PSI - Issue 64

Alessandra De Angelis et al. / Procedia Structural Integrity 64 (2024) 327–334 De Angelis et al./ Structural Integrity Procedia 00 (2019) 000 – 000

333

7

the comparison of experimental and numerical modal parameters of the structure, assuming frequencies and modal shapes as target parameters. A first finite element (FE) model was implemented to simulate the building in its as-built condition using the software SAP2000 (2016). In particular, columns and beams were modelled by beam elements, while concrete walls at the semi-basement were modelled by shell elements. For the RC elements, an elastic modulus of 25000 N/mm2 and a mass density of 2400 kg/m3 were assumed according to the results of in-situ tests. Infill walls and partitions were modelled through equivalent struts whose elastic lateral stiffness was evaluated according to eq. (2) as suggested by Dolsek and Fajfar (2008): = ∙ ⋅ ⋅ (2) where G m is the shear modulus of the wall, assumed equal to 0.4 E m ; L w , H w and t w are the length, height and thickness of the panel, respectively. The thickness of each strut was assumed equal to the thickness of the masonry panel, considering both the layers in case of double leaves walls, and the length is equal to the diagonal of the panel. The influence of openings such as doors or windows was taken in account by using the factor λ w proposed by Asteris et al. (2011), as a function of the opening percentage (opening area/infill wall area) according to eq. (3), =1−2 ⋅ 0 .54 + 1 .14 (3) in which α w is the infill wall opening percentage (area of the opening to the area of the infill wall). Starting from this FE models, other two models, were developed, representative of the condition of the building during AVT n. 2 and n.3, simply eliminating the masonry internal partitions in the transverse and longitudinal direction and the load of the floors in the model 2 and eliminating the infill wall along the staircase and some concrete wall at the semi-basement in the model 3. In addition, the thickness of the external infill walls, in the transverse direction, for which only the internal layer has been removed, was updated. Figure 7 shows the 3D views of the three models.

a)

b)

c)

Fig. 7. FE model: a) as-built condition; b) during AVT n. 2 and c) during AVT n. 3

The main purpose of the calibration process is to modify iteratively updating parameters to result in structural models that better reflect the measured data than the initial ones. The updating is carried out by a manual procedure because only a few number of parameters is considered. In particular, in this case only the stiffness of the infill walls was calibrated because the mass can be determined easily and with good precision given the type of construction known. The correlation between the final updated FE model and the test results are shown in Table 4 only for the first three global modes; as seen, a good correlation was obtained for all modes being D f less than 5%. Therefore, the correlation analysis provide a sufficient verification of the model assumptions, being a one-to-one correspondence between the mode shapes.