PSI - Issue 44

Gabriele Guerrini et al. / Procedia Structural Integrity 44 (2023) 1877–1884 Gabriele Guerrini et al. / Structural Integrity Procedia 00 (2022) 000 – 000

1880

4

The two building prototypes were incrementally tested on the shake-table, along the weak longitudinal direction, by scaling the same ground motion up to the attainment of near collapse conditions. The experimental results demonstrated that the proposed retrofit system is effective in improving the lateral response of URM buildings. The retrofit system allowed to reach twice the scaling of ground motion applied to the bare configuration of the specimen, guaranteeing a global box-type response, and preventing the onset of local failure mechanisms, which characterized the experimental response of the bare specimen. Further details can be found in Miglietta et al. (2021). 3. Material properties and masses The material properties were characterized at the Department of Civil Engineering and Architecture of the University of Pavia, Italy. Two groups of values of the main CS masonry properties are reported in Table 1, since two batches of masonry were used to build the retrofitted masonry pier and the retrofitted building prototype. CS masonry had average densities of respectively 1836 kg/m 3 for the first material batch, while 1862 kg/m 3 for the second one. Mechanical properties of the clay masonry are not included since they were not explicitly modeled. The timber of the retrofit was classified as S10/C24 according to EN 14081-1 (CEN 2016), with an average density of 517 kg/m 3 , a mean Young ’s modulus E t = 11000 MPa, a mean shear modulus G t = 690 MPa, and a mean tensile strength parallel to fibers f t = 21 MPa. The mean values of the elastic moduli were taken from EN 14081-1, while the tensile strength was estimated as 1.5 times the tabulated characteristic value (ASCE 2017). The OSB was classified as OSB/3 according to EN 300 (CEN 2006) and had an average density of 572 kg/m 3 . The same timber properties were assumed also for the second floor and roof framing systems. The tensile strength and axial stiffness of tie-down anchors were provided by the manufacturer as 12.8 kN and 12.6 kN/mm, respectively. The building specimen first- and second-floor masses were about 11.2 t and 1.9 t, respectively, while the complete roof structure had a mass of 2.8 t. The bare building had a total mass of 47.5 t, while the timber retrofit system had a total mass of about 1.6 t.

Table 1. Calcium-silicate masonry mechanical properties.

Component test Avg. C.o.V. (-)

Retrofitted building test

CS masonry property (unit)

Avg. 7319

C.o.V. (-)

Young’s modulus, E m (MPa) Compressive strength, f m (MPa) Bed-joint cohesion, c (MPa)

6593

0.09 0.06

0.15 0.11

10

10

0.62

- -

0.49 0.62

- -

Bed-joint friction coefficient.,  (-) 0.71

4. Numerical simulation of the retrofitted pier cyclic response The experimental simulation of the in-plane shear compression-test performed on the retrofitted CS masonry pier was carried out with the software TREMURI (Lagomarsino et al. 2013), which relies on the use of macroelements (Penna et al. 2014) to reproduce the flexural and shear damage mechanisms of masonry panels. There, the shear strength of a masonry panel is implemented as a Coulomb criterion, with parameters properly calibrated to be macroscopically representative of different shear failure modes. The numerical modelling of this experimental tests was fundamental to develop a proper approach to consider the influence of the timber retrofit system in the lateral response that could be extended to models of entire retrofitted buildings. To model the retrofitted masonry pier, the entire macroelement was discretized into 6 sub-elements, while the contribution of the vertical posts to the flexural resistance was modelled through nonlinear truss elements (with zero moment of inertia and shear modulus) with an elastic-perfectly plastic hysteresis model (Bracchi et al. 2020). Truss elements were discretized into segments of equal length and connected to the same nodes as the discretized macroelements (Fig. 3). The intermediate segments of each post were assigned cross-sectional area ( A p = 4800 mm 2 ), mean elastic modulus ( E t = 11000 MPa) and mean tensile strength ( f t = 21 MPa) of the timber members. The top and bottom end-segments were modelled with the same cross-sectional area A p but with equivalent values for Young