Issue 31

H.F.S.G. Pereira et alii, Frattura ed Integrità Strutturale, 31 (2015) 54-66; DOI: 10.3221/IGF-ESIS.31.05 62 have influence if a quite small value is adopted, i.e. equal to 0.00001. For a viscosity value of 0.001 the results were completely incoherent, with the maximum pullout load attaining the value of 200 kN, approximately three times the real value. Fig. 15 shows the results adopting a bond stress – slip law with an ultimate slip of 1000 mm. In the latter case, the viscosity coefficient only has influence for the higher viscosity value, i.e. 0.001, causing an increasing in maximum pullout load of approximate 10%. Based on these results, it was decided to use the smallest viscosity coefficient in all simulations (0.00001), this option leads to an increase of the computational time cost. Figure 14 : Bond stress-slip relationship displacement at failure equal to 5 mm simulated with different viscosity coefficients. Figure 15 : Bond stress-slip relationship displacement at failure equal to 1000 mm simulated with different viscosity coefficients. N UMERICAL MODELLING OF THE SENA ET AL . (2009) PULLOUT TESTS his section presents the numerical results within the finite element framework regarding the simulation of the experimental pullout tests performed by Sena-Cruz et al. [31]. Smooth rebar model The simulations were firstly performed assuming a smooth surface for the steel rebar. The pullout behaviour of three types of rebars, namely, non-alloy, galvanized and galvanized + epoxy rebar were modelled. The obtained numerical results were compared with the experimental results and with the results obtained using the analytical model proposed by Sena-Cruz et al. [31]. Tab. 7 includes the parameters that define the local bond – slip law (Fig. 6) used in the numerical simulations. These parameters were obtained by an inverse analysis procedure using an analytical shear-lag model [31]. The inverse analysis procedure consisted in fitting the numerical pullout force – slip relationship with the experimental correspondent one by varying the local bond law parameters. In these simulations the FE mesh depicted in Fig. 1 was used. Rebar  max  [-] s 1 [mm] Error [%] Galvanized + epoxy 0.73 ( fcm ) 0.5 0.4 2.00 3.9 Galvanized 1.46( fcm ) 0.5 0.62 1.30 2.5 Non-alloy 1.75( fcm ) 0.5 0.52 1.30 1.8 Table 7 : Parameters of the bond-slip relationship obtained by inverse analysis [31]. Fig. 16 depicts both the experimental and numerical results from the finite element analysis, for specimens with non-alloy steel rebars, in terms of pullout force vs . slip. The numerical results were inside of experimental envelope obtained experimentally, however the simulations results by ABAQUS slightly overestimates the results obtained by the analytical formulation of [31], in terms of stiffness and maximum pullout force. T