PSI - Issue 44

Corrado Chisari et al. / Procedia Structural Integrity 44 (2023) 1108–1115 Corrado Chisari et al./ Structural Integrity Procedia 00 (2022) 000 – 000

1111

4

(a)

(b)

(c)

(d)

Fig. 2. Experimental setup for the tests on the arches: (a) drawing, (b) steel support and handling frame, (c) application of vertical loads, and (b) detail of the restraint.

The arches were subjected to constant vertical loads simulating the weight transferred by the backfill usually encountered in vaults, applied on five points symmetrically placed. The loads were simulated by steel plates connected to the arch by means of couples of omega-shaped steel elements enclosing a single block (Fig. 2c). After the application of the vertical loads, one of the loaded points, approximately as high as the arch barycentre, was horizontally restrained to the external steel frame by means of a device to which a load cell was connected in series (Fig. 2d). The system was rigidly connected to a sliding base at the bottom, which could move horizontally with respect to the external frame. In this way, a concentrated force was applied to the loading point, and a monotonic displacement history was applied at the base of the arch. The experimental results, in terms of force-displacement of the loaded point with respect to the base, are reported in Fig. 3a. It is possible to notice that while ordinary mortar provided an increment of lateral strength (from 0.86 kN to 1.50 kN) thanks to the twofold effect of increasing the cross-section depth and decreasing the internal span, this enhancement was limited, and at large lateral displacements the horizontal capacity tended to a value similar to the unreinforced arch strength. On the contrary, thanks to its remarkable ductility, FRLBM provides a much larger ultimate strength to the arch (2.59 kN, 73% more than ordinary mortar) with limited degradation. The position of the hinges was almost equal for arches UR and RI1 and is depicted in Fig. 3b. Some differences were observed in specimen RI2, with sliding between blocks and reinforcement detachment at hinge D (Fig. 3c) and absence of hinge C. A fundamentally different mechanism thus developed in RI2 (Fig. 3d).