PSI - Issue 44

Laura Gioiella et al. / Procedia Structural Integrity 44 (2023) 1808–1815 Laura Gioiella et al. / Structural Integrity Procedia 00 (2022) 000–000

1811

4

damage detected for each school with two surveys and related forms is the maximum one between the filled forms. For what concerns schools that suffered damages after August, without being collapsed ( d ≠1 ) and that have been subjected to further inspections after October, but without correctly compiling the section of the AeDES form, there is no chance to evaluate the damage growth. Therefore, such schools are associated to the damage detected in August and the related intensity measure.

Fig. 2. Experienced PGA range during the 2016 seismic sequence with the location of both the schools and the epicentres of the four main events.

From the evaluation of the 658 AeDES forms acquired, no collapses have been detected among the analysed schools ( d <1), nor in concrete, neither in masonry. The schools that resulted undamaged and those that have not been inspected (AeDES form not compiled), have been associated with a null damage ( d =0). The damage index, for each school building, d j , was deduced based on the contents of "section 4 - damage to structural elements" of the AeDES form. For the evaluation of the index five structural components have been considered, that is vertical structures, floors, stairs, infills and partition walls, excluding any pre-existing damage. The procedure followed is known in literature, De Martino et al. (2017) and is based first on an estimate of the damage, ̅ suffered by the j- th structural component among those listed previously. Once known ̅ , it is possible to determine the damage index d j for the building analysed according to d j = 1 5 ∑ d � j 5 j=1 = 1 5 ∑ � ∑ DL∙e k,D DD 5=D 0 5 � 5 j=1 d j ∈ [ 0,1 ] (1) where DL is the level of damage derived from the form, while e k,D is a coefficient ranging between 0 and 1 which considers the three damage extension levels (<1/3, between 1/3 and 2/3, >2/3). For further details regarding the values of the coefficients DL and e k,D refers to De Martino et al. (2017). 3. The empirical damage probabilistic model and its application to the school of the seismic crater 3.1. Probabilistic model and related parameters In the proposed model the damage is described by a real-valued random index D in the range [0,1]. The model relates the damage index to a seismic intensity measure described by a real-valued random variable I in the range (0,∞). The boundary values of the interval describing D represent, respectively, the case of school building that did not suffer damages ( d =0) and the case of collapsed buildings ( d =1).