Bridges, Tunnels and other Transportation Structures
Full Session with Abstracts
Real-time fatigue health monitoring has the potential to serve as a valuable complement to structural health monitoring (SHM) for bridge inspections. It is an objective supplement to visual bridge inspections and provides data on a bridge’s fatigue condition between often infrequent visits from inspectors. Current methods of continuous structural health monitoring for condition assessment are performed by collecting measured bridge response subjected to operational traffic from an array of sensors installed on fatigue critical members of a bridge. The measured responses are used to determine the remaining fatigue life of the bridge.
The large amount of data involved in this process complicate the design of a system that will automate the data collection process at a bridge, analyze that data, and display information about bridge health to researchers and engineers. Variations in bridge structural systems and condition assessment algorithms also necessitate that such a system have a modular and distributed architecture to allow for expansion to new types of bridge structures. A new system is developed that separates bridge SHM from the data storage and communication system. This new software architecture creates a reliable interface for sending data from one or more bridges to a cloud server where it can be processed using modular algorithms that can be adapted for different types of structures. This design also allows researchers to create parallel platforms for the same bridge using different algorithms, which will allow for comparison and selection of the most efficient techniques as well speedy deployment of new innovations in bridge SHM. The cloud server makes data available to researchers at all stages of processing and allows for data to be securely and reliably backed up to the cloud.
In this research we have conducted bridge condition assessment using statistical bridge signatures and probabilistic hypothesis testing. A bridge signature envelope is generated using a bootstrapping method which gives the expected response of the bridge under live load operational traffic events. This signature envelope represents a confidence interval for the response of a healthy bridge that can be used to monitor bridges for potential damage. A bridge is considered healthy if a sample of recent events falls within range of this signature envelope. Hypothesis testing based on a nonparametric rank sum test is used to determine the likelihood that a sample of recent events represents a significant divergence from the expected bridge response based on all prior events. This test generates a single value that can be used to generate alerts for bridge managers based on the severity of the divergence from the expected response.
The processed bridge response to operational real-time traffic events can be accessed from the cloud server through a simple programmatic interface, allowing for the development of visualizations and other tools for researchers and bridge managers to assess the condition of the bridge. This system provides significant advantages over previous platforms for structural health monitoring and condition assessment, most notably in the areas of modularity and reliability.