Modeling of fatigue of dapped-end connections in existing bridges

Abstract

Due to the ease of assembly and the fact that they are statically determined, reinforced dapped-end connections are often used in precast concrete structures such as bridges. The main characteristic of this connection is the reduction of section at the two extremities of the beam. This reduction creates a re-entrant corner that is subjected to high shear forces. The flexural failure mode is characterized by an inclined crack that is generated from the re-entrant corner. Given that bridges are generally subjected to cyclic loading, the aim of this thesis is to focus on the effects of fatigue on the flexural behavior of dapped-end connections.

The fatigue of the dapped-end connection gathers the fatigue of three components: concrete, steel and bond stresses between steel and concrete. The latter is the main component studied in this thesis. Cyclic loading impacts the bond by creating an initial slip in the pullout behavior of the bond and by modifying the stiffness. The work consists thus in introducing these effects in the model implemented by [Rajapakse et al., 2023]. This model aims to represent the flexural behavior of this type of connection and the main input is the inclination of the crack.

The model used by [Rajapakse et al., 2023] to represent the bond is replaced by a finite element model in order to capture the variation of the stresses along the bar. Then, the monotonic and fatigue force-displacement responses of a bar subjected to pullout loading are modeled thanks to two bond-slip diagrams (splitting and pullout failure). Results show an increase in the displacements due to fatigue for a given load. The force-crack width response of the connection is then obtained by combining the bar responses of the half-bars on each side of the crack.

The finite element model is then validated at the dapped-end level by comparing the theoretical responses to the experiments conducted by [Rajapakse et al., 2023]. It appears that a bond strength reduction factor was needed to fit the experimental data. Then, the fatigue analysis is carried out for all the specimens of the experiments. From this study, it is shown that the crack width increases with fatigue but that the resistance of the connection was not impacted.

This study allows for providing some first results about the effects of fatigue on dapped-end connections. However, the fatigue of steel and concrete has to be introduced in order to complete the results and the lack of experimental data does not enable to validate the results. Besides, the models used are valid in a range that may not be suitable for bridges. For all those reasons further studies need to be conducted on this topic.