Proceedings from ASCE’s Structures Congress 2015
April 25, 2015
Instrumentation of bridges, through targeted short term test applications to long term health monitoring applications, has become more widespread and accepted as means for supporting traditional engineering where current practice fails to provide a clear path forward. While decreased costs associated with data acquisition and sensor hardware are supporting applications with large arrays of sensor nodes, there are still significant costs associated with data reduction, processing, and interpretation of large datasets.
Engineers involved with monitoring applications must be cognizant of this fact, and streamline system designs to focus a fixed quantity of manageable sensors to the areas of greatest uncertainty in terms of structural performance. This paper introduces a long span bridge in the United States which was found to have unsatisfactory ratings for structural elements at different locations along the main load carrying path. The structure contains many internal movement mechanisms, as it is a cantilevered truss with intermittent simply supported trusses. A load test was carried out using up to six fully loaded tri-axle dump trucks.
During the processing and interpretation of results, it was clear the structure was performing in a non-linear fashion, based upon the location of the load. Through a detailed finite element model calibration effort, using state of the art optimization techniques, it was found that depending on the location of load on the structure, the primary load carrying path of the truss changed, and assumptions regarding the performance of movement mechanisms were modified.
Through the detailed analysis, the engineer of record was able to provide updated ratings and recommendations to the owner of the structure, whose aim is to indefinitely preserve the structure due to its critical function within the regional transportation network.
A main contribution of this presentation is the important consideration to implementers of monitoring techniques to consider that such applications must be able to defend major assumptions placed on structural performance, such as them being linear and stationery.