ABSTRACT
The practical application of long term monitoring remains a challenge for many real world projects. When inspectors discovered cracking on numerous girders of an 8,500-foot long bridge in Louisiana, the Louisiana Department of Transportation and Development decided that it was necessary to develop a comprehensive testing and monitoring program. Soon after the cracks were discovered, the bridge owner initiated emergency live load testing and subsequent refined analysis to determine the effects of cracking on the structural integrity of the bridge.
During a regularly scheduled bridge inspection in March of 2012, cracks were discovered near the supports on many girders. Most of the cracks that were discovered were less the 1/16”; however, there were several cracks much larger, some of which exposed the longitudinal tendons. The cracks were found to have a flexural-shear type crack propagation/orientation, indicating that they were a result of a combination of load effects.
With the cracks suddenly developing in the girders approximately 20 years after the bridge was in service, the bridge owner initiated emergency actions to repair the cracks with a Carbon Fiber Reinforced Polymer (CFRP) wrap and epoxy injection. The emergency action plan also included live-load diagnostic testing and long-term monitoring program to try and determine the cause and effects of the cracking. The testing program involved performing diagnostic tests on the structure both before and after retrofit for comparison and to verify the effectiveness of the CFRP strengthening.
The live-load diagnostic testing focused on determining the effect of the cracks on the load carrying capacity and load distribution of the superstructure. The response data collected was used to calibrate a finite element model of the superstructure, which was in turn used to evaluate the load carrying capacity and the load distribution.
Analysis of live-load testing revealed that the cracks not only had a significant effect on shear capacity and also altered the load distribution of the bridge. However, the results ultimately confirmed that the live-load and dead lead stresses were not large enough to cause the cracking.
The monitoring program involved designing a system that would be capable of capturing long-term data on the superstructure as a function of temperature as well as live-load responses. To achieve this goal a network of Vibrating Wire sensors (known for their good long-term stability) was used for determining the buildup of stresses due to temperature over time, and conventional, mV output, sensors (which have faster response times) were used to capture live-load events from vehicular traffic. These two different types of sensors were combined into one data collection system such that readings could be synchronized allowing comparative analyses if required. This system was implemented initially on several continuous spans that contained some of the more severe cracks. A second system was also installed on several spans that displayed very little to no cracking so that a direct comparison could be made between a severely cracked sections and a relatively un-cracked section.
The results from the monitoring data suggested that the cracks were likely initiated or propagated as a result of a large thermal stress gradient in connection with the continuity diaphragm detail between continuous spans.
This ongoing comprehensive testing and monitoring program has allowed the bridge owner to obtain a quantification of actual bridge performance at different stages of cracking as well as to test and evaluate different rehabilitation strategies in real field conditions.
