PSI - Issue 44

Ernesto Grande et al. / Procedia Structural Integrity 44 (2023) 582–589 Grande et al. / Structural Integrity Procedia 00 (2022) 000–000

583

2

1. Introduction The observed damage in the aftermath of the most recent earthquakes highlighted the high vulnerability of existing reinforced concrete (RC) buildings designed only for gravity loads and without proper seismic details. In particular, brittle failure mechanisms of RC structures built in 70-80’s, under moderate to severe earthquake demands, demonstrated the involvement of the exterior beam-column joints because of some deficiencies in the structural detailing such as lack of transverse reinforcement in the joint core, inadequate anchorage of longitudinal steel bars crossing the joint, poor quality of the materials employed at the time of construction. Indeed, unreinforced joints, especially the exterior ones or those belonging to the façade frames, were, in many cases, responsible for the premature shear failure of the old-type RC frame structures designed before the introduction of modern seismic codes. In the recent years, the research community has focused the attention on analyzing the influence of the poorly detailed beam-column connections on the seismic performance of the RC frames. On the basis of different numerical modelling approaches, several studies aimed at predicting the joint shear strength and deformability through the calibration of shear stress-strain relationships (De Risi 2016, Hwang and Lee 1999, Kim and LaFave 2009). On the other hand, only few studies focused on the modelling of the joint hysteretic behavior under cyclic loadings (Lowes and Altoontash 2003); therefore, the marked pinching effect, in conjunction with the strength and stiffness degradation of the beam-column joints, are still challenging aspects in the numerical modelling procedures. The effect of the joint shear behavior on the global capacity of RC existing frames is deeply examined in the work of Del Vecchio et al. (2016), where the authors compared the results derived from an ad-hoc modelling strategy of exterior joints with those obtained from experimental tests; then, the numerical approach was applied to a building frame damaged at the exterior joints during the L’Aquila earthquake and, finally, the effect of an FRP strengthening on the global performance of the frame was also investigated. The study by De Risi et al. (2016) focused on the definition of a joint macro-model by employing a set of experimental tests on unreinforced exterior joints to calibrate a shear stress-strain relationship. A numerical investigation on the influence of joints on the seismic behavior of a RC frame was assessed and nonlinear dynamic analyses were performed by considering two different models, the former that explicitly includes the joint nonlinear modelling and the latter with joints modelled as elements with infinite strength and stiffness. Jeon et al. (2015) carried out dynamic analyses on a prototype non-ductile concrete frame, comparing the results of different frame modelling mechanisms: a) rigid beam-column joint; b) nonlinear joint shear response; c) nonlinear joint shear and bond-slip response; d) column shear failure. Fragility curves were then plotted to quantify the impact of different failure mechanisms modelling on the frame vulnerability. The aim of this paper is to numerically investigate the key role of the beam-column joints in the seismic assessment of existing RC frames designed and built without the principles of the capacity design approach. In particular, the study starts from the outcomes presented in Grande et al. (2021b), where a new macro-model of exterior beam-column joints is proposed following a calibration process of the joint shear stress-strain relationship. The joint model is, then, implemented in the numerical modelling of a RC frame selected from the study by Del Vecchio et al. (2016). The results of the nonlinear analyses are discussed and compared with those obtained from a modelling approach in which the nonlinear behavior of the joints is neglected and only rigid elements are introduced within the joint regions. 2. Exterior beam-column joints modelling The macro-model proposed by the Authors (Grande et al. 2021a) to reproduce the monotonic and cyclic behavior of reinforced concrete beam-column joints is employed herein to evaluate the seismic performances of a 2D frame. The model, briefly summarized in the following, was developed to evaluate the seismic performance of exterior beam column joints belonging to existing RC structures. As depicted in Figure 1a, the proposed macro-model is based on the so-called “scissors model” where two nonlinear rotational springs are introduced in series to simulate the two main mechanisms governing the behavior of exterior joints, namely the shear deformation of the panel zone and the “fixed-end-rotation” due to the bar-slip phenomenon of the longitudinal steel bars of the beams anchored into the joint. The model is implemented in the open-source finite element code “OpenSees” (McKenna et al. 2010), where the uniaxial material model “Pinching4” has been selected for the rotational spring of the scissors model. The parameters characterizing the multilinear backbone curve were derived by the Authors in a previous work (Grande et al. 2021b) in order to account for the following events generally