Behaviour and Modelling of Reinforced Concrete Dapped-End Connections

Reinforced concrete dapped-end connections are commonly used in bridges, parking garages, industrial buildings, and other precast concrete infrastructure. Due to their characteristic shape, they typically feature an inclined corner crack at service loads arising from high stress concentrations in the re-entrant corner. At the same time, these connections transfer high shear forces through significantly reduced sections and are susceptible to brittle failures. For these reasons and major durability concerns stemming from penetration of corrosive agents, the modelling and monitoring of dapped-end connections remain a challenging problem. To deepen the understanding of this critical connection, this thesis provides analytical and experimental research on the behaviour and modelling of dapped-end connections.
The thesis first presents a novel kinematics-based model to predict the strength of dapped-end connections failing along a re-entrant corner crack. The modelling framework is derived based on first principles - kinematics, equilibrium, and constitutive relationships – and facilitates the direct use of on-site measurable data to assess the strength of the connection. The model utilizes measured angles of the inclined crack as an input, and explicitly accounts for kinematic parameters such as the width and the length of the corner crack. Upon validating using a database of 47 tests from the literature, it is shown that the model captures well the peak resistance of dapped-end connections governed by the opening of the re-entrant corner crack, leading to an average experimental-to-predicted strength ratio of 1.10 and a coefficient of variation of 8.6%.

In order to investigate the main failure modes of dapped-end connections, the thesis then presents experimental research on eight specimens. These specimens, which are among the largest in the literature, featured the orthogonal reinforcement layout. The main test variables are the amount of dapped-end reinforcement and the ratio of horizontal to vertical reinforcement area. The reinforcement amount is gradually increased to capture both flexural and shear failures of dapped-end connections. It was found that, for the same total area of dapped end reinforcement, connections featuring more horizontal reinforcement are slightly stronger. While the orthogonal layout did not provide sufficient crack control under service loads, all tests exhibited substantial rotation capacity, including the shear critical connections.

Further experimental research on eight companion specimens featuring the diagonal reinforcement layout is then presented. The two sets of eight specimens had almost identical test parameters except for the reinforcement layout, allowing for direct comparisons. Crack displacements, crack patterns and the elongation of main dapped-end reinforcement are reported in detail via 56 continuous measurements of deformations. Comparisons of all sixteen specimens showed that, for the same amount of total dapped-end reinforcement, specimens with diagonal reinforcement are considerably stronger than the corresponding connections with orthogonal reinforcement layout. It was found that under both reinforcement layouts, the corner crack widths were in excess of typical code provisions under service conditions.

Utilizing the observations and measurements of the experimental campaign, the kinematics-based model is further extended to predict the complete behaviour of dapped-end connections failing along a re-entrant corner crack. The model captures the pre-peak, peak and post-peak response, and explicitly predicts the deformations of the re-entrant corner crack under service conditions. The modelling assumptions are discussed in detail, and the crack width predictions are extensively validated using 24 tests from the literature. The results show good agreement between measured and predicted crack widths and peak capacity of dapped-end connections. As crack width predictions can be directly compared with on-site measurements, this offers a suitable framework for crack-based assessment of existing connections.