Design And Evaluation of Flexibility-Based Structural Damage Localization Using Wireless Sensor Networks

Abstract

The health of civil structures is very important and sometimes life-critical. While there are different ways to monitor their health, wireless sensor network (WSN) has the advantage of easy deployment and low cost, which make it feasible for most structures. We designed and implemented a system to localize damages on structures with a WSN by detecting the change in structure flexibility. This method has been validated to work well on bridges like a cantilever beam and a truss. It is also possible to be extended to other type of structures. Different from other systems, in network data processing was applied to lower the bandwidth requirement of large amount of raw sensing data. Only the intermediate computation results, that capture the flexibility related information, were transmitted back to the base station. We also divide the detection and localization into multiple levels. Lower level acts as the sentinel to detect the existence of damage; and higher levels, which consume more energy, are then triggered when necessary to get a higher resolution of localization. This design helps to further extend the lifetime of the system. Type of Report: MS Thesis Department of Computer Science & Engineering Washington University in St. Louis Campus Box 1045 St. Louis, MO 63130 ph: (314) 935-6160 ii ABSTRACT OF THE THESIS DESIGN AND EVALUATION OF FLEXIBILITY-BASED STRUCTURAL DAMAGE LOCALIZATION USING WIRELESS SENSOR NETWORKS by Weijun Guo Master of Science in Computer Science Washington University in St. Louis, 2009 Research Advisor: Professor Chenyang Lu The health of civil structures is very important and sometimes life-critical. While there are different ways to monitor their health, wireless sensor network (WSN) has the advantage of easy deployment and low cost, which make it feasible for most structures. We designed and implemented a system to localize damages on structures with a WSN by detecting the change in structure flexibility. This method has been validated to work well on bridges like a cantilever beam and a truss. It is also possible to be extended to other type of structures. Different from other systems, in network data processing was applied to lower the bandwidth requirement of large amount of raw sensing data. Only the intermediate computation results, that capture the flexibility related information, were transmitted back to the base station. We also divide the detection and localization into multiple levels. Lower level acts as the sentinel to detect the existence of damage; and higher levels, which consume more energy, are then triggered when necessary to get a higher resolution of localization. This design helps to further extend the lifetime of the system.