The initiation of intergranular stress corrosion cracking (IGSCC) on Alloy 600 in a simulated pressurized water reactor primary water environment was investigated by the characterization using an electron back scattering diffraction (EBSD) on flat tensile specimens subjected to a slow strain rate testing. The IGSCC initiation was evaluated for many grain boundaries in terms of the grain boundary characteristics and the local stress generated at each grain boundary; the latter is estimated by considering the specific slip deformation determined by the Schmid factor of two grains adjacent to a grain boundary. This simplified local stress evaluation for IGSCC susceptibility based on EBSD analysis is introduced for the first time in this study. IGSCC tends to occur as a result of induced tensile stress rather than shear stress at the grain boundary, whereas no IGSCC occurred when compressive stress was applied at the grain boundary. Similar results were observed for both 10% and 20% cold worked (CW) specimens in the stress analysis. It was noted that crack initiation depends not only on the stress at the grain boundary but also on the strain concentrated around the grain boundary for the 20% CW specimen.

cold work, intergranular stress corrosion cracking, nickel alloys, pressurized water reactor, slow strain rate tests

Alloy 600 has been widely used as component materials in pressurized water reactors (PWRs), including steam generator (SG) tubes, control rod drive mechanism (CRDM), safe ends, and various instrumentation ports. Intergranular stress corrosion cracking (IGSCC) occurred on Alloy 600 in PWR primary water environments,1-3  resulting in a drastic reduction in the lifetime of components made from the alloy. This has been one of the critical issues faced by the nuclear industry.1,4-5  Although most applications have replaced Alloy 600 with thermally treated Alloy 600 (Alloy 600 TT)6-7  or more Cr containing Alloy 690, there are still some facilities using Alloy 600.6-10  Therefore, extensive efforts have been dedicated to understanding and preventing IGSCC of high nickel alloys including Alloy 600 in the PWR primary water environment. Several mechanisms such as internal oxidation,11-12  hydrogen embrittlement,13-17  and slip dissolution18-20  have been proposed for IGSCC of nickel-based alloys. In addition, it has been reported that stress,7,21-22  grain boundary characteristics,23-24  cold work (CW),25-27  temperature, and dissolved hydrogen (DH) concentration18,28-30  affect IGSCC. Although IGSCC occurs at the grain boundary, cracks do not necessarily initiate at all grain boundaries,31-32  implying that grain boundary characteristics affect the IGSCC susceptibility of high nickel alloys.

Grain boundaries can be classified as random or coincidence site lattice (CSL) boundaries. It is well known that Σ3 boundaries exhibit superior resistance to IGSCC.33-35  Gertsman and Bruemmer examined the grain boundary characteristics of austenitic alloys and concluded that only coherent twin Σ3 boundaries can be considered as special grain boundaries, and twin interactions with random boundaries may suppress crack propagation.36  Crawford and Was demonstrated that CSL boundaries in Alloy 600 are more resistant to cracks than other boundaries in argon or highly deaerated pure water.37  Moreover, Alexandreanu and Was reported that grain boundary deformation behaved as a precursor of IGSCC on a Ni-based alloy in high-temperature water.35  Hou, et al., showed that the grain boundary deformation and crack growth rate increased on a 20% CW specimen owing to the nonuniform strain concentration induced by the slip deformation and high dislocation density.27  In addition to the above-mentioned results and discussion, mechanisms such as intergranular oxidation that have been developed from the internal oxidation model proposed by Scott, et al.,11,-12  and hydrogen embrittlement have been related to the grain boundary characteristics. Bertali, et al., reported that variations in the morphology of the oxidized surface of Alloy 600 were observed at triple points of grain boundaries and at the intersection of twin and high-angle grain boundaries, indicating that the grain boundary characteristics are important in the oxidation susceptibility.38  Furthermore, the preferential intergranular oxide penetration was reported to occur along the newly migrated and solute-depleted grain boundary rather than the original grain boundary.39  In contrast, the grain boundary characteristics have been reported to affect hydrogen segregation, leading to crack initiation. The molecular dynamics and Monte Carlo simulation for Ni indicated that boundaries vicinal to the coherent twin exhibited significant changes in the structure of the boundary as well as the amount of segregated hydrogen.40  The effect of hydrogen on the resistance of CSL boundaries against crack initiation was controversial. Seita, et al., indicated that coherent twin boundaries were most susceptible to crack initiation.41  In contrast, Bechlte, et al., reported that the increase in the fraction of twin boundary enhanced the tensile ductility and fracture toughness when the hydrogen concentration in commercially pure nickel was ranged from 1,200 ppm and 3,400 ppm.42  Seita, et al., further examined crack initiation at coherent twin boundaries in hydrogen-charged Alloy 725 crytstallographically, focusing on the angle between the coherent twin boundary plane normal and the tensile axis as well as the smallest angle between a <110>-type direction in the coherent twin boundary plane and the steepest direction along the coherent twin boundary plane. They analyzed data using the Kullback-Leibler divergence and proposed probabilistic failure criteria.43  Conversely, grain boundary fracture is known to be triggered by the stress operating at the grain boundary. Therefore, both the grain boundary characteristics, as well as stress operating at grain boundary, may affect the crack initiation on Alloy 600.

