PSI - Issue 44

Marco Gaetani d’Aragona et al. / Procedia Structural Integrity 44 (2023) 1760–1767 Marco Gaetani d’Aragona et al./ Structural Integrity Procedia 00 (2022) 000–000

1761

2

IP

Infill, Partition and integrated component

PFA PGA

Peak Floor Acceleration Peak Ground Acceleration

PRC Predicted Repair Cost RC Reinforced Concrete MDOF Multi-Degree-Of-Freedom NTHA Nonlinear Time History Analysis Stick-IT Stick for Infilled frame Typologies

1. Introduction Economic losses represent a burdening consequence in the aftermath of strong earthquakes. Indeed, the restoration of involved communities often requires the allocation of huge economic resources (Daniell et al., 2012). Refined loss assessment procedures requiring the analytical modeling of buildings can be employed to explicitly link the building seismic performances, expressed in terms of EDPs such as IDRs and PFAs, to expected damage and related quantitative-based loss assessment. In this context, the FEMA P-58 (2012), which represents the state of the art for the calculation of monetary losses at the building scale, embraces the PEER framework (Porter et al., 2003) to estimate the expected losses involving a four-step procedure: hazard analysis, structural analysis, damage, and loss analysis. By using explicit structural analyses into performance and loss assessment allows to incorporate the effect of suitable retrofit strategies in modification of building response and consequent estimated losses (Gaetani d’Aragona et al., 2018; Polese et al., 2019). Several alternatives were proposed to estimate EDPs in the context of the PEER framework, depending on the complexity of the analytical models and the analysis type. However, for analyzing large building stocks at the territorial scale it is not possible to employ detailed nonlinear models developed building by building. For this reason, simplified MDOF models have been proposed. To this end, fish-bone models (Khaloo et al., 2013; Soleimani et al., 2019; Jamsek and Dolsek, 2020) or stick models (Lu et al., 2014, 2018; Xiong et al., 2016, Gaetani d’Aragona et al. 2020, 2021a, 2022a, b, c) were recently proposed to condense the behavior of the whole structure in equivalent MDOF systems, allowing the investigation of seismic performances of large building portfolios with a suitable estimation of component damage level and related losses. In Gaetani d’Aragona et al. (2020) the Stick-IT model was recently proposed and calibrated to simulate the behavior of infilled RC moment-resisting frame buildings typical of the European Mediterranean region. The model has been demonstrated to predict the building response and direct economic losses with sufficient accuracy when compared to more refined finite element models Gaetani d’Aragona et al. (2021a). However, while in the latter paper a building scale validation for the proposed methodology has been performed, the derivation of geometrical and structural parameters required to generate the Stick-IT model proposed in Gaetani d’Aragona et al. (2020) can be non-trivial when dealing with the seismic assessment of large building portfolios and, thus, expert-informed assumptions or suitable simulated design procedures need to be adopted. Further, when the assessment of direct economic losses has to be performed at the large-scale, the variability of parameters influencing the loss assessment must be appropriately taken into account (e.g., intra-typology, seismic input, seismic response, damage definition, and loss functions) by introducing a fully probabilistic approach.To th is end, in Gaetani d’Aragona et al. (2022a) a probabilistic framework is proposed to allow the use of the Stick-IT model for large scale assessment purposes, by suitably extending the procedure proposed at the building scale (Gaetani d’Aragona et al. 2021a). The present paper resumes the main findings from Gaetani d’Aragona et al. (2022a) and highlights the potentiality of the Stick-IT model for estimating the performance and related losses for a large building portfolio. A database of 120 real reinforced concrete buildings damaged and repaired after the 2009 L’Aquila earthquake, including the reconstruction costs, are employed as a benchmark for the assessment. Although some features required to build the relative Stick-IT model are available from the database in such application, other quantities needed to complete the loss assessment are not known and thus a probabilistic-based assessment is performed. The applicability of the proposed probabilistic performance-based loss assessment based on the use of the Stick-IT model is evaluated by