PSI - Issue 44

Alessandra Gubana et al. / Procedia Structural Integrity 44 (2023) 512–519 A. Gubana and A. Mazelli / Structural Integrity Procedia 00 (2022) 000–000

516

5

Fig. 4. Pushover curve in +Y direction under Group 1 force distribution.

4. Incremental Dynamic Analyses (IDA) Incremental Dynamic Analysis (IDA) consists of the execution of several nonlinear dynamic analyses under a multiple-scaled set of ground motion records. Every IDA curve represents the relationship between the Damage Measure DM and the Intensity Measure IM. In the present work, the Interstorey Drift Ratio (IDR) was selected as DM ( Masi et al. 2021; Kwon and Elnashai 2006; Nanos and Elenas 2006) . Five different Damage Level DL were defined in Table 1, for Non Structural Damage NSD and Structural Damage SD. The IDR limits and the decision to merge DL4 and DL5 are consistent with other works in the literature ( Masi et al. 2009 ).

Table 1. Damage Levels and corresponding IDRs.

DL

NSD None Slight

SD

IDR [%]

0 1 2 3

None None Slight

<0.1

0.1-0.25 0.25-0.5 0.5-1.0

Moderate

Heavy

Moderate

4-5

Very heavy

Heavy

>1.0

Two different IMs were selected to perform a comparison. They are both wide used measures in the literature: the Peak Ground Acceleration PGA (Karapetrou et al. 2016) and the spectral pseudo acceleration in correspondence of the first vibration mode of the structure S a (T 1 ) (Vamvatsikos and Cornell 2002; Vamvatsikos and Cornell 2004). A set of 15 couples of accelerograms was chosen by means of Rexel software (Iervolino et al. 2009) and based on the disaggregation results (Fig. 5). In order to perform IDAs on such a complex building (more than 4500 elements), the initial nonlinear FEM model of the pushover analyses was iteratively changed. A first model with only the hinges which reached collapse or significant plasticization was build. Then, the sections without hinges were checked after each IDA step to ensure that the stresses remained below the elastic limit or that the yielding was negligible. Otherwise a hinge was added. The model was enriched by hysteretic laws: the Takeda hysteresis (Takeda et al. 1970) was chosen for beams, and the Pivot hysteresis (Dowell et al. 1998) was chosen for the shear hinges. For the fibre hinges, Takeda hysteresis was chosen for concrete, and kinematic was chosen for steel. This modelling choice describes the stiffness degradation, but for the necessity of an FEM model that was not too complex, the post-peak strength degradation was neglected