The authors of this article also studied IGSCC of Alloy 600 in a simulated PWR primary environment to correlate IGSCC initiation with the grain boundary characteristics. We reported for the first time that IGSCC initiated on a flat tensile specimen of mill-annealed Alloy 600 in less than 50 h in a simulated PWR primary water environment by a slow strain rate testing (SSRT), if the specimen surface was finished with colloidal silica suspension.44-45  Then many crack initiations were characterized by a field emission-scanning electron microscope (FE-SEM) equipped with an electron backscatter diffraction (EBSD) apparatus. The results were statistically analyzed to reveal that the crack initiation probability exhibited a maximum for grain boundaries with a mis-orientation angle of 30° to 40°.44-45  Furthermore, the angle between the tensile-axis and grain-boundary plane was determined by FE-SEM and EBSD measurements at successive depths which were achieved by repeated gentle polishing and FE-SEM/EBSD observations. Through the above-mentioned experiments, we found that IGSCC tends to occur at grain boundaries with an approximate 40° grain-boundary plane angle, regardless of the misorientation angle as well as the type of grain boundary (random or CSL).45  However, in polycrystalline alloys, homogeneously loaded remote stress induces slip deformation in grains owing to shear stresses acting on a slip plane along a slip direction determined by the Schmid law in each grain. Therefore, the stress at the grain boundary is generated by the shear stresses in adjacent grains; that is, the stress generated at each grain boundary depends on the slip deformations of adjacent grains. To the best of our knowledge, IGSCC initiation, not limited on Alloy 600, has not been characterized experimentally in terms of the stress operating at the grain boundary.

The purpose of this study is to evaluate the effects of grain boundary characteristics and stress operating at grain boundaries on the IGSCC susceptibility of Alloy 600 in a simulated PWR primary water environment. We characterized the crystal orientation of many grains by using an EBSD, then determined the slip deformation direction of each grain and estimated the stress generated at the common grain boundary adjacent to the two grains. The results were statistically analyzed to discuss the IGSCC initiation mechanism.

Flat tensile Alloy 600 specimens of 2 mm thickness with a gauge section of 4 mm width and 10 mm length were fabricated from alloy sheets which were subjected to mill-annealing and subsequent CW with the reduction rates of 10% and 20%. The chemical composition of the alloy was as follows (mass%): C: 0.01, Si: 0.31, Mn: 0.36, Ni: 75.01, Cr: 15.71, Fe: 7.35, P: 0.009, and S: < 0.001. The surface of the specimens was ground using SiC abrasive papers up to #2000, and then polished with 9 μm and 1/4 μm diamond pastes, followed by mirror-finishing using a colloidal silica suspension for 20 min. This surface treatment could result in crack initiation on the Alloy 600 flat tensile specimens in a considerably short time through SSRT as previously reported.44-46  After polishing, the tensile specimens were ultrasonically cleaned using acetone, ethanol, and deionized water for 5 min, successively. Then, SSRT was performed to examine the IGSCC of Alloy 600 tensile specimens in a simulated PWR primary water environment. The simulated primary water comprised 500 ppm of B and 2 ppm of Li in the form of H3BO3 and LiOH, respectively. The DH concentration was controlled at 2.75 ppm, whereas the dissolved oxygen concentration was limited to less than 1 ppb. This condition is typical of the simulated PWR primary water environment used for laboratory tests because this condition is located near the Ni/NiO equilibrium, where Ni-based alloys are known to exhibit maximum IGSCC susceptibility.47-50  The SSRTs were performed at a temperature of 633 K and at a pressure of 20 MPa. During the SSRT, the tensile specimens were elongated up to 10% of a strain with the strain rate of 5 × 10−7 s−1. Then, the strain was maintained for 50 h and the SCC test was terminated. Oxide films, typically 20 nm to 50 nm in thickness, were formed on the surface of the specimens in the simulated PWR primary water environments as previously reported.51  The oxide films were removed by Ar+ ion beam sputtering before FE-SEM/EBSD characterizations. The sputtering was operated at 400 eV and 500 μA/cm2 for typically 90 s. This sputtering condition was determined by preliminary tests to completely remove the oxide films formed in the high-temperature aqueous solution. The EBSD measurements were performed at five different locations on the specimen surfaces for an area of 250 μm × 250 μm with a step size of 1 μm at an accelerating voltage of 25 kV. SEM images of identical locations were also obtained at the same accelerating voltage.

