The adjacent box beam bridges (ABBB) are considered as an ideal solution for short to medium span bridges and for routes with low to medium traffic volumes. The ABBB system has been utilized and is popular in several states in the United States. However, this bridge system has long term durability issues caused by shear key failure and reflective cracking in the topping. The means and methods to alleviate the problems in connections between members of the ABBB were researched and the development of new connection details was pursued. Diagnostic tests to study the in-service behavior of ABBBs was performed. Two bridges with varying magnitude of joint deterioration were investigated. Both bridges were instrumented extensively and were subjected to known loads in the form of tandem trucks. The response of these bridges was studied and conclusions were made about the state of the bridges and the behavior after shear key failure. A finite element (FE) model of one of the tested bridges was developed to study the response of an ABBB with sound joints. The results of the finite element analysis (FEA) were compared with the results of the bridge diagnostic test. Conclusions about the FE model were made on the basis of this comparison. Another FE model, referred as the full scale bridge (FSB) was developed. The FSB model was used to simulate the behavior of an ABBB with the proposed connection details. This FSB model was subjected to design truck loads and the response was studied. The behavior of FSB model was replicated through a three beam sub-assembly that was supported on elastic supports. The stiffness of the elastic supports was calibrated such that the state of stress in the joints and the relative displacements between adjacent box beams in the sub-assemblage matched those in the FSB. The three beam sub-assembly was constructed with the proposed connection details. Two new connection details were proposed in this research. A Kevlar and epoxy connection and a spliced connection with fiber reinforced self-consolidating concrete are proposed. A total of six specimens, with different connection details, were constructed and tested for strength and durability in the laboratory. The behavior of the proposed connections and the connection materials were studied in detail. Additional FEA was performed to study the effect of shrinkage and temperature on the proposed connection details. / Ph. D.

Non-contact lap splices placed within a single concrete placement are often used and have been studied in previous research projects. However, non-contact lap splices used with each bar in a different concrete placement such that there is a cold joint between the bars, have not been investigated. This situation is found in the repair of adjacent box beam bridges and in the construction of inverted T-beam systems, among others. It is vital to understand whether the same mechanisms are present across a cold joint with two different types of concrete as are present in traditional non-contact lap splices. In this research, eight T-beam specimens with non-contact lap splices were tested. The spacing between the bars, the splice bar blockout length, and presence of transverse bars were varied to study the effectiveness of the splices. The beams were tested in four point bending so that the splice region was under constant moment and the tension forces in the spliced bars were constant. End and midspan deflections were measured along with surface strain measurements at midspan and at the quarter span points, top and bottom. Gap openings were also measured at the ends of the blockouts. The main conclusions found from this research are that beams containing non-contact lap splices were able to develop nominal capacity with the bar spacing less than or equal to 4 in. and the blockout between 17 and 20 in. long. Extending the blockouts and adding transverse bars underneath the splices did not add to the capacity. / Master of Science