Issue 49

R. Suresh Kumar et alii, Frattura ed Integrità Strutturale, 49 (2019) 526-535; DOI: 10.3221/IGF-ESIS.49.49 533 245964 in the case of plate geometry. The total number of elements can also be further reduced by adjusting the boundary of the sub model without affecting the results. The SIF variation extracted from FRANC3D at the critical location (B from Fig.10) during FCG is shown in Fig.11. It indicates that K I increases with an increase in ‘a’ up to ‘a/t’ ratio of 0.4 (a = 6 mm) and then it decreases, due to the influence of the neighbourhood compressive stress zone as evident from Fig.6. Figure 11: SIF variation for the pipe bend at the critical location-B The numerically predicted estimate of FCG behaviour of the pipe bend is compared with the experimental data and depicted in Fig.12. It is clear that the numerically predicted crack growth is in good agreement with the experimental results. The reduction in the rate of crack growth beyond 6 mm is due to the influence of low K I value. Figure 12: Comparison of ‘a’ Vs ‘N’ on pipe bend S UMMARY AND C ONCLUSIONS wo representative geometries have been selected from the literature towards validating numerical simulation of FCG behaviour. The selected geometries are a plate specimen and a representative full-scale pipe bend. The important observations are summarised below.  SIF is computed at the mid-side nodes along the crack front based on M-integral concept. The deployment of sub- modelling concept has made the computation of FCG behaviour more effective. The crack growth direction is decided by maximising the ‘Kink angle’ measurement at the base of the crack front using the maximum tensile stress. This also enhances the accuracy of the numerical model. T