332 T. Marinone et al. Fig. 31.2 Nacelle supported (upper left), nacelle excited by shakers (upper right), and test-display model showing node locations (bottom) Signal processing for the shaker runs used a Hanning broad window with overlap averaging. The impact runs were processed with no window since the excitation and response signals were near zero at the beginning and end of each frame of data. 31.7 Test Results This section documents the test results. The rigid body modes were estimated for the two nacelles and one hub. The mass-line approach was only applied to the hub. Figure 31.3 shows the power spectrum mode-indicator function (PSMIF) for the nacelle with a single reference. While two shakers were used during the test, a single reference provided better fits and mass estimation than both references. The PSMIF, as the summation of all FRFs multiplied by each FRF conjugate, is an ideal function to visualize the overall response of the structure. The six rigid body modes are observed in the six peaks of the response. To evaluate the quality of the rigid body modes obtained, the PSMIF was then synthesized using only the six rigid body modes. The synthesized PSMIF shows a high level of correlation with the test data. As such, there is confidence that the extracted rigid body modes will accurately describe the inertia properties of the nacelles. Figure 31.4 shows the complex mode-indicator function (CMIF) for testing the hub with two shakers. Again, all six rigid body modes can be seen in the peaks of the response (the CMIF is plotted for both shaker references in order to clearly distinguish modes 5 and 6). To evaluate the quality of the rigid body modes obtained, the CMIF was then synthesized using
RkJQdWJsaXNoZXIy MTMzNzEzMQ==