PSI - Issue 14

Aman Arora et al. / Procedia Structural Integrity 14 (2019) 790–797 Aman Arora/ Structural Integrity Procedia 00 (2018) 000–000

794

5

about grain size, crystallographic texture distribution and grain boundaries network. It can be seen that there is a large population of ∑3 special grain boundaries in blue about 87%, rest are random grain boundaries and average grain size of 28 µm.

(a

(b)

∑3

Fig. 5. EBSD map of annealed, uncharged Nickel. (a) Inverse pole figure (IPF) in loading direction (b) Grain boundaries distribution, 87% of CSL mainly ∑3 grain boundaries.

For uncharged sample three cracks were initiated corresponding to nine grains. Elastic modulus map, kernel average misorientation (KAM) map and schmid factor maps for the {111} <110> slip systems according to the macroscopic loading direction are shown in Fig. 6. It can be seen from Fig. 6 (c) that crack in uncharged sample is initiating at grain boundaries with neighbouring grains having high schmid factor such as grain 2, 3, 4, 7, 8, 9. Moreover, elastic anisotropy is also playing role as can be seen in Fig. 6 (d) as elastic modulus difference in neighbouring grains can induce high strain localization which in turn creates high incompatibility stresses at grain boundaries. Kernel average misorientation (KAM) map in Fig. 6 (e) shows that there is a local change is orientation within the grain 1 which is having low elastic modulus as compared to the neighbouring grains. For hydrogen charged sample, three cracks were initiated corresponding to thirteen grains. Elastic modulus map, misorientation average map and schmid factor maps for the {111} <110> slip systems according to the macroscopic loading direction are displayed in Fig. 7. It can be seen in Fig. 7 (c) that cracks in hydrogen charged sample is also initiating at grain boundaries. Only grain 1 and 11 have high schmid factor but there is random distribution in schmid factor values in other crack neighbouring grains. Moreover, it seems that elastic anisotropy is not playing a significant role as can be seen in Fig. 7 (d) as the elastic modulus difference is negligible in crack neighbouring grains. Kernel average misorientation (KAM) map in Fig. 7 (e) shows that there is not much local change in orientation within the grains. However, a local change in orientation due to localized plasticity is only near the grain boundaries where hydrogen concentration is supposed to be more. Thus, it can be inferred that near grain boundaries HELP mechanism is also playing a role. Crack is just nucleating due to grain boundaries decohesion as hydrogen also weaken the nickel-nickel bond strength. Further analysis of schmid factor and elastic modulus of crack neighbouring grains provides a better understanding of effect of hydrogen on crack nucleation in nickel. As can be seen in Fig. 7 (a), crack is nucleating near the grains with high schmid factors which is in the range of 0.3 to 0.5 in both the samples. Elastic modulus is observed to be randomly distributed for uncharged nickel sample but have a narrow range of elastic modulus for hydrogen charged nickel sample. However, after analysing the difference in schmid factor and elastic modulus in crack neighbouring grain, in Fig. 7 (b) it can also be observed that schmid factor difference and elastic modulus difference is more for uncharged nickel samples than the hydrogen charged sample. In hydrogen charged sample the difference in schmid factor and elastic modulus is almost negligible. Thus, it can be concluded that elastic anisotropy is not playing any role in crack nucleation in hydrogen charged sample, as can also be seen in Fig. 7 (e) that shows no local plasticity within the grain.