*3.3. Fractography*

As already discussed, and shown in Figure 12, the room-temperature fracture behavior of studied T92/T92 weldments during tensile straining is primarily controlled by the used PWHT conditions, whereas the effect of electrolytic hydrogenation is less critical. The observation of macroscopically ductile failures (Figure 12a,b) within the SC-HAZs of both the hydrogen-free and hydrogen-charged T92/T92 weldments after the tempering PWHT-1 and tensile straining, can be related to higher deformability of the over-tempered SC-HAZs compared to the BM and WM regions. The corresponding fracture analyses for the hydrogen-free and hydrogen-charged weldments initially processed by the PWHT-1 are shown in Figures 13 and 14, respectively. Although the both hydrogen-free and hydrogen-charged weldments in PWHT-1 material state were broken in a recognizably ductile manner in their SC-HAZs (see the fracture paths in Figures 13a and 14a), the signs of hydrogen embrittlement are clearly visible for the hydrogen-charged weldment. In Figures 13a and 14a the SC-HAZ microstructures beneath final fractures are characterized by intensive deformation of grains in a direction of dominating tensile stresses. The microstructure of hydrogen-free weldment after the PWHT-1 (Figure 13a) shows randomly distributed precipitates and signs of microvoid coalescence along the elongated grain boundaries indicating the origin of the formation of ductile dimples on the fracture surface (Figure 13b). Whereas the fracture surface of hydrogen-free T92/T92 weldment in PWHT-1 material state shows completely ductile dimple fracture (Figure 13b), the fracture surface of the hydrogen-charged weldment shows quasi-cleavage areas with typical "fish-eye" morphology due to the local hydrogen-embrittlement beside ductile dimple tearing areas (Figure 14b). The hydrogenated sample after the PWHT-1 shows in its microstructure beneath the fracture surface hydrogen-assisted cracks and even some grain fragmentation (Figure 14a). The "fish-eye" fractographic object on the fracture surface (Figure 14b) represents the hydrogen embrittled quasi-cleavage zone created radially along the central inclusion.

**Figure 13.** Fracture analysis of hydrogen-free T92/T92 weldment initially processed by the PWHT-1 and subsequently ruptured in SC-HAZ after the room-temperature tensile test: (**a**) fracture path and microstructure beneath the fracture; (**b**) fracture surface showing pure ductile dimple tearing.

**Figure 14.** Fracture analysis of hydrogen-charged T92/T92 weldment initially processed by the PWHT-1 and subsequently ruptured in SC-HAZ after the room-temperature tensile test: (**a**) fracture path and microstructure beneath the fracture; (**b**) fracture surface showing mixed fracture consisting of both transgranular quasi-cleavage with a "fish-eye" morphology and ductile dimple tearing.

On the other hand, the observation of rather brittle failures (Figure 12c,d) within the WMs of both the hydrogen-free and hydrogen-charged T92/T92 weldments after the renormalizing-and-tempering PWHT-2 can be related to their lower deformability. It appears as a result of thermal embrittlement and possibly remaining microstructural heterogeneities of the original as-cast WMs microstructures (e.g., dendritic segregation) due to insufficient homogenization (i.e., short time for the alloying

elements redistribution during performed reaustenitization, see Figure 1). Thus, the WMs represent the most heterogeneous regions within the renormalized-and-tempered weldments. This finding indicates the reason for their high propensity for the localization of plastic deformation and final failure occurrence during tensile straining. The corresponding fracture analyses for the hydrogen-free and hydrogen-charged weldments initially processed by the PWHT-2 are shown in Figures 15 and 16, respectively.

**Figure 15.** Fracture analysis of hydrogen-free T92/T92 weldment initially processed by the PWHT-2 and subsequently ruptured in WM after the room-temperature tensile test: (**a**) fracture path and microstructure beneath the fracture; (**b**) fracture surface showing mixed fracture consisting of both ductile dimple tearing and transgranular quasi-cleavage.

**Figure 16.** Fracture analysis of hydrogen-charged T92/T92 weldment initially processed by the PWHT-2 and subsequently ruptured in WM after the room-temperature tensile test: (**a**) fracture path and microstructure beneath the fracture; (**b**) fracture surface showing mixed fracture consisting of transgranular quasi-cleavage with a "fish-eye" morphology besides ductile dimple ridges and some inter-lath decohesion areas.

The fracture paths in Figures 15a and 16a indicate lower ductility failures for both the hydrogen-free and hydrogen-charged T92/T92 weldments after the renormalizing-and-tempering PWHT-2 compared to the weldments after the tempering PWHT-1 (Figures 13a and 14a). The clearer differences between the fracture mechanisms related to the hydrogen-free and hydrogen-charged weldments after the PWHT-2 are distinguishable on the fracture surfaces in Figures 15b and 16b. Whereas the fracture surface of hydrogen-free weldment exhibits mixed fracture mechanisms including the ductile dimple tearing and transgranular quasi-cleavage (Figure 15b), the fracture surface of hydrogenated weldment shows even more complex fracture mechanisms involving transgranular quasi-cleavage with a "fish-eye" morphology besides ductile dimple tearing and some inter-lath decohesion (Figure 16b). The reason for the observed fracture morphology changes can be explained by considering local microstructural characteristics beneath the fracture surfaces. In Figures 15a and 16a the WM microstructures beneath final fractures are characterized by polygonal grains and sub-grains almost continuously decorated with coarsened secondary phase precipitates indicating thermal embrittlement induced by PWHT-2. That is why the both hydrogen-free and hydrogen-charged samples show mixed fractures consisting of both ductile dimple tearing and brittle transgranular quasi-cleavage (Figures 15b and 16b). The reason for the additional occurrence of "fish-eye" fracture morphology on the fracture surface of the hydrogenated sample treated by PWHT-2 (Figure 16b) is basically the same as in the case of hydrogenated sample after the PWHT-1. However, the higher level of thermal embrittlement after the PWHT-2 was at the same time the reason for higher grade of hydrogen embrittlement due to the existence of more nucleation sites for the creation of "fish-eye" fractographic objects.

As already shown in previous section, the hardness profile of undeformed c-w sample of T92/T92 weldment after the PWHT-2 (Figure 10) showed somewhat lower hardness in WM compared to the rest BM portions. This observation may also be related to lower alloying of WM compared to BM. Thus, the preferential deformation of softer WMs during room-temperature tensile deformation of the welds after the PWHT-2 gave rise to their higher propensity for localization of plastic deformation and final failure occurrence. It can be concluded that the use of renormalizing-and-tempering PWHT-2 was found to be rather inconvenient for improving the microstructure and mechanical properties of the investigated T92/T92 weldments. However, based on the results obtained from the cross-weld tensile tests, the hydrogen embrittlement susceptibility of the investigated weldments was rather low for both the applied PWHT conditions.