Figure 1 presents the distribution of the grain boundary characteristics in the tensile specimens with respect to the misorientation angle. In the figure, the random and CSL boundaries were counted separately for 10% CW and 20% CW specimens. The distributions indicate that in both specimens, a considerably large number of Σ3 grain boundaries was present. The number of CSL boundaries was in the order of Σ3 boundary > Σ9 boundary > Σ27 boundary in both specimens.
FIGURE 1.
Number of grain boundaries with respect to their misorientation angles; (a) 10% CW specimen and (b) 20% CW specimen.
View largeDownload slide

Number of grain boundaries with respect to their misorientation angles; (a) 10% CW specimen and (b) 20% CW specimen.

FIGURE 1.
Number of grain boundaries with respect to their misorientation angles; (a) 10% CW specimen and (b) 20% CW specimen.
View largeDownload slide

Number of grain boundaries with respect to their misorientation angles; (a) 10% CW specimen and (b) 20% CW specimen.

Close modal
Figure 2 shows typical SEM images and the corresponding inverse pole figure (IPF) maps of a site on the 10% CW specimen as well as on the 20% CW specimen after the SSRT. It should be noted that the oxide film formed on the specimens during SSRT was removed by Ar+ ion sputtering to obtain a clear IPF map. The horizontal direction of the images is parallel to the loading direction of the tensile stress applied to the specimens by SSRT. Apparently, many cracks initiated in the specimens. In addition, most of the cracks initiated vertically against the loading direction. The black arrows in Figure 2(a) represent the locations of the cracks, whereas the white arrows in Figure 2(b) represent identical locations in their black counterparts. This comparison indicates that cracks initiated along the grain boundaries. Similar results were also obtained for the 20% CW specimen.
FIGURE 2.
SEM image and the corresponding IPF map of (a) and (b) 10% CW specimen and (c) and (d) 20% CW specimen, after the SSRT. The arrows in (a) and (b) indicate corresponding cracked boundaries. The numbers in (c) correspond to the numbers in Figures 5(b) and 8.
View largeDownload slide

SEM image and the corresponding IPF map of (a) and (b) 10% CW specimen and (c) and (d) 20% CW specimen, after the SSRT. The arrows in (a) and (b) indicate corresponding cracked boundaries. The numbers in (c) correspond to the numbers in Figures 5(b) and 8.

FIGURE 2.
SEM image and the corresponding IPF map of (a) and (b) 10% CW specimen and (c) and (d) 20% CW specimen, after the SSRT. The arrows in (a) and (b) indicate corresponding cracked boundaries. The numbers in (c) correspond to the numbers in Figures 5(b) and 8.
View largeDownload slide

SEM image and the corresponding IPF map of (a) and (b) 10% CW specimen and (c) and (d) 20% CW specimen, after the SSRT. The arrows in (a) and (b) indicate corresponding cracked boundaries. The numbers in (c) correspond to the numbers in Figures 5(b) and 8.

Close modal

The observed cracks were crystallographically analyzed based on the EBSD analysis. To ensure the reliability of the analysis, the EBSD measurements were performed at five different sites on each specimen. All of the grain boundaries present within these sites were analyzed; however, grain boundaries of less than 5 μm were excluded from the analysis.

