotational speed can be used for the identification. Although such a solution would separate the previously coinciding real eigenfrequency and the P harmonics, it is not always easy to find the required measurement period because of the reasons mentioned above. As a second alternative, Mohanty and Rixen (33), (34) proposed a modified version of OMA which was extended to identify modal parameters in the presence of harmonic excitations even with frequencies close to the eigenfrequencies of the system. 5. Identified System Parameters This section presents the results of the infield tests performed on the 2.5 MW – 80 meter diameter - wind turbine whose technical properties were given in the previous sections. The main features of the utilized measurement systems, tests durations, and sampling frequencies are summarized in Table 1. The acquired dynamic response was then analyzed by using LSCE (Least Square Complex Exponential) method which is in theory based on NExT (Natural Excitation Method) explained before. It should be noted that in order to protect the interests of the turbine manufacturer, the frequencies mentioned in this work were normalized and not given explicitly. 5.1. Operational Modal Analysis of the Parked Turbine The measurements and analysis results summarized in this section was obtained using the measurements from the strain gauges installed in the turbine and the LDV (laser Doppler vibrometer). Several laser measurements were performed both on the tower and the blades. Figure 9 shows the PSD (Power Spectral Density) graph calculated from the laser measurements taken on the tower. Although there are 2 different tower modes having almost the same frequencies, the clear peak shown in the figure represents the longitudinal tower frequency because the instantaneous orientation of the laser was almost the same as the longitudinal tower vibration while the data series were being acquired. The obtained data series were also processed by using LSCE method. The lateral tower vibration could not be identified. The average damping ratios were calculated as 0.3% with minimum and maximum values of 0.1% and 0.5%, respectively. The turbine was observed to be yawing slowly during the measurement which produced the maximum damping ratio of 0.5 %. Therefore, it is thought to be caused by the drag phenomenon mentioned above whereas the minimum damping of 0.1 % is believed to be due to insufficient excitation and very low wind speed experienced during the test. Since the damping value is very low, the measurement duration of 294 sec (approximately 100 cycles) was long enough to identify tower and several 267
RkJQdWJsaXNoZXIy MTMzNzEzMQ==