PSI - Issue 44

Michele Angiolilli et al. / Procedia Structural Integrity 44 (2023) 870–877 M. Angiolilli et al./ Structural Integrity Procedia 00 (2022) 000 – 000

871

2

other earthquake-prone countries, proved that a considerable number of existing RC structures were unable to withstand earthquake forces. This was mainly owing to a lack of capacity design principles as well as the structural degradation due to the use of original materials featuring low mechanical characteristics, such as low-strength concrete and low-adherence smooth bars, that are often combined with poor reinforcement detailing, such as insufficient column longitudinal reinforcement, lack of transverse reinforcement in joint regions and inadequate anchorage detailing. Furthermore, critical joint failure can also be caused by an unbalanced thrust of the masonry infill and a higher displacement demand caused by global torsional effects. Some of the construction details can be identified as possible essential causes of brittle failure mechanisms, which diminish the overall structure ductility and result in insufficient lateral strength. Different failure modes for beam column joints have been recognized during recent earthquakes, namely the joint panel shear failure, which is characteristic of a weak-column/strong-beam combination, and the bond-slip failure mode, which is dependent on bond properties. In general, force distribution within joints generates diagonal cracking, resulting in a reduction in joint strength and stiffness (Ricci et al. 2011). Diagonal stress caused by parts intersecting in the joint cannot exceed concrete compressive stress. When concrete cracks, the existence of transverse reinforcement permits loads to be transmitted via a strut and tie mechanism. This mechanism can be developed if longitudinal/ transverse reinforcement and concrete struts contribute to truss formation. Preventing brittle failure in joints allows for the development of more ductile mechanisms in other structural parts if capacity design prescriptions are followed. Post-earthquake reconnaissance (e.g. Ricci et al. (2011)), experimental studies (e.g. Beschi et al. (205), Hakuto et al. (2000), Hassan and Moehle (2012), Masi et al. (2013), Murad et al. (2020), Sharma et al. (2013)), and numerical analyses (see §4) have all shown that existing RC joints constructed according to old standards/construction practices often collapse prematurely. Furthermore, they have confirmed the experimental evidence in which key parameters influencing the joint shear strength include concrete compressive strength, joint aspect ratio, joint width, column axial load and joint transverse reinforcement. Many tests were carried out on sub-assemblies with interior or exterior beam column joints characterized by typical details of Italian construction practice in the 1960s-70s, namely using smooth bars with hooked-end anchorages and poor reinforcing details. Most studies considered ribbed bars bent in the joint while few tests focused on sub-assemblies with hooked-end smooth bars and only some of them were correctly designed to develop a joint shear failure. Among them, Calvi et al. (2002) demonstrated the brittleness of this type of external beam-column joint through tests on a 2:3 scaled RC frame, showing that the development of a shear failure mechanism distinct from that provided in the case of a rigid joint behaviour, for which a soft floor mechanism would be expected. The significance of adequately designing the test specimen reflects the likelihood of capturing shear failure in the joints. Indeed, despite the lack of transverse reinforcement in the panel region, some poorly detailed sub assemblies showed flexural hinges in some experiments (e.g. Russo and Pauletta (2012) and Masi et al. (2013)), owing to the low reinforcement ratio used in the beam, which reduced the shear demand in the joint panel. The shear transfer mechanism in exterior beam-column joints with smooth bars and without transverse reinforcement in the joint panel is based on a compression strut mechanism, whose efficiency depends on the concrete strength and the anchorage adopted for longitudinal beam reinforcement. If hooked anchorages were adopted, the joint strength would be impaired by the expulsion of a concrete wedge, due to the pushing action of the hooked-end anchorages in compression and caused by bar slip within the panel region (Calvi et al. (2002). Different available analytical models can be found in the current literature, including those developed solely based on experimental evidence and those developed based on equilibrium and congruence considerations and then modified by using simplifying assumptions derived from experimental evidence. Murad et al. (2020), Nicoletti et al. (2022) describe some of the more well-known models available in the literature. In this article, the attention is focused on the seismic behaviour of the exterior (façade) beam-column joint, characterized by the lack of stirrups in the joint panel. Moreover, beams and columns were adequately designed to remain in the elastic field during the tests. Hence, both the specimen design strategy and test setup were conceived to emphasize the vulnerability of the joints and, therefore, to achieve pure shear failure in the joint panel without any plastic hinges forming in beams or columns. The main goal is to understand whether the weakness of the joints affects the overall behaviour of the sub-assemblies as well as if the models present in the current literature can predict correctly the effective performance of the joints. A numerical 3D model was also developed to facilitate the design phase. The validation of the model to the desired behaviour under monotonic loading compared to the cyclic behaviour of the experimental test is then reported.