PSI - Issue 23

Solveig Melin et al. / Procedia Structural Integrity 23 (2019) 137–142 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

141

5

Fig. 4. Dislocation density vs. axial strain for: a) [100]-[110], b) [100]-[111], c) [110]-[111] grain boundaries. Red curve is mobile dislocations, blue immobile dislocations, and green total number of dislocations.

3.3. Dislocation evolution during loading

To, in detail, monitor what happens inside the material during loading, the dislocation arrangements inside the beams were recorded. In Fig. 5 some snapshots of dislocation arrangements early on in the loading process is seen for two of the studied cases, the grain boundary [100]-[110] in Figs 5a-d and [110]-[111] in Figs 5e-i. The color coding in Fig. 5 marks the type of dislocation, as defined in Stukowski et al. (2012). For the [100]-[110] beam it can be seen that the first dislocations are Shockley partial dislocations, formed at the top corners at the beam at the grain boundary interface in the [110] grain, cf. Fig. 5a. For a somewhat higher load, Fig. 5b, more dislocations, mostly Shockley, has formed close to the interface in the [100] grain, corresponding to the first dip in the stress curve and to the first maximum in the dislocation density curve in Fig. 3a. Figures 5c and 5d correspond to the following minimum and maximum of the dislocation density curve in Fig. 3a. A general observation is that only the first dislocation was formed in the [110] grain and all other dislocation related events took place in the [100] grain. Slip events, due to passage of dislocations to the surface casing slip steps on the surface, in the [100] grain occurred as slip in close packed <110> directions, as observed by Ahadi et al. (2017). For the beam with the [110]-[111] grains a different dislocation pattern emerged as seen in Figs 5e-i. Here the first dislocations were formed already during the relaxation of the beam, at zero applied strain, cf. Fig. 5e. After the relaxation, already at low strains, a specific pattern of Shockley dislocations was formed at the grain boundary, shaped like a fence, seen in Fig. 5f. This explains the first top in the dislocation density found in Fig. 4c and the small initial steps in the stress curve in Fig. 3c. The fence remains up to some percent of straining, where after some of them disappear and the dislocation density drops to a minimum, see Fig. 4c and Fig. 5g. The fence seems to hinder further dislocation generation in the [111] grain. For higher strains the dislocation density increases again and the structure of Fig. 5h is formed, followed by another drop in dislocation density seen in Fig. 5i. It can also be observed that as the strain increases, more dislocations of type immobile are observed as compared to mostly mobile dislocations for low strains. This is consistent to what was observed in Fig. 4c. A general observation for the [110]- [111] case is that many dislocations are formed close to the grain boundary, primarily in the [111] grain, and that slip mostly occurs in the [110] grain in close packed <110> directions. For the [100]-[111] beam the first dislocations were formed at the grain boundary and the following dislocations were formed in the [100] grain only. Also only slip in the [100] grain during loading, was found for this case.