PSI - Issue 44

Giorgia Cianchino et al. / Procedia Structural Integrity 44 (2023) 219–226 Giorgia Cianchino et al. / Structural Integrity Procedia 00 (2022) 000–000

224

6

Figure 3. Comparison between capacity and demand with identification of Performance Point (PP).

Damage thresholds were first identified on the representative capacity curve with reference to the worst mechanism of the archetype building (in blue). Different shades of gray are used to identify different damage thresholds from light damage (light gray) to total collapse (dark gray). PP was then defined by comparing the capacity curve against the representative demand spectrum of a seismic event with PGA of 3,18 m/s2. As highlighted in the Figure, the PP falls within the range of moderate damage. 4. Fragility curves Following the described methodology, the corresponding Damage Level, for each archetype, was evaluated for each seismic input. This procedure accounts for the record-to-record variability as the main source of uncertainty. Then by estimating the frequency at which damage is activated for each family and for each archetype, Damage Probability Matrices (DPM) were defined. By accumulating these frequencies, a points cloud was obtained. These points were intercepted by a lognormal curve through a cumulative lognormal distribution, implemented in ®MATLAB as a function of 2 parameters: the median value of the intensity measure that induces a damage equal or greater than Dk and the dispersion β . The first parameter defines the median peak ground acceleration values (PGA) related to the 5 damage levels, determining the trend of the fragility curve. The second, on the other hand, indicates the statistical dispersion, i.e. the greater or minor slope of the curve. Regarding this last parameter β , it should be noted that the fitting procedure leads to different dispersion values among damage levels. This might cause, in turn, the intersection of the curves, which is not acceptable since the probability of exceeding a given damage level (for example D3) cannot be higher than the probability to exceed a lower damage level (for example D2) (Lagomarsino et al., 2021). For this purpose, the beta value was standardized by taking the dispersion of the D2 damage curve as the reference value. This is considered to be representative of significant damage, since exceeding the D2 curves results into the activation of the kinematic mechanism. Fig.4 shows the analytical fragility curves of the six analyzed archetype buildings displayed by type of vertical load-bearing structure and by increasing number of stories. 5. Conclusion This paper dealt with the seismic vulnerability assessment at regional scale, evaluated through analytical analysis carried out on the most fragile macro-elements of representative archetype buildings, which were derived from the CARTIS database. The Abruzzi Region has been considered as case study. Fragility curves have been found and the main conclusions of the study can be summarized as follows: • The kinematic analysis showed that the mechanism associated with the lowest acceleration is the complete overturning of the façade. This is because there are no elements for restraining possible out-of-plane mechanisms in the analyzed archetypes. • The most vulnerable buildings are those made of brick masonry. In fact, such buildings are likely to suffer more severe damage for the same acceleration level as compared to stonemasonry buildings. This is due to the reduced