Load Testing of a Deteriorated Prestressed Concrete Girder Bridge Without Plans
Sebastián Castellanos-Toro, Diana Millán, Albert R. Ortiz, Johannio Marulanda, Peter Thomson, Eva O.L. Lantsoght
Appears on pages(s):
dynamic characterization, load distribution factors, load testing, prestressed concrete girder bridge, strain measurement, visual inspection
In this study, a prestressed concrete girder bridge without plans and with severe levels of deterioration, located in Cali, Colombia, was load-tested to quantify, experimentally, its live-load behavior. The bridge consists of seven prestressed I-girders covered with a reinforced concrete deck, and four diaphragm beams. A geometric survey was performed to obtain the dimensions for a shell-based linear finite-element model (FEM) representing the bridge superstructure. In this survey, it was observed that the diaphragm beams in the span are geometrically inadequate to contribute to the structural system. Based on the experimental modal properties and the design regulations enforced at the time of bridge design and construction, a first update was made. Modifying the effective stiffness of selected elements to model girder deterioration, a second update was performed based on strain-gauge data from three load tests and visual inspection (VI) of the elements. The three models (basic, modal updated, and load-test/VI updated) were compared with the load distribution factor (DF) obtained from the load test and AASHTO distribution factor estimations. Visual inspection, dynamic characterization, and load testing response of this structure indicated severe deterioration of the girders and the absence of the effect of the diaphragms in the overall structural behavior. The results show that the AASHTO recommendations overestimate the LDF in comparison with the FEM without girder deterioration. When girder deterioration is included in the model, the LDFs change drastically, showing that AASHTO estimations are not in line with the experimental results. As such, for cases of bridges with severe levels of deterioration, it is recommended to use field data to estimate the distribution factors.