46 S. D. R. Amador et al. 2.5 2 1.5 1 0.5 0 Frequency [Hz] 0 10 20 30 40 50 60 70 Model order [-] Stable pole Vector pole Estimated nat. freqs. Singular Values (a ) 0.21 0.214 0.218 Frequency [Hz] 0 10 20 30 40 50 60 70 Model order [-] (b ) 0.504 0.505 0.506 0.507 Frequency [Hz] 0 10 20 30 40 50 60 70 Model order [-] (c ) 0.8 0.82 0.84 Frequency [Hz] 0 10 20 30 40 50 60 70 Model order [-] (d) Fig . 4. 3 Stabilization diagram constructed with the novel pCF-MM by the identifying models with order ranging from 1 to 70 (a), and details of the closed spaced modes automatically identified with HC algorithm around 0.2 Hz (b), 0.5 Hz (c), and 0.82 Hz (d ) After automatically identifying the operation factor vectors, the mode shape vectors were estimated in a subsequent step by means of the so-called Leas t Square s Comple x Frequency (LSFD ) algorithm [12]. The summary of the identification results obtained for the TV tower with the novel pCF-MM technique is shown in Fig. 4.4. 4.5 Conclusion This chapter shows the results of a multi-dataset modal identification performed on the Riga TV and Radio transmission tower to the estimate the global modal properties of the structure. The vibration responses of the tower measured at different storeys underwent a comprehensive signal processing to (i) remove the phase between reference and roving responses and (ii) estimate the global HS matrix used as primary data in the identification process. The novel pCF-MM was then applied to the prescaled global HS matrix to estimate the tower’s natural frequencies, damping ratios, and operational factor vectors in the frequency range of 0.0 to 2.5 Hz in a first step of the identification process. In a subsequent step, the mode shape vectors were estimated by making use of the so-called LSFD algorithm. The application of the pCF-MM to the Riga TV Tower vibration data led to the construction of a clear stabilization diagram from which eleven vibration modes were clearly and accurately identified. The accuracy and robustness provided by the pCF-MM algorithm were particularly observed in the estimation of the very closed–spaced modes around 0.2, 0.5, and 0.82 Hz.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==