Fiber Optic Method for Bridge Health Assessment Based on Long-Gauge Sensors

Princeton University, Princeton NJ

Investigators

Abstract

For civil infrastructure there is a need for structural health monitoring to provide an optimized maintenance, lifetime extension, and improved safety. Structural health monitoring can provide accurate and in-time information concerning condition and residual life of civil infrastructure. It consists of instrumentation of the structure with sensors and transformation of the collected data into meaningful and actionable information for the owner or manager of the structure, so decision makers can undertake informed and economical maintenance and repair actions. This project focuses on the typical bridge beam systems that represent 60 percent of all US bridges. The objective of this project is to pursue research in creating and validating a universal health assessment system for typical beam bridges based on static and dynamic measured strains using long-gauge fiber optic sensors. This research will provide means to avoid disruptive closures and catastrophic failures, while also ensuring on-going reliable health assessment. The success of the project stands to have profound societal impact by addressing public safety and minimizing economic disruption. The neutral axis and the deformed shape are universal beam parameters that reflect bridge health and performance condition. Real-time determination of these parameters has potential to enable a robust bridge health monitoring assessment method. Fundamental questions associated with determination of these parameters addressed in this project are related to: 1) variability of these parameters and uncertainty in their determination under damaged and undamaged conditions, and 2) transformation of the collected data into meaningful and actionable information to the bridge owner or manager. The objective of the project is to create a reliable method for performance and health assessment of beams in bridges based on determination of the position of the neutral axis and on deformed shape. State-of-the-art long-gauge fiber optic strain sensors have inherent advantage to acquire data relevant to the global structural scale. Algorithms will be created based on general probabilistic approaches, Bayesian structural identification, and performance prediction models taking in to account variability and uncertainties in collected data. Experimental validation will be performed both through laboratory tests and controlled tests on two bridges in-use. The method will be applicable to existing and new beam bridges and will not be material dependent.