PSI - Issue 44

Sergio Ruggieri et al. / Procedia Structural Integrity 44 (2023) 1964–1971 Sergio Ruggieri et al./ Structural Integrity Procedia 00 (2022) 000–000

1968

5

For the set of generated mechanical models, structural and damage analysis is performed, deriving seismic fragility curves. With this regard, it is worth making some comments about the record selection, the analysis method, and the limit-state for which fragility curves are computed. Regarding record selection, it is recommended to use natural records of medium-high seismicity (e.g., Kohrangi and Vamvatsikos, 2016), instead of considering a specific target spectrum (in which the localization assumes high significance). For the analysis method, any method that provides a good distribution of the engineering demand parameter (EDP) vs. intensity measure (IM) can be employed, such as cloud, multi-stripe of incremental dynamic analysis. For the limit state to investigate, ultimate ones are of interest, assuming parameters of IM and EDP that can provide a reliable estimate of the building fragility, as well as a limit state threshold related to a specific failure criterion (e.g., failure of the first vertical element). The obtained result consists of a fuse of fragility curves, from which a mean value (and fractiles) can be computed and attributed to the building in the photo. In the described process, only data processed by Bi VULMA have been used. For improving the estimates, additional freely available data can be used (e.g., Census and Regional Technical maps), as previously discussed. In this way, the year of construction, area, and the localization of the building can be identified, and the fuse of fragility curves can be reduced. 5. Case study The case study presented in this paper consists of two existing RC buildings located in the municipality of Bisceglie, displayed in the pictures of Figure 2. Both images have been processed by Bi VULMA , which provided the information in Table 1. In Table 2, all values are shown, and, in addition, data regarding year of construction and indication of the base area are reported, extracted by the census database and CTR.

Fig. 2. Case study buildings, named B1 and B2, respectively. Black boxes are placed to cover the shop brands in the second photo.

Table 1 shows that VULMA provides results that the human eye can confirm. It is also interesting to highlight that some parameters, such as the total number of openings, could not be defined, since the VULMA dataset did not allow to identify this parameter. Using in the first phase the data provided by VULMA , the simulation and modelling campaign was performed, selecting the unknown parameters firstly and after automatizing the procedure of combination, modelling and analysis. All parameters (known and unknown) are reported in Table 2. The number of storeys, the base area and the aspect ratio are the parameters that have been varied in the models. Two values of concrete compressive strength, σ c , have been defined according to the minimum and maximum limits provided by the reference building codes. The same assumption has been made for the tensile strength of steel ( σ s ), but in this case, a unique value has been assumed, considering the greater standardization of this material and the lower variability that this parameter can provide to the input than other mechanical ones. Regarding the masonry infill parameters, “strong” infills have been considered in the modeling (in order to obtain a significant interaction), assuming a variability of the compression strength of masonry ( σ m ) given by actual experimental results (Uva et al., 2012 and Hak et al., 2012). The assumed service loads are those typical of residential buildings, and variability is not considered in this case.

Table 1. Features for the case study buildings and additional information from other sources