Journal of Scientific and Industrial Research

Open Access Open Access  Restricted Access Subscription Access

Modal Parameter Estimation of Tall Structures using HHT with Improved EMD


Affiliations
1 CSIR-Structural Engineering Research Centre, CSIR Campus, Taramani, Chennai 600 113, Tamilnadu, India
2 Dept of Civil and Environmental Engineering, Carnegie Mellon University, Pittsburgh, PA 15213, United States

In this paper, identification of the natural frequencies, damping ratios and mode shapes of the structures using the measured ambient responses are proposed using time-frequency analysis. The impulse responses are obtained from the measured acceleration time history data through cross-correlations. Empirical Mode Decomposition (EMD) is employed on these generated impulse responses to obtain Intrinsic Mode Functions (IMFs). Finally, modal identification of the structure is carried out by performing Hilbert Transform (HT) on these generated IMFs. To avoid the problem of mode mixing during EMD of the signal, an improved version with intermittency criteria along with treatment to end effects during sifting is proposed in this paper. Experimentally measured data of Guangzhou New TV (GNTV) Tower is used to test and verify the proposed algorithm. The studies indicate that the proposed HHT based algorithm can be applied quite effectively for the modal identification of practical engineering structures.
User
Notifications
Font Size

Abstract Views: 149




  • Modal Parameter Estimation of Tall Structures using HHT with Improved EMD

Abstract Views: 149  | 

Authors

K Lakshmi
CSIR-Structural Engineering Research Centre, CSIR Campus, Taramani, Chennai 600 113, Tamilnadu, India
Varun Kasi Reddy
Dept of Civil and Environmental Engineering, Carnegie Mellon University, Pittsburgh, PA 15213, United States
A Rama Mohan Rao
CSIR-Structural Engineering Research Centre, CSIR Campus, Taramani, Chennai 600 113, Tamilnadu, India

Abstract


In this paper, identification of the natural frequencies, damping ratios and mode shapes of the structures using the measured ambient responses are proposed using time-frequency analysis. The impulse responses are obtained from the measured acceleration time history data through cross-correlations. Empirical Mode Decomposition (EMD) is employed on these generated impulse responses to obtain Intrinsic Mode Functions (IMFs). Finally, modal identification of the structure is carried out by performing Hilbert Transform (HT) on these generated IMFs. To avoid the problem of mode mixing during EMD of the signal, an improved version with intermittency criteria along with treatment to end effects during sifting is proposed in this paper. Experimentally measured data of Guangzhou New TV (GNTV) Tower is used to test and verify the proposed algorithm. The studies indicate that the proposed HHT based algorithm can be applied quite effectively for the modal identification of practical engineering structures.