PSI - Issue 44

Agnese Natali et al. / Procedia Structural Integrity 44 (2023) 2326–2333 Agnese Natali, Francesco Morelli / Structural Integrity Procedia 00 (2022) 000–000

2328

3

simplified using the methodology described in (Tsarpalis et al., 2021). Specifically, the full cross-aisle frame is substituted with a pair of connected upright frames (i.e., one “macro-column”) with calibrated horizontal springs at the top level, to simulate the stiffness of the roof. A Rayleigh damping formulation is employed, using a viscous damping ratio of 3% for the first and second eigen periods, and a + 0.8 ⋅ combination for mass and gravity loads. The uprights are simulated as elastic beam elements with a P- Δ formul ation to account for the effect of geometric nonlinearities. In the POS re-design, the only source of material nonlinearity is the plastic ovalization of the diagonal bolt hole. The opening of the bolt hole is simulated using a zero-length element that incorporates the elastic-perfectly-plastic gap material of OpenSees. Ultimately, the diagonal and its bolted connections are simulated by a macro-element that contains (a) a zero-length element with gap behaviour that simulates bearing failure and (b) an elastic beam element that captures the axial stiffness of the diagonal member and factors (other than bearing failure) that reduce the shear stiffness of the upright frame (Fig. 2a). The control parameters of the elastic-perfectly plastic gap material are bilinearly fitted to the analytical equation of prEN 1993-1-8:2021 (2021), as illustrated in Fig. 2b. The complete procedure for the derivation of the macro-element that simulates bearing failure of the diagonal is described in Tsarpalis (2022).

14.30

1.30 0.65

25.45

1.20

2.50

1.10

0.30

1.90

Fig. 1. Cross-aisle view of case study (dimensions in m).

a

b

Fig. 2. (a) Numerical modelling of a diagonal element with bearing failure behaviour in OpenSees (Mazzoni, 2017); (b) Bilinear fitting of the elastic-perfectly-plastic gap material, to match the analytical predictions of prEN 1993-1-8:2021 (2021), as described in Tsarpalis (2022). In post-processing, for all the inputs and for all the intensity levels, safety checks of all the components are executed. Resistance values are evaluated according to Eurocode 3 part 3 (2006). Due to the prevalence of brittle failure modes, it is assumed that the failure of any member/connection (i.e., when a utilization factor exceeds the threshold value of 1.0) leads to a global collapse of the structure. An additional global collapse rule is considered: if