Identification of structural changes using symbolic representations of modal quantities
Abstract »A structural health monitoring strategy based on the control of structural frequency data over time, obtained from operational modal analysis, is presented in the present paper. The strategy relies on modal estimation based on Stochastic Subspace Identification and clustering methods and, unlike most methodologies found in previous works, it does not require the tracking of each structural mode through time. Instead, it relies on extracting histograms of frequency data and in quantifying the dissimilarities between sets of these histograms, over time.
The strategy is tested and validated on modal estimates obtained from the monitoring system of the suspended 25 de Abril bridge, located in Lisbon, Portugal. The obtained results show that the proposed strategy is capable of highlighting small-magnitude changes in a few number of mode shapes, while controlling a large range of structural frequencies (and, consequently, a large number of structural modes). When applied to smaller frequency subranges, the strategy proves capable of identifying the frequency values more susceptible to the damage being observed, thus contributing for the localization and magnitude assessment of the changes monitored on site.
Biography: Postdoctoral Researcher at LNEC (Portuguese National Laboratory for Civil Engineering). Obtained the PhD from the University of Lisbon, on the topic of smart structural health monitoring techniques applied to bridge structures.
Affiliation: National Laboratory for Civil Engineering (LNEC) Structures Department 1700-066 Lisbon Portugal