GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 2.2 GNGTS 2023 According to previous studies (Mattei et al. , 2021), the selected parameter as IM is the spectral acceleration at the fundamental period T 1 (S a ) that provides a closer approximation to the structural response of case-studies than the absolute estimation of the peak ground acceleration (PGA). On the contrary, the identification of EDP is slightly more complex due to the poor knowledge of the seismic performance of glass non-structural components, which are considered only as an added contribution in terms of masses but without any specific definition of damage limitations at any limit state. In this paper, the considered EDP is the maximum inter-story drift that has been demonstrated to be well correlated with key damage levels for secondary elements. According to Section 7.3. of the Italian seismic code, Norme Tecniche per le Costruzioni (2018), the check of structural elements in terms of damage control to non-structural elements states that the limiting IDR values must not exceed 0.002 times the story height, in the most conservative case. Since Cloud analysis is strictly dependent on the distribution of the selected suite of ground motions, the 40 unscaled records were identified in order to cover uniformly a wider range of potential seismic demand. It can be noted that the conducted nonlinear time history analyses lead to the blue curves in Figure 2 by considering an EDP threshold of 0.004m. The analytical and experimental curves were derived from the available data in the literature, in functions of θ and β , for the ‘fallout’ damage state. In particular, the analytical derivation (Nuzzo et. al , 2020) was based on the commonly used model proposed by Bouwkamp and Meehan (1960) according to which the behavior of a glass panel under racking loads follows specific steps. By the comparison between the fragility curves obtained by numerical analyses with the ones derived based on the experimental and analytical data obtained in O’Brien et al. (2012) and Nuzzo et al. (2020), respectively, it is possible to observe that the simulations always generate higher probabilities of failure by taking into account the code provisions in terms of prescribed drift for serviceability limit state. This observation is even more evident in the case the clearance increases. Therefore, this parameter greatly affects the seismic performance of the glass panel considering the geometry unchanged (i.e., aspect ratio, width). It is worth to mention that, in accordance with the ATC-58 Project, O’Brien et al . (2012) employed “Fragility Function Calculator” (version 1.02) software by Porter et al. (2007) to compute the fragility parameters, θ and β , based on the results from the performed “crescendo tests”. Thus, of significant interest is the possibility to adopt the median ( ϑ exp ) from the empirical method as the actual value of the parameter, chosen as EDP, corresponding to the mean of drift obtained by experimental tests for each configuration (i. e., Figure 2(e), (f), (g) and (h )) . In this case, an important consideration can be made by analysing the numerical random dispersion values computed by Eq. 3: for C#1 β num = 0.505; for C#2 β num = 0.421; for C#3 β num = 0.402; for C#4 β num = 0.216. T he β values are very similar to those obtained experimentally: for C#1 β exp = 0.359; for C#2 β exp = 0.262; for C#3 β exp = 0.315; for C#4 β exp = 0.311 , and by analytical approach β anal = 0.4. Finally, two small-scale configurations from the literature have been modelled in dynamic finite element simulations of a four-story building (full-scale application) with the aim of considering the actual differences in terms of seismic performance and providing some realistic reference