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

1765

6

damageable components is properly accounted within the Monte Carlo simulation. For the generic simulation, the extension of infills, partitions, and the opening percentage is then simulated by entering in the adopted inverse cumulative probability distribution with a random number generated using the uniform distribution over the interval [0,1] (fig. 4 (b.2)). The artificial EDPs, separated for each story, direction and simulation, are adopted to quantify the level of damage attained (DS) by entering in the in the DS fragility curve adopted for the component with a random number generated in the interval [0,1] (fig. 4 (c.1)), this way the DS experienced by each component is defined. Note that a unique random number is extracted for both infills and partitions for each simulation (i.e., correlated DSs), while in each separate direction it is assumed that the DS in the longitudinal direction can be different from that in the transverse one and may differ for each story depending on the EDP prediction (i.e., uncorrelated DSs).

Generation of additional EDPs (a)

Infill/partitions with openings (IP with )

∑

(

)

L

L ds DS ≥

=

P(DS | edp j ) 1

int

1

i

, STORY DIRECTION

Unit repair cost for ds=DS i (c k )

Max cost

Uncertainty

€ i

0

edp j

Mincost

Infill/partitions without openings (IP w/o )

P(DS | edp j ) 1

L int

Max Q

Q i

Min Q

For each DS, IP ra

0

edp j

randn(0,1) For each STORY, DIRECTION

Story

Story

Dir

DS

Dir

DS

1 1 1 j k = = = ∑∑∑ i

1 1 1 j k = = = ∑∑∑ i

(

)

, C sym CAL

, k IPwith i k IPwith c L , ,

, k IPw o i k IPw o c L / , , /

=

+

tot

Fig. 4. Probabilistic loss framework proposed (from Gaetani d’Aragona et al., 2022a)

By summing the extension of all damaged components (ds