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

1762

3

comparing the results of the simulated damage and loss assessment with the actual direct cost data for the buildings in the database. 2. Stick-IT model Gaetani d’Aragona et al. (2020, 2021a), proposed the use of the Stick-IT to predict the behavior of existing infilled RC moment resisting frames in each main building direction (fig.1(a)). The Stick-IT is a MDOF Non-Linear Shear Model consisting of a series of masses lumped at the story level connected in series by means of nonlinear shear springs. Based on the cyclic response analysis of a database consisting of 5632 refined Finite Element Models, in Gaetani d’Aragona et al. (2020) a regression study was performed to derive the story-level backbone and hysteretic parameters for the Stick-IT. The story-level backbone of nonlinear shear springs is approximated by a four-linear backbone curve defined by 7 parameters (fig.1(b)). To simulate the pinched response and the strength and stiffness cyclic degradation of the story-level spring link element, the Pinching4 material available in OpenSees (2021) is adopted.

(a) (b) Fig. 1. (a) Stick-IT model for each main building direction; (b) four-linear backbone for nonlinear shear springs (adapted from Gaetani d’Aragona et al., 2022). The regression formulas allowed to obtain the Stick-IT model starting from parameters that could be easily obtained for large-scale studies. the building dimensions along the analyzed direction, the normalized story number for the generic i th level and the opening percentage (i.e., the area of openings for the infills with respect to of the infill area for each panel). Finally, the infill elastic shear modulus is used to represent the infill consistency. The model parameters for envelope curves and hysteretic parameter are expressed according to an exponential formulation. By assuming a lognormal distribution, the backbone parameters, are also provided with suitable statistics in order to account for intra-model variability. Further details and formulation may be found in Gaetani d’Aragona et al. (2020). The procedure to derive Stick model backbone curves has been recently improved (Gaetani d’Aragona et al., 2022b, c) to simulate the behavior of infilled RC building characterized by different levels of design (gravity load, obsolete seismic codes), age of construction, number of storeys and in-plan shape and dimensions. The proposed procedure allows to explicitly simulate typical failure modes of non-conforming RC elements, local frame-infill interaction phenomena, and brittle collapse modes for existing buildings. Finally, the possible implementation of retrofit intervention at the building level to both increase the structural capacity and the stiffness of the building is also considered. 3. Loss assessment framework The procedure developed in Gaetani d’Aragona et al. (2021a) has been extended in Gaetani d’Aragona et al. (2022a) to allow the application at the large scale in a probabilistic perspective. This section clarifies how Stick-IT parameters can be derived when dealing with the loss assessment of large building portfolios, and how to explicitly account for intra-typology variability of mechanical and structural properties, structural response, definition of the seismic input, uncertainties regarding the damageable components, damage states, and quantification of the repair costs (fig.2). In