Optimized finite element analysis and strengthening assessment of the I-39 Kishwaukee bridge utilizing proof load testing

Many countries worldwide face a common problem with the aging bridge infrastructure that is being demanded to carry increasing loads. With the cost and the difficulties associated with replacing and rehabilitating these bridges, it is necessary to make the most efficient use of the existing infrastructure. Proof load testing (PLT) proved to be a reliable non-destructive method to assess the bridge and reflect its actual behavior, especially the old bridges. The advancement on the Internet of Things (IoT) technology concerning sensors and data acquisition systems for sensing, collecting, and storing the data in conjunction with Finite Element Modeling has resulted in combining analytical models and field test results for better assessment of the bridge condition. It would be insightful to combine the field-testing data with Finite Element Modeling to optimize the outcomes from proof load tests. In this paper, the case of the I-39 Kishwaukee, a five-span twin post-tensioned segmental concrete box girder bridge, has been studied. Kishwaukee bridge was built in 1970. Several retrofits were carried out on the structure to solve the cracks and slippage at the shear key between the pier segment and the adjacent cantilever segment caused by non-hardened epoxy at the joint during the time of construction. In 2006, The Illinois Department of Transportation investigates the structural behavior of the bridge and determined that the crack growth along the webs is caused by principal tensile stresses higher than the code limits. Using the FEA, the model shows an estimated permanent strain of 85 mu e caused by the dead load only at the shear key. This strain added to the strain caused by the HS20 truck live load led to a total strain of 180 mu e higher than the strain corresponding to the modulus of rupture of the concrete (135 mu e). A total estimated deflection of 3.84 in. at midspan of Span #3 caused by HS20 truck live load exceeded the AASHTO allowable limit for deflection (3.74 in). Since the bridge was deficient, IDOT decided to strengthen the structure using four-12 strands, 15 mm external post-tensioning tendons placed inside the box girders to reduce the shear forces acting across the webs. This paper illustrates a proof of four different trucks loading weights of 76 tons (167 k), 90 tons (200 k), 122 tons (268 k), and 136 tons (300 k) conducted on the bridge. Nine testing scenarios were successfully completed with a maximum of two testing trucks of approximately 136 tons (300 k). The data obtained from the field test (Measured strains near the pier, where shear and negative moment are critical, and at midspan, where the positive moment is crucial, and measured deflection profiles) were used to optimize a non-linear finite element model for the bridge. This paper provides a comprehensive guide on how to conduct load rating assessments based on the AASHTO MBE method for PLT. It outlines a step-by-step procedure for conducting field operations, implementing instrumentation, and interpreting test results. The data obtained from the field test are used to develop a Finite Element Model showing the impact of the recently introduced external post-tensioning tendons on the structural performance of the bridge. In conjunction with the FEA, this research demonstrated that the rehabilitation of Kishwaukee I-39 bridge using the post-tensioning system reduced the deflection by 88.72%, and minimized the principal tensile strain of the shear key by 80 mu e. Based on these findings, this paper provided a significant allowance for accommodating future traffic load increases on the Kishwaukee I-39 River bridge.