Normalized energy index-based signal analysis through acceleration trendlines for structural damage detection

Abstract

Utilizing the Trend line-based Signal decomposing Algorithm (TSA), this paper develops an innovative Structural Health Monitoring (SHM) method to locate damages in structures on the basis of vibration signature analysis. Firstly, a suitable vibration mode of the structure should be selected. The TSA is then used to determine special trend lines of the structural response according to the selected vibration mode. The damage is located through a normalized energy-based damage index calculated for these special trend lines. The accuracy of the developed method is examined using two examples: an experimental five floor laboratory building model, and a numerical arch-tied bridge, both subjected to seismic excitations. For the building model, the seismic excitation was exerted to the building model using a shaking table and the seismic acceleration responses were recorded using a set of wireless accelerometers installed at the center of each floor. For the arch-tied bridge, the numerical simulation was performed in ABAQUS software and the seismic acceleration responses were recorded on the deck near the cables. For the building model, damages were modeled as loss of stiffness in columns on different floors. For the arch-tied bridge, damages were modeled as loss of stiffness in cables. The results show that, provided that the response records and reflects damage-induced changes in the monitored structure, the developed method can accurately locate the damaged floor or damaged cable in both examples. Additionally, using the normalized energy-based method to locate the loss of stiffness in a structure makes the developed method a good candidate for rapid and simple assessment of a structure’s health state.