Abstract
This paper describes an experimental investigation into the inelastic cyclic performance of reinforced concrete members incorporating mechanical reinforcement splices. Based on a survey of available mechanical splicing forms, two types of threaded couplers with different geometric configurations, namely ‘parallel threaded couplers’ and ‘parallel threaded sleeve couplers’ are selected for detailed experimental assessment. The fundamental uniaxial monotonic and cyclic response of reinforcement bars connected with threaded mechanical splices is firstly examined through bare (in-air) and embedded (in-concrete) tests, and their response is compared with that of their non-spliced counterparts. This is followed by an experimental study on four large scale reinforced concrete specimens subjected to lateral inelastic cyclic displacements, which provides a direct comparison between the performance of members with or without the two types of reinforcement couplers as well with or without a co-existing axial load. The test results enable a direct comparative assessment of the key response characteristics of the specimens including stiffness, strength, ductility and energy dissipation. Overall, the results show that members incorporating reinforcement bars connected through threaded mechanical couplers can provide considerable ductility and energy dissipation. However, the coupler-concrete interaction behaviour, which depends primarily on the geometry and location of the threaded coupler, has a significant influence on the inelastic cyclic response and ductility levels exhibited by reinforced concrete members. The experimental results indicate that the presence of the slender coupler alters the plastic hinge behaviour by localising the curvatures between the coupler and interface, reducing the rotational capacity. In contrast, the response of members with compact couplers, and without axial load, is virtually identical to that of the specimen with continuous reinforcement. The axial load enhances the stiffness and lateral capacity yet reduces the ductility performance due to a response strongly characterised by crushing. It is shown that whilst uniaxial in-air reinforcement performance may be adopted as a basis for selecting suitable mechanical couplers for use in dissipative members, the expected levels of ductility implied in current codified seismic procedures need to be carefully and explicitly evaluated when relatively non-compact couplers are employed.