PSI - Issue 11

Siro Casolo et al. / Procedia Structural Integrity 11 (2018) 20–27 Siro Casolo & Giuseppina Uva / Structural Integrity Procedia 00 (2018) 000–000

23

4

height of 18.05 m and small openings in correspondence of the symmetry axis. The original façade was built with three leaf masonry of thickness equal to 0.80 m, with a masonry bond-texture made of roughly squared stones. Such texture could is quite similar to texture (I). After the first main shock it was heavily damaged, and a survey of the geometry and masonry walls was conducted (Doglioni et al, 1994).

Fig. 4. Comparison between texture (I) and texture (II). From left: top displacement history; hysteretic and kinetic energy; damage pattern: the brown segments (flexure) and crosses (torsion) corresponds to values of curvature larger than ( c max - c y )/ c y = 3. (Casolo & Milani, 2013)

The masonry façade has been modelled with quadrilateral rigid elements as shown in Fig. 4 that also reports the calculated damage patterns that indicate the activation of different collapse mechanisms depending on the different textures. Indeed, in the case of texture (I), the damage involves the entire façade (Fig. 4, first row). This fact is evidenced by the vertical segments indicating cracking along the symmetry axis of the façade. For texture (II), instead, the damage pattern corresponds to the activation of the overturning of upper part (tympanum) of the facade (Fig. 4, second row). The real damage pattern is very similar to that obtained for texture (I). Displacement time histories are also quite different for the two textures, whereas the levels of hysteretic and kinetic energies involved, which are generally much more stable (Uang & Bertero, 1990) are comparable. At the macro-scale, the crack patterns resulting from the RBSM dynamic simulations clearly show that the church façade built in texture (I) is weaker and tends to suffer a global collapse because of a predominant development of vertical flexural hinges combined with torsion, which confirm the inadequate interlocking of the texture. With a better texture (II), in high seismicity zone, it is rather evident a more localized collapse of the upper part of the facade (tympanum) due to insufficient vertical bending resistance, but better interlocking, which induce a preferential formation of horizontal hinges. It can be therefore stated that the presented approach has been able to satisfactorily reproduce the crucial role played by blocks’ shape and arrangement in the determination of the out-of-plane flexural – torsional response of the façade under seismic excitation both along the external leaves and inside the thickness of the walls. 3. Going for Dynamic analyses Another fundamental question is the choice of the method of analysis for the assessment of the out-of-plane response of masonry walls under seismic action. To provide a first answer, we present a systematic comparison between two methods that are well established in the field of earthquake engineering: non-linear dynamic (NLD) analysis and non-linear static procedure (NSP). Two case studies, representative of typical Italian church façades have been selected: the Church of Rosario, in Guastalla (Reggio Emilia, Italy) and the Church of Trasfigurazione in Moggio Udinese (Friuli, Italy), already analysed in Section 2.