The aim of this research was to investigate how wireless technology, combined with data analytics, can be used for effective structural health monitoring (SHM) of civil infrastructure. Two main applications were investigated, for which wireless sensor networks (WSNs) were integrated into complete SHM solutions: (1) long-term quasi-static monitoring on a suspension bridge, and (2) temporary monitoring of pedestrian bridge vibration.
In the first application, a commercial off-the-shelf WSN was used to acquire and transmit data from extensometers measuring the longitudinal deck displacement of the Tamar Bridge in the UK. Six months of displacement data were analysed in conjunction with the ambient and structural temperature data acquired from a separate monitoring system on the bridge. Empirical models were fitted to relate the deck displacement to various combinations of temperatures. Comparisons of each model’s prediction accuracy showed that the practice of estimating a suspension bridge deck’s thermal expansion based solely on the air temperature is overly simplistic. The deck displacement was predicted more accurately by considering instead the temperatures of the deck itself and of the underlying structure.
In preparation for the second application, a number of indoor tests and short-term deployments on full-scale structures were carried out using an existing prototype WSN, to assess its suitability for vibration monitoring. Subsequently, an embedded data processing method was developed by adapting various signal processing techniques and combining them in sequence. The method was then programmed on the WSN, which was integrated into autonomous SHM systems deployed to monitor two in-service, multi-span pedestrian bridges in Singapore for two weeks. The wireless sensor nodes periodically acquired ambient vibration response data and processed them in a decentralised manner to extract and transmit useful results pertaining to the bridges’ response and modal properties. These results showed that the dynamic properties of the bridges were not affected significantly by the diurnal usage pattern or by the vibration amplitude. The maximum vibration levels recorded on both bridges were found to be within the limits recommended in design guides.
Wireless technology has the potential to make SHM viable for a much broader range of civil structures than it is at the moment. While some WSNs are readily applicable for quasi-static monitoring, considerable development and system integration effort are required to use existing wireless technology in a reliable SHM system for dynamic monitoring.
