Modelling the behavior of dapped-end connections with fiber-reinforced concrete

Abstract

This Master Thesis is aimed at modelling the behavior of fiber-reinforced concrete (FRC) dapped- end connections. These connections are subjected to high stress concentrations due to their abrupt change of cross-section and it usually leads to sudden collapse. The use of FRC increases the duc- tility of the material and allows to prevent the rupture. The kinematic model takes into account the serviceability of the dapped-end and is not only a tool for design at ultimate limit state. In reality, this method assesses the structural state of the dapped-end by collecting data in situ. The studied model applies to one failure mode, a flexural failure governed by an inclined crack starting from the re-entrant corner. This mode of failure is the most frequent for experimental struc- tures and more frequent when the amount of fibers added in the concrete is high. The model compares its predictions to experimental data collecting from different scientific papers. These data have been analyzed on the most accurate way as possible to confirm the validity of the results. At first, the kinematic model is developed by a simplified approach to predict the peak capacity of a dapped-end. Globally, this approach is quite similar to the one developed by Rajapakse et al. ap- plied to reinforced-concrete structures. The predicted results with FRC compared to the experimental results are promising. Then, the complete behavior of the dapped-end connections is modelled. This approach aims at not only reaching the real peak capacity but also at modelling the main inclined crack for any loading condition. Some parameters have been calibrated for modelling the experimental results. Kinematic effects taken into account in the model allows to the model to tend to what occurs in reality. Indeed, the predicted results obtained with the model agree quite well with the experimental results for any loading condition. The model works well with the experimental data and the kinematics of dapped-end connection is taken into account. For all these reasons, this model gets a good perspective to be more developed in the future by including factors or agents that contribute to the serviceability of the dapped-end such as water or salt that are the first causes of damage of these structures.