A Comparative Assessment of In-Operation Modal Analysis and Frequency Domain Decomposition Algorithm Using Simulated Dat
Recently a new In-Operation modal identification algorithm has been proposed by one of the authors (INOPMA). The algorithm is based on considering the correlation sequence of the outputs of the structure as an impulse response but with a certain phase shi
- PDF / 203,930 Bytes
- 7 Pages / 439.37 x 666.142 pts Page_size
- 52 Downloads / 188 Views
Abstract Recently a new In-Operation modal identification algorithm has been proposed by one of the authors (INOPMA). The algorithm is based on considering the correlation sequence of the outputs of the structure as an impulse response but with a certain phase shift. A numerical modal appropriation is then performed and it is shown that a mode is isolated at a certain characteristic frequency. The estimation of the modal parameters is then done in a standard way. We investigate in this paper the performance of INOPMA under realistic excitation and we compare it to the Frequency Domain Decomposition algorithm (FDD). The example is a 3DOF system under El-Centro seismic excitation. It is shown that both methods perform equally well. Keywords In-operation modal identification analysis Identification
Modal appropriation
Modal
1 Introduction The algorithms estimating the dynamic parameters of structures just based on the output responses became popular as operational modal analysis (OMA) or output-only modal analysis or ambient vibration analysis or in-operation modal analysis. Primary studies about OMA were established in 1990s. Researchers, particularly civil engineering community, deeply focused on OMA techniques since M. Abdelghani (&) Department of Mechanical Engineering, ISSAT-University of Sousse, Sousse, Tunisia e-mail: [email protected] M. Ghalishooyan A. Shooshtari Department of Civil Engineering, Ferdowsi University, Mashhad, Iran e-mail: [email protected]; [email protected] A. Shooshtari e-mail: [email protected]; [email protected] © Springer International Publishing Switzerland 2017 T. Fakhfakh et al. (eds.), Advances in Acoustics and Vibration, Applied Condition Monitoring 5, DOI 10.1007/978-3-319-41459-1_21
215
216
M. Abdelghani et al.
about 15 years ago. Because, mostly there is no need to artificially excite the structures which will be expensive in the case of large structures. On the other hand, applying this method doesn’t interrupt the daily usage of the structures. Over the years, OMA has evolved as an autonomous discipline and have been attracting great research interest for many years. According to the domain of implementation, OMA methods can be categorized into time domain and frequency domain approaches. For the past 20 years, different OMA procedures have been proposed by researcher. Natural excitation technique (NExT) is regarded as one of the earliest algorithms of OMA which presented in the 1990s (James et al. 1995). To identify the structural modal parameters when applying NExT, several basic system estimation time domain methods such as Least Square Complex Estimation (LSCE) (Brown et al. 1979), Ibrahim Time Domain (ITD) (Ibrahim and Mikulcik 1977), Polyreference Time Domain (PTD) (Vold and Rocklin 1982) and Eigensystem Realization Algorithm (ERA) (Juang and Pappa 1985) are utilized. Auto-Regressive Moving Average (ARMA) model was widely used for estimating the modal parameter of civil engineering structures excited by white noise (Lju
Data Loading...
