PSI - Issue 44

Marta Faravelli et al. / Procedia Structural Integrity 44 (2023) 107–114 Marta Faravelli et al. / Structural Integrity Procedia 00 (2022) 000–000

111

5

3.1. Static data

The platform allows to display, for a single selected school, some pre-loaded static data. These are the data derived from the exposure database and the seismic hazard at the site. After selecting a school, it is possible to view the data of the database mentioned in section 2 and summary information relevant for the assignment of structural type and fragility curve. Specifically: number of stories, covered area, age of construction, vertical and horizontal structure, type of roof, gravity or seismic design, assigned soil category, and the Vs30 according to Mori et al. (2020) map. This map provides, for a grid of 50x50 m, the shear wave velocity at 30 m depth (Vs30) over the whole Italian territory. Each school building was associated with the nearest grid point and then to the soil category based on the “Vs30-soil” correspondence defined in the regulation (D.M.17.01.2018). The second group of static information displayed on the platform is the seismic hazard. As anticipated, the nationally adopted hazard model MPS04 (Stucchi et al. 2004) was used as the hazard map. For the selected school building, the acceleration values taking into account litho-stratigraphic amplification for the return periods of the regulation (D.M.17.01.2018) are displayed in the platform. 4. Calculation tools of the platform Two types of damage maps can be calculated in IRMA: conditional damage and unconditional damage. For conditional damage, hazard maps with selected return period are adopted. For the calculation of unconditional damage, the entire hazard curve is adopted and the exceedance probabilities in an observation time window are considered. The observation windows are 1, 10 and 50 years. IRMA enables to perform the calculation on the whole country or on a single Region of interest, accounting for soil influence or neglecting it. The tool calculates, for each school building, the probability of reaching or exceeding the five damage levels from D1 to D5. Then, the risk tool allows to combine damage results in order to calculate direct economic losses and impact in terms of usable, unusable in the short period, unusable in long period and collapsed buildings. To move from damage to risk, it is necessary to define specific consequences functions, i.e. relationships between attended damages and corresponding losses. Default values of these coefficients are provided in the platform (those adopted in NRA 2018, Dolce et al. 2021) but the user can also customise the coefficients. The latest tool available allows researchers to run deterministic damage scenarios calculated based on preloaded shakemaps produced by INGV (Michelini et al. 2020) for several scenario events. Again, the user can combine different exposure matrices, sets of fragility curves, and specific consequences functions, obtaining damage and risk maps for a scenario event. 4.1. Results presentation The representation of results is fundamental and deeply related to the aim of the analyses performed and the assumptions made. For the large-scale analyses of school buildings in the MARS project: (i) the exposure database has gaps and was compiled by non-technical personnel, (ii) the fragility curves were produced by mechanical, empirical, and hybrid methods and are not calculated for a specific building adding further uncertainties to the computation of vulnerability, and (iii) the hazard associated with each school is derived from an interpolation of grid points. Then, the visualization of the result relative to the individual building can be misleading. Analyses performed with IRMA for school buildings, in fact, increase their value when the results, although computed item by item, are combined on macro-area. With this aim, the platform allows three levels of visualisation, in addition to the scale of the individual building: Municipality, Province, and Region (Figure 4). This visualisation consists of showing the average values of damage levels from D1 to D5. This average has been calculated as a function of the number of buildings in the selected geographic area or weighted on the floor area. If the floor area has not been indicated in the exposure database, the average floor area for the Region is considered. However, if in a Region there are more than 30% of school buildings for which the floor area is not indicated, its average value is considered not representative, and therefore for that Region it is not possible to calculate the average damage value weighted on the floor area. For this reason, in the map reported in Figure 5 there are five uncoloured Regions. For school buildings, aggregation by area is of particular interest, as it allows to more realistically account also for the exposure associated to the number of students. Moreover, in general, it is expected the area is also