PSI - Issue 64

Salvatore Verre et al. / Procedia Structural Integrity 64 (2024) 1508–1515 Salvatore Verre / Structural Integrity Procedia 00 (2019) 000 – 000

1511

4

experimental P - g responses. Fig. 2a also shows the average P - g responses, obtained by averaging the applied forces associated with any global slip g for specimens with the same number of layers. The P - g responses of specimens with 1 and 2 layers of textile had similar characteristics, with a pseudo-linear initial branch, followed by a branch with reduced stiffness and a plateau.

Table 2. Results of single-lap shear tests.

Specimen P * (kN) P avg

* (kN) (CoV) σ * (MPa) σ avg

* (MPa) (CoV) g

1 (mm)

P plat (kN) G f (N/mm)

DS-1L-1 DS-1L-2 DS-1L-3 DS-1L-4 DS-1L-5 DS-1L-6 DS-2L-1 DS-2L-2 DS-2L-3

8.21 7.90 7.67 8.05 7.67 8.12

1907 1834 1782 1871 1782 1887 1390 1387 1393

7.94 (2.9%)

1844 (2.9%)

0.83

7.59

0.695

11.96 11.94 11.99

11.96 (0.2%)

1390 (0.3 %)

0.46

11.26

0.765

a

b

c

Fig. 2. (a) Experimental P-g response, (b) transversal and (c) vertical cracks observed in DS-2L-2.

The peak load of specimens with two textile layers was 1.51 times the peak load of specimens with one layer. The initial stiffness was greater in the case of two layers than in the case of a single layer. The plateau was less extended in the case of two layers than in the case of a single layer. This was probably associated with a longer effective bond length in the case of two layers. In fact, the extension of the plateau branch corresponds to the self-similar propagation of the interfacial crack at a constant load when the shear stress transfer is fully established. During the loading process, transversal cracks initially formed near the loaded end on the external matrix layer. As the shear stress transfer propagated towards the free end, further transversal cracks formed near the free end. The failure of the specimens was caused by an interlaminar crack formed at the fiber-matrix interface (internal steel layer) near the loaded end and propagated along the bonded length up to the complete detachment (Figs. 2b and c). Fig. 3a shows a scheme of the debonding mechanism observed. With the same SRG composite, similar failure mechanisms were observed by Sneed et al. (2016), Ascione et al. (2020) and Ombres and Verre (2020), whereas the fiber failure was also observed in some specimens by Ascione et al. (2020) and Thermou et al. (2021). The average P - g responses were used to determine the global slip g 1 at the beginning of the plateau, the plateau load P plat , and the interfacial fracture energy G f with the procedure defined in ASTM (2021). The value of g 1 was determined as the minimum global slip corresponding to d P /d g ≤0.5 kN/mm. It should be noted that this slope is smaller than the value suggested in ASTM (2021) (5kN/mm), since this standard refers to FRP composites, which are characterized by higher slopes of the P - g responses compared to SRG composites. The ratio of the plateau load with two layers to the plateau load with one layer is equal to 1.48. 2.4. Effect of the number of textile layers on the bond mechanism Fig. 3b shows the ratio ξ of P avg * of specimens with two layers to P avg * of specimens with one layer in function of the bonded length, including the results obtained by Ascione et al. (2020) and Thermou et al. (2021) with the same SRG composite considered in this paper.