74 W. Fladung and K. Napolitano Fig . 7. 8 Singular values of the imaginary part of an FRF matrix (QMIF) versus singular values of the complex FRF matrix (CMIF) 7.6 Conclusions With the current state of data acquisition and computer capabilities, we can acquire and process large volumes of data without waiting too long for the results. And while it might be tempting to use all the data all the time for parameter estimation, a recurring theme in this chapter has been “Don’t use data that doesn’t help your cause.” Excluding frequency ranges containing noise (be it of the variance, periodic, or bias variety) is a simple yet effective way to heed this advice. In addition, selecting frequency bands around the poles that only use data that constructively contributes to estimating their residues is another. Although residues are, in their most general form, complex—and some software packages may use vector complexity as a quality-of-fit metric—you can also solve directly for normal modes using the quadrature part of the FRF. In doing so, modal parameters can oftentimes be more accurately estimated. The tips, tricks, and obscure features described in this chapter are based on actual events that necessitated their development over the years for real-world test data. However, due to the nature of the authors’ business, these data belong to our customers and cannot readily be shared in an open forum (at least not without a lot of paperwork). Although the data presented herein are taken from nonproprietary laboratory test articles (e.g., “iron birds”), the contrived examples are meant to be representative of the types of situations in which these techniques have proven useful. Bibliography 1. Fladung, W., Vold, H.: An orthogonal view of the polyreference least-squares complex frequency modal parameter estimation algorithm. In: Proceedings of the 33rd International Modal Analysis Conference, pp. 171–182. Springer, Cham (2015) 2. Fladung, W., Vold, H.: An improved implementation of the orthogonal polynomial modal parameter estimation algorithm using the orthogonal complement. In: Proceedings of the 33rd International Modal Analysis Conference, pp. 157–170. Springer, Cham (2015)
RkJQdWJsaXNoZXIy MTMzNzEzMQ==