PSI - Issue 44

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

110

Marta Faravelli et al. / Structural Integrity Procedia 00 (2022) 000–000

4

3. Description of the tool

The IRMA platform (Borzi et al. 2021) is accessible online with the most commonly used browsers providing username and password. The engine to perform risk calculation is OpenQuake (OQ), a software developed by the Global Earthquake Model (GEM) team. The tool dedicated to schools allows the scientific community to produce seismic damage and risk maps for Italian school buildings. Figure 3 shows the logical scheme of the platform. As is well known, three ingredients are needed for risk calculation: hazard, exposure, and fragility.

Fig. 3. Logical scheme of the IRMA platform.

The hazard is fixed and preloaded and it is the MPS04 hazard model (Stucchi et al. 2004), developed by INGV and adopted, up to now, by the Italian regulation (D.M.17.01.2018). The exposure is based on the inventory described in the previous section. From it, the user must define, by an excel file, an exposure matrix that associates taxonomies of school buildings with behaviours in reference to the vulnerability classes of the EMS98 scale (Grünthal 1998). More specifically, in Masi et al. (2021) and Lagomarsino (2022) it was established to consider the six vulnerability classes from A to F of the EMS98 scale, doubled to account for the ductile and brittle behaviour of the buildings for a total of 12 classes (A ductile , A brittle , B ductile , B brittle , etc.). The user associates the different probabilities of behaviour with the building taxonomies. Building taxonomies are defined according to: vertical structure, horizontal structure, roof type, number of storeys, age of construction, seismic or gravity design. It is not mandatory to consider all these fields. As an example, to a seismically designed two stories reinforced concrete school building with a floor of girders and bricks built after 1976, the user could assign the following behaviour: 30%C ductile , 10%C brittle , 50%D ductile , 10%D brittle . Finally, building vulnerability is defined by a combination of fragility curves corresponding to the behaviours A ductile , A brittle , B ductile , B brittle , etc. and for storey classes ranging from 1 to >=4. In another excel file, the user enters the median PGA (in g) and the β dispersion of the lognormal distribution of the fragility curves with reference to the five damage levels provided by the EMS98 scale, ranging from D1 (low damage) to D5 (collapse), for each vulnerability class (from A to F ductile and brittle). IRMA allows the user to combine different exposure matrixes and different sets of fragility curves to produce conditional or unconditional damage and risk maps. For the purpose of checking whether a vulnerability model is adequate, IRMA also allows to calculate damage scenarios using the shakemaps provided by INGV (Michelini et al. 2020) as ground shaking scenario. These scenarios can be compared with observed damage data (Dolce et al. 2019) in order to evaluate whether the model is representative of the reality of the “as built”. The results of the analyses can be viewed on a map, in a table, and downloaded as shapefiles and excel. IRMA also allows sharing the input files and maps obtained with other users of the platform in the perspective of collaboration that underlies the MARS project (Masi et al. 2021).