Bridge evaluation utilizing structural health monitoring data

Date
2016
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
The safety and serviceability of bridges is of paramount concern for bridge owners and for the traveling public. As our bridge infrastructure continues to age, there is a growing need for new methods and technologies that can enable transportation agencies to better evaluate their bridges to ensure their structural safety and to optimize their maintenance and inspection procedures. Following the collapse of the I-35 Bridge in Minnesota in 2007, the Federal Highway Administration (FHWA) began requiring a bridge load rating for all bridges in the United States. According to the American Society of Civil Engineers, one out of every nine bridges in the United States is classified as structurally deficient and is in urgent need of repair. The required maintenance of these and other bridges is very expensive. In fact, the FHWA estimates that it would cost nine billion dollars per year more than what is currently being spent on bridge maintenance to repair and maintain our deficient bridges. Structural Health Monitoring (SHM) is a technique that has been evolving and has been used in recent years to measure the loading environment and response of bridges in order to assess serviceability and safety. There are several examples around the world that have demonstrated the benefits of SHM using both short- and long-term monitoring. However, transportation agencies still lack the ability to directly implement SHM data into their maintenance and decision making processes. More specifically, transportation agencies are generally not capable of implementing the existing complex methods for using short- or long-term SHM data for bridge evaluation. ☐ The primary objective of this study was to develop new methods for utilizing SHM data that are analogous to more traditional methods and can be easily implemented by transportation agencies to better evaluate their bridges to achieve optimal maintenance and effective decision making. In developing the new methods, two approaches were taken. ☐ The first approach, referred to as the Continuous Rating Factor-Structural Health Monitoring method, uses SHM data to compute continuous rating factors. This approach applies SHM data directly into the Load Resistance Factor Rating (LRFR) equations to produce continuous rating factors for specific bridge components. To do this, the continuously recorded SHM data is converted into structural forces and/or stresses and incorporated directly into conventional rating equations to calculate continuously rating factors over time. More specifically, this new approach converts the measured strain and temperature data to live loads, thermal loads, prestressing losses, i.e. to yield accurate site-specific rating factors for various critical bridge components. ☐ The second approach, referred to as the Reliability Analysis-Structural Health Monitoring method, uses a reliability analysis framework combined with SHM data. In this approach loads and resistances were expressed as Probability Distribution Functions (PDF), where loads and resistances are treated as random variables. The concept of estimating the probability of failure or probability of exceedance is utilized and expressed as a reliability index for a specific bridge component. The reliability analysis was conducted first using design loads and then using long-term SHM data. The analyses were performed using Monte Carlo simulation and Rackwitz-Fiessler method and considered a variety of limit states. In the first type of analysis (using design information), the resistance model, dead load model, and live load model used in the reliability analysis were based solely on design information. In this analysis, the same statistical parameters used to develop the load effects and resistances in the AASHTO LRFD calibration were applied. In the second type of analysis (using SHM data), the load effects consisted of dead, live, and thermal loads. A live load statistical model was created based on data from Weigh-In-Motion (WIM) stations close to the location of the IRIB and a 3-D finite element model. The thermal load statistical model was created based on data from Delaware Environmental Observing System (DEOS) and correlation analysis between measured SHM strain and temperature data from the IRIB. In both cases, reliability indices for the west edge girder were computed along the bridge for various limit states. ☐ In order to demonstrate the two methods, the Indian River Inlet Bridge (IRIB), a prestressed concrete cable-stayed bridge located in Sussex County Delaware, was used as a study case. The research showed that the two methods can serve as possible evaluation approaches for bridges that have SHM systems. Both methods are successful in taking huge amounts of SHM data and translating them into simple and well understood evaluation parameters (ratings and reliability indices). ☐ The primary findings from results given by the continuous rating factor method were (1) SHM data can be used to directly compute bridge load ratings, (2) the developed technique provides results that can be easily understood and utilized by transportation agencies, and (3) the ratings show that thermal effects can have a significant effect on load ratings for long-span bridges. The primary findings from results given by the reliability method based on SHM data were (1) the method can be used to determine whether or not the monitored bridge meets the design code standards in terms of reliability by allowing a comparison of the target reliability indices to indices computed based on SHM data, (2) the developed reliability-based methodology using SHM data can be applied to other bridges, (3) the developed method shows promise for enabling SHM data to be directly incorporated into the maintenance, inspection, and decision making processes, and (4) the work suggests how reliability analysis results can be integrated with bridge field inspection results.
Description
Keywords
Citation