Experimental verification of a novel system identification technique that can detect defects at the element level is successfully accomplished. The method can be used for in-service health assessment of real structures without disrupting normal operations. This study conclusively verifies the method.Analytical verification of the proposed algorithm has been successfully completed by the research team at the University of Arizona. Vo and Haldar (2004) experimentally verified the method by conducting tests on fixed-ended and simply supported defect-free and defective beams. The purpose of this research was to validate the method by conducting experiments with more realistic structures. A three-story one-bay steel frame, built to 1/3 scale to fit the experimental facility, was considered. The frame was excited by harmonic or impulsive excitation forces. The transverse acceleration responses were collected using capacitive accelerometers. The angular displacement responses were measured using an autocollimator. The dynamic responses of the frames were collected by a data acquisition system with simultaneous sampling capability. Using only experimentally collected response information and completely ignoring the excitation information, the stiffness of all the structural elements were identified. The method identified the defect-free frame very accurately. Defects, in terms of removing a beam, reducing cross sectional area over a small segment of a beam, and cutting notches in a beam, were introduced. The method correctly identified the defect location in all cases. Additional sensors were placed around the location of the defect in an effort to identify the defect spot more accurately. The proposed method also successfully identified defect with improved accuracy. To increase the implementation potential of the proposed method, the defect-free and defective frames are then identified using limited response information. A two-stage Kalman filter-based approach is used. It is denoted as Generalized Iterative Least Square Extended Kalman Filter with Unknown Input (GILS-EKF-UI) method. A sub-structure approach is used for this purpose. The GILS-EKF-UI method also identified the state of the structure using only limited response information. As expected, in this case the error in the identification goes up as less information is used. However, the error is much smaller than other methods currently available in the literature, even when input excitation was used for the identification purpose. The method is very robust and can identify defects caused by different types of loadings. The method can be used as a nondestructive defect assessment technique for structures.
| Identifer | oai:union.ndltd.org:arizona.edu/oai:arizona.openrepository.com:10150/193974 |
| Date | January 2005 |
| Creators | Martinez-Flores, Rene |
| Contributors | Haldar, Achintya, Haldar, Achintya, Richard, Ralph, Fleischman, Robert, Contractor, Dunshaw N. |
| Publisher | The University of Arizona. |
| Source Sets | University of Arizona |
| Language | English |
| Detected Language | English |
| Type | text, Electronic Dissertation |
| Rights | Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author. |
Page generated in 0.0109 seconds