Seismic behaviour of engineered cementitious composites beam-column joints

Reinforced concrete (RC) beam-column joints are one of the most important components that may lead to collapse or major damage of an RC frame structure when subject to earthquake loading. Structural engineers tend to pay attention to the detailing of the joints, which may lead to reinforcement congestion and thus making it difficult to be constructed. The emergence of engineered cementitious composites (ECC), which features ultra-ductility and damage-tolerance, offers a potential solution to the abovementioned problem. Therefore this study evaluates the feasibility of using ECC in the joint cores as a means to enhance the seismic behaviour of the RC beam-column sub-assemblages. Two series of experimental tests were conducted on interior and exterior RC beam-column sub-assemblages under lateral cyclic loading. In the first series, two RC interior beam-column sub-assemblages were detailed to seismic design while another two were detailed to non-seismic design. The control specimens had normal concrete in the joint core while ECC was incorporated into the joint core of the seismic and non-seismic designed sub-assemblages. Joint stirrups were totally eliminated from the joint core when ECC was used in the beam-column sub-assemblages, to study the performance of ECC joints without joint stirrups. On the other hand, four exterior beam-column sub-assemblages were adequately designed to seismic in the second series of test. The effects of joint stirrups and column axial load on the seismic behaviour of the exterior beam-column sub-assemblages were investigated. Comparisons on the hysteretic response, failure mode, analysis of inter-story drift contribution, distribution of beam and column longitudinal bars strain, joint shear stress and energy dissipation were made between sub-assemblages with joint cores either made of ECC or RC materials. In order to study the bond-slip behaviour of steel reinforcement embedded in ECC, a series of pull-out tests was conducted on short and long reinforcement. Thereafter an analytical model for bond-slip was proposed based on experimental results to predict the force-slip relationship of long reinforcement either anchored in concrete or in ECC. Moreover, a series of parametric study was conducted to estimate required embedment length of reinforcement in concrete and ECC according to four types of failures. Finally, a beam-column joint model available in OpenSees was employed in simulating seismic response of the beam-column sub-assemblages. Bond strength of steel reinforcement embedded in ECC obtained from pull-out test was incorporated in the bar-slip component for ECC joints. Results of simulation were discussed.Doctor of Philosophy (CEE