Issue 78

A. A. Frantsuzov et alii, Fracture and Structural Integrity, 78 (2026) 474-488; DOI: 10.3221/IGF-ESIS.78.27

The same pipe material was examined after industrial heat treatment and after an additional laboratory heat treatment designed to increase PAGS while maintaining approximately the same yield strength. The EBSD results for both conditions are shown in Figure 6. In the sample subjected to the laboratory heat treatment (Figure 6b), PAG boundaries can be identified relatively easily even before reconstruction, because several martensite packets are formed within a single PAG. Changing the inverse pole figure projection direction can further enhance the color contrast between neighboring PAGs and improve manual interpretation. For the sample with smaller PAGS (Figure 6a), the PAG boundaries are much more difficult to identify directly from the raw EBSD map. With decreasing PAGS, each prior austenite grain is less likely to contain several distinguishable martensite packets after quenching, which reduces the visibility of the parent grain structure. Increasing the EBSD magnification may therefore appear to be a possible solution for manual PAG identification. However, the results presented and discussed below show that this approach has limitations.

Figure 6: EBSD and reconstructed PAG maps of the same pipe steel after: a) industrial heat treatment; b) additional laboratory heat treatment aimed at increasing PAGS. Since EBSD magnification affects both spatial resolution and the number of grains captured within a single map, its influence on PAG reconstruction was evaluated. Higher magnification may improve local boundary resolution, but it also reduces the analyzed area and therefore decreases the statistical representativeness of the measurement. For this reason, the comparison was performed in two steps: first, by comparing PAGS distributions obtained at different magnifications, and second, by overlaying reconstructed PAG maps from the same sample region. The first comparison was performed using EBSD maps acquired at ×500, ×1000, and ×3000 magnifications with areas of 250×188, 125×94 and 42×31 μ m respectively. The reconstructed PAG maps in Figure 7 show that the MTEX algorithm produces physically reasonable parent grain structures at all investigated magnifications. However, the resulting PAGS size distributions (Figure 8) indicate a systematic shift toward smaller grain sizes with increasing magnification. The ×500 and ×1000 distributions remain relatively close to each other, whereas the ×3000 map contains a higher fraction of small reconstructed grains. Reconstructed grains smaller than 3 μ m were excluded from the statistical analysis because PAGs of this size are difficult to validate reliably by independent methods. This difference may be related not only to the higher spatial resolution, but also to the smaller mapped area and the increased probability of subdividing individual PAGs during reconstruction.

480