13th International Conference on Fracture June 16–21, 2013, Beijing, China showed no evidence of texture heterogeneities such as orientation clustering or correlation. The microstructural and texture characteristics estimated from these maps, such as grain size and pole figures and ODFs, agree with the results previously found during the metallographic inspection and texture analysis of the studied samples.3 Based on these findings, and on the fact that the number of grains in the orientation maps of the test samples is relatively large (of the order of, or greater than 104), one can be concluded that the local EBSD-derived distribution of GBs is representative of that of the bulk sample. Figure 6 shows the grain boundary statistics or mesotexture in the HIC-free and HIC-stricken groups of samples. The first result to stress from the results in Fig. 6a is that the proportion of CSL boundaries observed for these two groups of samples is not statistically significant. Therefore, in Fig. 6b, where the average proportion of HABs and LABs are shown for both groups of samples, CSLs boundaries are accounted for as HABs. 20 30 40 50 60 70 80 90 High-angle (>15 deg.) Low-angle (<=15 deg.) Coincidence Site Lattice 0 10 Random HIC-free HIC-stricken WRA HRA 20 60 80 100 WRB CRA HRB Frequency (%) Grain boundary type LABs between {011}ND grains High-angle Low-angle 0 40 Freque ncy (%) s and departs considerably from that of the (theoretically-deduced) random the HIC-stricken group. In contrast, the proportion of HABs the HIC-free group is about 45% lower than in the HIC-stricken group. For the group of HIC-stricken samples (except for HRB) the mesotexture is the closest to that of the random polycrystal. The fraction of LABs in samples HRA and HRB, on average, close to 13%; this proportion is about half of that observed for the rest of the test samples. In a very particular case, sample HRB shows a remarkably higher fraction of LABs. However, given the results shown in Fig. 5, and under the assumption of absence of orientation correlation, it is reasonable to expect that the majority of these GBs occurs between grains with orientations close to {001}ND. The implications of this for HIC propagation is discussed later. Samples HRA and CRA show a high fraction of HABs, which, on average, is about 45% higher than the proportion observed in the HIC-free samples. Once again, sample HRB stands out as a remarkable exception, where the small fraction of HABs and CSLs can be related to the high fraction of LABs shared by {001}ND-oriented grains. Grain boundary type (a) (b) Figure 6. Mesotexture of the investigated samples. (a) By sample, considering with CSLs separately accounted for. (b) By group of samples with CSLs accounted for as high-angle boundaries. For the group of HIC-free (warm-rolled) samples the GB statistics shows the highest proportion of low-angle boundarie polycrystal. This result is in complete agreement with the results shown in Fig. 4 for the crystallographic texture of these samples, and is an indication that no significant orientation correlation [12] occurs in them. Similarly, orientation correlation was also neglected for the group of HIC-stricken samples. Fig. 6b also shows that, on average, the fraction of LABs in the HIC-free group of samples almost doubles that of in 3 Given this result, and for the sake of space economy, the pole figures and ODF presented in this paper are those obtained from the EBSD measurements -5-
RkJQdWJsaXNoZXIy MjM0NDE=