This Partnerships for Innovation (PFI) project will address the serious problem of aging of the highway system infrastructure and the possibility of structurally deficient bridges in the United States today. The proposed solution to this problem involves a shift in the bridge design protocol to include baseline finite element modeling, continuous sensor-based monitoring and testing, and the intellectual post-processing of collected data to determine the structural health of bridges. The 1956 Interstate Highway Program expanded the U.S. highway system to over 500,000 bridges, but there was no monitoring or condition assessment included in the initial design and building effort. This proposal focuses on a condition assessment tool that integrates baseline finite element models and parameter estimation for model updating. The baseline model will include documentation of design approaches and methodology and is intended to facilitate the use of that baseline information over the bridge's life cycle to monitor condition and structural health. Thus the project would develop a process to consider how the, long term requirements such as future inspection, corrosion, and reconstruction impact the structural condition of the bridge in a meaningful way.
The research, if successful, will be applicable to a significant fraction of the national infrastructure in the US and beyond. The benefits from a successful project would be realized through improvements in safety, as well as improvements in the economics of bridge design, fabrication, and maintenance.
Partners include Tufts University (lead institution); University of New Hampshire; State of New Hampshire Department of Transportation; Fay, Spofford, & Thorndike Engineers; Geocomp Corporation (consultants on instrumentation for risk); Bridge Diagnostics, Inc.; and the U.S. Department of Transportation, Federal Highway Administration.
?" research team successfully developed instrumentation, testing, modeling, and data analysis protocol using the now highly instrumented Powder Mill Bridge (Vernon Avenue over the Ware River) in Barre, Massachusetts, Figure 1. Research results will assist bridge engineers and owners for analysis and design for long-term bridge conditions, not just evaluation for "opening day" that is more the focus of current design practice. This approach considers long term structural health monitoring in the initial design and will help bridge owners appropriate dwindling maintenance resources. Several partnerships were formed throughout this research project ranging from other universities to industry partners and even state organizations and local communities in order to have a successful project. The partnerships were an important component of the research project’s success. The project also successfully supported the educational development of several graduate students at Tufts University and the University of New Hampshire. By example, this research progress has also educated several future bridge engineers into the idea of understanding of true bridge behavior via structural health monitoring to aid in their design process and load rating. This research team performed three load tests on the Powder Mill Bridge since it opened in September of 2009. These tests included both static and dynamic components to look at the full spectrum of bridge testing. Measurements from these bridge tests will be available to bridge engineers and researchers for years to come as investigations into bridge behavior are examined. Additionally, long term bridge response measurements were collected under operational conditions, Figure 2. The data was used to examine the long term behavior of the bridge subjected to various truck loadings and environmental effects such as diurnal temperature changes. The measured data, Figure 3, from these load tests was also used to form a comprehensive comparison of bridge load rating techniques and ideas for use by bridge owners and load rating engineers. This comparison showed both the pros and cons of selecting different load rating methods and approaches, Figure 4. The load test data was also used to calibrate a finite element model, Figure 5, that has lived with the bridge for three years with great success and shows that, as expected, there has been little deterioration of the bridge in its first three years of service. The full benefit of having this model and instrumentation plan will not be seen for years to come. However, the systems are in place to be able to monitor and update the model as time progresses, Figure 6. The use of data for long term monitoring was investigated and it was found that the data acquisition systems are highly affected by diurnal temperature change. An in-depth investigation ensued to determine the root cause of this problem and a vast repository of data and knowledge has been created on this subject. Two shaker-dynamic dynamic tests were successfully completed on PMB during summers of 2010 and 2011. This is the first time the research team has performed a dynamic test on a full-scale bridge using an electrodynamic vertical shaker with seismic accelerometers. The research team is currently using this data in frequency domain for FE model updating using frequency response functions. The NSF-PFI research team forged additional partnerships with other projects and groups. The team collaborated with: the Massachusetts Port Authority (now Massachusetts Department of Transportation) for the modeling, analysis, instrumentation and monitoring of the Tobin Memorial Bridge in Chelsea, MA, the FHWA Long Term Bridge Performance (LTBP) program for adding the measured data and findings of the Powder Mill Bridge to the US national inventory of bridges, the New Hampshire Department of Transportation for overpass bridge instrumentation and monitoring and real time monitoring, the University of Miami for supervision of one doctoral student, and several other partners for steel girder fabrication, steel erection, and construction. The Powder Mill Bridge in Barre, MA is instrumented with over 200 sensors. The information obtained from those sensors has helped to advance the field of structural health monitoring by showing how static and dynamic measured data can be used for bridge load rating and calibration of a finite element model. The Powder Mill Bridge is a seemingly non-descript bridge in a beautiful setting, crossing the Ware River at a wooded site next to a waterfall. Drivers crossing the new bridge may not be aware that the span is now one of the most densely instrumented bridges in the world. The research team has enjoyed conducting NSF-PFI supported research in this bridge. The data set collected to date, and in the near future, will help with development of SHM systems. The research work is leading the way towards developing approaches that will leverage improving technologies to more efficiently help address the challenges posed by our aging infrastructure.
