Infrastructure Technology Institute
There are over 597,000 publicly owned bridges in the United States, but resources available to maintain them are limited. University Transportation Center researchers can provide a better understanding of how structures respond to real world conditions by applying structural health monitoring techniques to selected in-service bridges. Engineers and policy makers can then use this additional information to make more cost effective choices, benefiting the entire transportation system.
In July of 2009 the ITI Research Engineering Group began a multi-year cooperative research project in partnership with the Wisconsin Department of Transportation (WisDOT) to deploy and evaluate advanced structural health monitoring techniques on a typical highway bridge. The target of the study is a five-girder continuous-span steel bridge with a concrete deck which carries US Highway 2 over the Montreal River between Hurley, Wisconsin, and Ironwood, Michigan. The bridge design is typical of many small river crossings throughout the country. However, this bridge is of special interest because it is regularly subjected to truck loads well in excess of its posted 40 ton limit due to nearby logging operations. WisDOT contracted with a commercial supplier to install a state of the art real time weigh-in-motion (WIM) system adjacent to the bridge to measure vehicle loads and as a supplement to enforcement efforts. This presented a unique opportunity for research collaboration, since the actual vehicle loads on instrumented structures are rarely available for comparison with instrument readings.
During the second half of 2009, REG engineers, along with Northwestern University graduate and undergraduate students, designed, constructed, and installed a comprehensive automated, structural health monitoring system on the Hurley bridge. The system consists of 14 strain gages, four accelerometers, two displacement transducers, and six environmental sensors applied at critical locations on the structure. The sensors feed into a computerized data logger located in an enclosure at the end of the bridge, along with supporting communications and electronic equipment. Real time data, including weigh-in-motion data, are automatically transmitted back to ITI servers, where they are processed and made available in near real time on a password-protected project Web site. In addition to making data available to WisDOT personnel, ITI researchers and Northwestern students are analyzing bridge performance using time series statistical models and comparing measured strain values with those predicted by finite element models. The comprehensive set of load-response measurements gathered during this multi-year project will be a unique resource for bridge engineering studies.
Northwestern civil engineering undergraduate Ken Fuller, an ITI student employee who actively participated in the installation of the remote monitoring system at Hurley, is working on an independent study project with ITI researcher Professor David Corr to investigate the effects of truck loads on the structure. Specifically, Fuller is investigating the correlation between input measures (such as vehicle speed, axle configuration, gross weight, and axle weight), and outcome measures, such as the observed strains, accelerations, and deflections of the bridge and its components.
As a member of the installation team, Fuller is able to combine skills learned in his classes, his field experience, the data gathered autonomously by ITI’s structural health monitoring system, the traffic and weight data from the weigh-in-motion-system, industry standard structural modeling software, and Professor Corr’s experience to understand the behavior of this simple structure in incredible detail. Because the design of the Hurley bridge is typical of many small river bridges throughout the country, Fuller’s findings may have far-reaching impact on the understanding of thousands of other structures.
As the REG collects data from the Hurley bridge, Professors Pablo Durango-Cohen, David Corr and graduate student Yikai Chen are developing a time-series statistical framework for analyzing that data. Because the response of the bridge is influenced by external factors such as temperature, it is helpful to use advanced statistical tools to help reveal an underlying trend that those external factors may be obscuring.
Thus far, statistical analysis has shown that the bridge is gradually moving sideways (perpendicular to the flow of traffic) by tiny fractions of an inch under vibrations from traffic. This is consistent with anecdotal evidence and field observations. While the sideways movement does not currently have any implications for management of the bridge, it validates the analysis methods, and provides confidence that any other trends that may occur will also be detected.
The discovery of trends in the data may help inform future data acquisition and instrumentation plans, while the data acquisition and instrumentation systems provide a concrete link between the analysis and a real-world physical system. This collaborative effort in data acquisition and statistical analysis is representative of the cross-linking between research areas that is common among the work of ITI and its researchers.