PSI - Issue 44

I. Boem et al. / Procedia Structural Integrity 44 (2023) 1260–1267 Boem I. and Gattesco N. / Structural Integrity Procedia 00 (2022) 000–000

1263

4

drops and progressive recovering. The number of cracks resulted lower in respect FRCM. As the number of cracks stabilized, a final branch with a slope close to that of the bare glass-fibers mesh was monitored until the rupture of the yarns (ultimate tensile strain of 1.95%, C.o.V. 15.6%). The horizontal gap between the CRM and the bare mesh curves evidenced a tension stiffening effect due to the intact mortar between cracks. The connection with the masonry abutments was made with three M6 threated bars, AISI 316 (resistance T b = 16.4 kN and axial stiffness EA b = 3417 kN) injected in the masonry and embedded in the mortar coating for a length of about 320 mm (Fig. 3). In FRCM-reinforced samples, this implied an increase of the mortar thickness at the vault ends (from 10 to about 30 mm), so to ensure an adequate covering. The two vault samples reinforced at the intrados (Fig. 3b) were provided also with an additional couple of transversal fiber-based composite connectors at the ends, deeply injected in the masonry abutments, so to oppose to possible inward sliding. In particular, in the sample vRI-C, L-shape Glass FRP connectors (7x10 cross section, 32.4 mm 2 fiber area) were applied; in the sample vRI-F, carbon-based FRP bars (10 mm diameter, 43 mm 2 fiber area) provided with anchor fans embedded in the mortar coating were introduced.

a

b

Fig. 3. Arrangement of connections at the vault ends, in case of the reinforcement applied (a) at the extrados and (b) at the intrados.

3. Experimental tests The experimental tests consisted in applying a horizontal transversal cyclic loading to the samples, with increasing amplitude, so to reproduce the effects of the horizontal inertia forces induced by an earthquake on masonry vaults supporting their own weight only and rising from fixed abutments. Two hydraulic actuators, installed at the opposite ends of a steel contrast structure, were activated simultaneously in the same direction to apply the load distributed along the vault through pinned frame systems (Fig. 1). The loading arms were hanged on the upper beams of the contrast frame through pre-tensioned springs, so to avoid additional gravity loads on the samples. Loading cells and potentiometer transducers allowed to monitor actual force distribution and the vaults displacement during the tests. More details concerning the test setup were reported in Boem and Gattesco (2021). The results of the tests are resumed in Table 2 in terms of horizontal peak load, F max , ultimate horizontal displacement (measured at the keystone),  u , and ratio between the peak load of the reinforced and the unreinforced sample, F max(R) / F max(U) , which provides an indication of the effectiveness of the strengthening intervention in terms of resistance. Table 2. Main experimental results on masonry vaults. ID Configuration

Ultimate displ.  u [mm]

Peak load F max [kN]

Effectiveness F max(R) /F max(U) [-]

vU

Unreinforced

0.55 11.8 10.8

~1

-

vRE-C CRM at extrados vRE-F FRCM at extrados

115.5 86.5

21.5 19.6

vRI-C vRI-F

CRM at intrados FRCM at intrados

+7.3 (-8.2) +9.5 (-6.5)

+105.3 (-27.9) +74.2 (-23.7)

13.3 (14.9) 17.3 (11.8)

In Fig. 4, the trend of the horizontal load F h against the horizontal displacement at the keystone  h of the four strengthened samples are reported. The curve of the unstrengthened sample was omitted due to the very low performances emerged during the test: the failure mechanism, with cracks at the abutments and at two opposite haunch sections, activated at F h = 0.55 kN for a very small displacement (  h ~1 mm).