GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 2.2 GNGTS 2023 glazing case studies consist of a 6 mm annealed (AN) monolithic panel; an aluminum mid-rise framing system and rubber gaskets are used for the dry connection between the two materials along the entire perimeter. Moreover, a laminated glass unit with a thickness of 6.76 mm (3 mm annealed glass + 0.76 mm Polyvinyl Butyral or “PVB” + 3 mm annealed glass) has been also considered keeping a constant clearance of 11 mm, that corresponds to the recommended value for practical building installations. All the analyzed configurations have been selected from an study conducted by O’Brien et al . (2012) aimed at developing empirical fragility curves giving fundamental indications related to the damage state involved in the seismic performance of a glazing system. For sake of clarity, the included experimental data have been derived from several empirical campaigns (Behr and Belarbi (1996), Behr (1998), and Memari et al. (2003)) carried out at the Pennsylvania State University and the University of Missouri. According to AAMA 501.6 (2009), the testing phase consists in performing ‘crescendo tests’ where a series of ramp-up intervals and constant in-plane drift amplitude intervals was applied to the panel until the achievement of glass fallout, considered as the ultimate failure. Finally, the final test purpose is the identification of the drift relative to an event when a glass fragment with an area larger than 645 mm 2 breaks away from the glass panel and falls out. According to Porter et al. (2007), there are several derivation methods in order to address fragility functions: empirical, analytical, and expert opinion. In particular, the first one is undoubtedly the most desirable although not always the most feasible due to the high costs in terms of material, suitable facilities, and time. Consequently, the purpose of this study is to provide a comparison of fragility curves obtained by numerical approaches and by the experimental data available in the literature. Of paramount importance is the fine-tuned modelling in Abaqus/Cae as shown in Figure 1. All glass configurations were characterized by the same glass size (hxb=1829x1524 mm). In particular, the three-dimensional solid model consists of a panel modelled with a perfectly linear elastic behaviour; while an elastoplastic mechanical type was chosen for the aluminium frame and the transparent interlayer of PVB. However, the simplified assumption on glass behaviour is monitored by controlling that the tensile strength has never been achieved. Finally, the stiffness of axial connectors used for reproducing gaskets performance in the plane of the glass panel was calculated as a function of typical mechanical parameters, mesh dimensions and different clearances.