Proc. SPIE. 8692, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2013
KEYWORDS: Detection and tracking algorithms, Visualization, Sensors, Databases, Linear filtering, Data archive systems, Signal processing, Structural health monitoring, Damage detection, Sensing systems
For the application of structural health monitoring (SHM) system to the post-earthquake damage screening of building structures, an immediate evaluation of the degree of damage in primary structural components is a challenging task. To increase the resolution in damage detection above a certain level to detect damage in individual components, a SHM requires the use of a dense array of sensors deployed to building structures. In order to deal with a large amount of data acquired by the sensing network and to distribute quick safety alerts on the condition of earthquake-affected buildings, a SHM system that is connected with a cyberinfrastructure specifically designed for the autonomous structural integrity assessment of buildings is developed. In the system, big data transferred from a dense sensing network is automatically stored and processed to extract damage features using a PostgresSQL relational database and embedded local damage detection algorithms. In a benchmark study, the schema of the SHM system is specifically designed to function with a built-in local damage detection algorithm that needs a comparative study of current dataset with past reference dataset. To visualize the results of the damage detection analysis, a PHP-based web-viewer is also designed for the SHM system. Finally, the performance of the developed cyber-based SHM system is evaluated through a series of the damage detection tests on a 5-story steel testbed frame that can replicate damage in beams and columns.
A structural health monitoring system that aims to extract local damage information (i.e., existence, location and severity) in buildings may require a dense array of transducers due to the high complexity and high degree of statical indeterminacy of their structural system. While monitoring systems for building applications are mostly consisted of seismographs or tremor sensors, a technique to pragmatically and accurately capture strain information of structural members is efficacious for detecting damage in individual members. This paper presents the use of polyvinylidene fluoride piezoelectric films as dynamic strain sensors for detecting local damage in steel moment-resisting frames. First, a damage detection methodology that monitors the changes in the relative distribution of the bending moments in structural systems is presented. Next, an array of dynamic strain sensors networked by wireless sensing units is developed in consideration of its installation cost and efforts when it is applied to real buildings. Finally, the performances of the developed methodology and its sensing system are evaluated through a series of vibration testing using a 5-story steel testbed frame that can simulate seismic damage at beam-to-column connections.