A Method for Modal Identification of Existing Structures Using Microtremors


Abstract eng:
Evaluating the structural performance of the existing structures is important to confirm the structural soundness after natural disaster or validity of the seismic retrofit. Modal identification techniques based on vibration records are usually applied to evaluate dynamic characteristics of the target structure. In this technique, we will determine natural frequency, modal vector and modal damping ratio, called modal characteristics as parameters. For architectures and civil engineering structures, modal identification method using microtremor records is desirable, since these records can be observed readily and frequently, comparing with other vibration records such as earthquake records and excitation records. Although the modal identification is generally implemented based on frequency response function which can be obtained from input-output relation, it is difficult to specify and record input vibrations into a structure from microtremor data. Some alternative approaches have been proposed, which is premised on using only output records without the use of the input records. The Frequency domain decomposition (FDD) method proposed by R.Brincker is one of the output records only modal identification method. In this method, the modal parameters are identified by taking the singular value decomposition (SVD) of a power spectral density matrix under assumption of white noise input. Despite the difficulty of input specification, many of methods, which use only output motions, assume input vibrations as white noise. Therefore, accuracy of parameters identified by these methods depends on whether the assumption of input motions are satisfied. A new technique is proposed for the modal identification, which does not require any assumption to the input vibrations. Cross correlation analysis is applied to three records in order to calculate spectral ratio between records at two target sites. The proposed technique for modal identification is presented on the basis of the cross correlation analysis. It is demonstrated analytically under an assumption of individual modes that ratios of cross correlations can reduce the influences from incoherent noise and input motions, which are unknown. A closed form relationship is derived between the ratios of cross correlations and modal vectors. An inversion technique is applied to the relationship and the modal vectors are estimated. The proposed method is applied to a numerical simulation data with five-degree-of-freedom system. It is confirmed that the method can identify modal vectors without any information on input motions. Some difficulties, however, are also recognized for the identification of modal damping ratios. This means that some improved algorithms are required for more accurate identification of structural damping.

Contributors:
Conference Title:
Conference Title:
16th World Conference on Earthquake Engineering
Conference Venue:
Santiago (CL)
Conference Dates:
2017-01-09 / 2017-01-13
Rights:
Text je chráněný podle autorského zákona č. 121/2000 Sb.



Record appears in:



 Record created 2017-01-18, last modified 2017-01-18


Original version of the author's contribution as presented on USB, paper 677.:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)