Slab-on-girder prestressed concrete bridges: linear and nonlinear finite element analysis and experimental load tests

“A diaphragm is a transverse stiffener, which is placed between girders in order to maintain section geometry”, (AASHTO, 2002). Intermediate diaphragms, usually placed at the midspan or third points of a bridge, are thought to contribute to the overall distribution of live loads in bridges. Cast-in-place concrete intermediate diaphragms were investigated in this study in order to assess their load distribution effectiveness in prestressed concrete I-girder bridges. Finite element packages (Ansys and GT STRUDL) were used to perform the analyses for multiple bridge configurations, including a parametric study of span length, girder spacing, and concrete strength. It was found that intermediate diaphragms do not contribute significantly to live load distribution and that they are not needed if there is no possibility of impact by an over-height truck. If the risk of a lateral collision to the girders is present, intermediate diaphragms should be placed over the respective traffic lane so that the potential impact takes place at the diaphragms location. Temporary steel diaphragms can be used to stabilize girders during construction. Static and Dynamic live load tests were performed and experimental results were compared to the finite element analysis. Various pavement roughnesses were simulated with the use of wood boards up to 1.5” in depth. A dump truck, driven at 30 MPH, 38.5 MPH and 40 MPH, produced strains and deflections which were processed for load distribution and dynamic allowance (IM) factors. According to the results obtained, there is a more uniform live load distribution as the speed and roughness increase. However, driving at 40 MPH over the 1.5” wood board produced an impact factor almost twice as large as the one specified by AASHTO LRFD (2004). Finally, nonlinear analyses of the tested bridge were performed in order to predict its ultimate capacity. Two bridge configurations were analyzed. The first one was modeled without any diaphragms or edge stiffeners, while the second one with end diaphragms and edge stiffeners. Ultimate loading results show that AASHTO LRFD presents conservative values.