PSI - Issue 44

Guadagnuolo M et al. / Procedia Structural Integrity 44 (2023) 942–949 Guadagnuolo et al. / Structural Integrity Procedia 00 (2022) 000–000

946

5

a) d) Fig.1 Pull out test - OLY ROD GLASS L with OLY RESIN 10 in (a) rockfill; (b) masonry; (c) clay-brick; (d) concrete. b) c)

Table 1. Results of tests in the concrete support. Specimen Adhesive N exp (kN)

Failure Mode bar breaking bar breaking bar breaking bar breaking bar breaking

Specimen

Adhesive N exp (kN)

Failure Mode bar breaking bar breaking bar breaking bar breaking

ORG_CLS_1 ORG_CLS_2 ORG_CLS_3 ORG_CLS_4 ORG_CLS_5

RESIN 10 RESIN 10 RESIN 10 RESIN 10 RESIN 10

37.30 36.10 38.16 36.71 35.50

ORG_CLS_6 ORG_CLS_7 ORG_CLS_8 ORG_CLS_9 ORG_CLS_10

RESIN I RESIN I RESIN I RESIN I RESIN I

36.81 35.95 37.87 36.92 35.72

bar breaking RESIN 10: epoxy resin; RESIN I: vinyl-ester resin

4. Comparison between experimental tests and theoretical values Table 2 summarizes the literature pull-out tests used for comparison with the experimental tests performed. Achillides performed pull-out tests of GFRP bars from 150 mm cubic mould concrete (Achillides 1998). Baena et al. executed 88 pull-out tests of GFRP bars from 200 mm cubic mould concrete, according to ACI 440.3R-04 (2017) and CSA-S806-02 (2002) standards. The length of the GFRP bars was five times the rebar diameter (L eff = 5d b ) (Baena et al. 2009). Gao et al. performed pull-out tests of GFRP bars from 150 mm concrete cubes according to CSA-S806-02 (2002). The GFRP bars were made of unidirectional E-glass fibres and vynilester resin, with an embedment length of 5d b (Gao et al. 2019). By analyzing all the experimental data of the above-mentioned Authors presenting the same pull-out failure, the same type of bar, and the same surface treatment, it emerges very clearly that there is a relevant gap between the experimental pull-out forces and the theoretical ones. To address the above inconsistency, some theoretical models propose a modified bond-slip relationship, but it depends on several parameters that can only be calibrated through experimental analysis (Baena et al. 2009; Gao et al. 2019; Pepe et al. 2013). Moreover, the aforementioned approach does not allow the prediction of the failure mode. In this area, organizing, according to the main parameters (grout, bar surface treatment, failure behavior), the literature data of pull-out tests of GFRP bars from concrete support that were performed by (Baena et al. 2009; Barbieri et al. 2016; Ceroni et al. 2016; Gao et al. 2019), it emerges that the relation between pull-out strength and bar diameter diverges strongly from the theoretical one.

Table 2. Scheme of literature data tests considered Author Tests d b (mm) L eff (mm)

f c (MPa)

Grout

Surface treatment

Achillides

47

8.50÷13.50

64.00÷135.00 15.00÷41.00 Vinylester hybrid resin

Rough external surface

Baena et Al.

18

8.55÷19.14

34.20÷95.70

49.55÷58.20 Urethane vynilester, Vynilester

Grooves, helical wrapping surface texture Sand coating, surface texture, helical wrapping, helical wrapping with sand coating, grooves

Gao et Al.

92

8.22÷21.21

32.88÷84.84

28.20÷47.60

Vynilester resin