Comparison of analytical and numerical models for predicting the capacity of dapped-end connections

Abstract

Reinforced concrete (RC) dapped-end connections are widely used in precast concrete structures; however, their abrupt geometric discontinuity results in complex stress distributions and brittle failure mechanisms, making the accurate prediction of their load-carrying capacity a significant engineering challenge. This study compares the predictive performance of representative analytical and numerical models for estimating the ultimate capacity of RC dapped-end connections. Three analytical models, namely the classical Strut-and-Tie Model (STM) proposed by Schlaich and Schäfer, the Precast/Prestressed Concrete Institute (PCI) design method, and the Kinematics-Based Model (KBM), were evaluated using the experimental results of eight large-scale, orthogonally reinforced dapped-end specimens from a previous experimental programme. In addition, a nonlinear finite element model was developed in VecTor2 using the Disturbed Stress Field Model (DSFM) and validated against the experimental data. The predicted capacities obtained from each model were compared with the measured ultimate loads. The results showed that the KBM provided the most accurate analytical predictions, while the VecTor2 model also demonstrated excellent agreement with the experimental results. The findings highlight the effectiveness of mechanics-based analytica l and nonlinear numerical models for the reliable assessment of reinforced concrete dapped-end connections.