Experimental assessment and numerical modelling of exterior non-conforming beam-column joints with plain bars

The seismic performance of existing Reinforced Concrete (RC) frames is significantly influenced by the behaviour of beam-column intersections, especially in non-conforming buildings with poor structural details and completely unreinforced joints. In literature, a very limited number of studies deals with specimens reinforced with plain hook-ended longitudinal bars, widespread in Italian and Mediterranean building stock, or with the analysis of local aspects, such as the experimental evaluation of joint shear strains. The almost totality of the models proposed in literature for simulating the cyclic behaviour of RC joints was developed and calibrated by means of tests performed on elements with deformed bars, and, thus, these models may be not adequate for elements with hook-ended plain bars, especially due to the peculiarities in terms of failure mode and interaction mechanisms between concrete and steel. This study analyses the experimental cyclic behaviour of four full-scale exterior unreinforced RC beam-column joints with plain reinforcing bars in beams and columns, which differ for joint aspect ratio and beam longitudinal reinforcement ratio. First, experimental global and local responses of such tests – including energy dissipation capacity and observed damage evolution – are analysed. The main deformation mechanisms ascribable to the joint – namely rotation at the interface between beam/columns and joint, and shear deformation of the joint panel – are experimentally evaluated to provide a realistic support for the numerical modelling of this typology of beam-column joints. Then main joint shear strength models existing in codes and literature are compared with the experimental results. Finally, the numerical modelling of the specimens is carried out under monotonic loading to reproduce the envelope of the experimental responses. The joint panel constitutive parameters are defined to reproduce the experimental joint shear stress-strain relationships. Additionally, modelling of bond-slip is particularly taken into account due to the poor quality of steel-concrete interaction in the specific case of plain reinforcing bars