In the present study, the authors consider the stress generated at the grain boundary. This can be analyzed based on the crystal plasticity theory proposed by Taylor,52  considering elastic deformation until the yield point, shear strain rate, and strain hardening that exhibit anisotropy for each grain. The authors of the present study previously reported on the stress generated at the grain boundaries of an Alloy 600, which caused IGSCC in a PWR primary water environment using a multiscale finite element method (FEM).44  In the analysis the geometry and crystallographic orientation of all grains in a 100 μm × 100 μm area were determined using an FE-SEM/EBSD, then the stress and strain generated at all grain boundaries were evaluated by FEM. The results correlated with IGSCC occurrences. In the FEM analysis, to minimize geometry noise, small grains were excluded, and grain boundaries were revised to smooth lines. This approach is not straightforward and requires significant effort in preprocessing for the analysis. Additionally, a certain level of computational resources is required for the FEM. Therefore, in the present study, the authors introduced a simplified stress evaluation method for a large number of grain boundaries without considering elastic deformation and yielding.

Generally, when a polycrystalline metal or alloy is elongated beyond its elastic limit, slip deformation occurs along the slip plane and slip direction; these are, in turn, determined by the crystal orientation and stress direction for each grain. Therefore, the slip deformations along different directions in two adjacent grains can induce local stress at the grain boundary between these grains. In the present study, the initiation of IGSCC is discussed considering the stress generated at the grain boundary.

First, crystal orientations of all grains were identified using an EBSD. Subsequently, a slip system, including slip planes and slip directions, was determined for each grain. As Ni-based alloys with a face-centered cubic structure possess 12 equivalent slip systems, slip deformation basically occurs in the most deformable slip system, that is, the primary slip system. This can be identified as the one having the maximum Schmid factor. The shear stress acting on a slip plane in a slip direction within a grain is the product of the tensile stress applied to the specimen and the Schmid factor of the grain (Schmid law). For simplifying, assuming that the applied remote tensile stress is common for all grains in a specimen, the magnitude of the shear stress vector is proportional to the maximum Schmid factor in each grain when the primary slip system is active. In the deformation of polycrystalline metal, the rotation of grains, as well as that of lattice within a grain, will occur due to the restraint by adjacent grains, resulting in the deformation with multi-slip systems within a grain as well as in the grain subdivision. This causes this stress analysis more complicated. Therefore, we assume further that the rotation of grain is ignored, and a single slip system is considered to be operative. Local stresses operating to grain boundaries were estimated under these assumptions as follows.

The shear stress on a slip plane in grain is assumed to cause certain stress to grain boundary, which is in proportion to the magnitude of the shear stress. Although the stress generated at the grain boundary plane may induce IGSCC, the angle between the grain boundary plane and the shear stress direction was unknown because in the present study the grain boundary plane under the surface of the specimen was not observed. Therefore, the stress operating at the grain boundary plane could not be evaluated. In the present work, instead, the shear stress vector determined for each grain was projected onto the surface, as illustrated in Figure 3(a). The stress acting on a grain boundary at the surface may induce cleavage or slip of grain boundary, resulting in IGSCC. For these reasons, the shear stress vector projected on the surface is analyzed in the following.
FIGURE 3.
(a) Schematic illustration of the analysis of crack initiation in terms of stress operating at a grain boundary, (b) examples of shear stress scalars projected on the surfaces of the specimens, and (c) schematic illustration on an example of the decomposition of the projected shear stress vectors. Note that the shear stress vectors operating in grain 2 adjacent to grain 1 and grain 3 are in an identical magnitude but in opposite directions.
View largeDownload slide

(a) Schematic illustration of the analysis of crack initiation in terms of stress operating at a grain boundary, (b) examples of shear stress scalars projected on the surfaces of the specimens, and (c) schematic illustration on an example of the decomposition of the projected shear stress vectors. Note that the shear stress vectors operating in grain 2 adjacent to grain 1 and grain 3 are in an identical magnitude but in opposite directions.

FIGURE 3.
(a) Schematic illustration of the analysis of crack initiation in terms of stress operating at a grain boundary, (b) examples of shear stress scalars projected on the surfaces of the specimens, and (c) schematic illustration on an example of the decomposition of the projected shear stress vectors. Note that the shear stress vectors operating in grain 2 adjacent to grain 1 and grain 3 are in an identical magnitude but in opposite directions.
View largeDownload slide

(a) Schematic illustration of the analysis of crack initiation in terms of stress operating at a grain boundary, (b) examples of shear stress scalars projected on the surfaces of the specimens, and (c) schematic illustration on an example of the decomposition of the projected shear stress vectors. Note that the shear stress vectors operating in grain 2 adjacent to grain 1 and grain 3 are in an identical magnitude but in opposite directions.

Close modal
Figure 3(b) presents examples of the distribution of the projected shear stress direction on the surface of the 10% and 20% CW specimens after SSRT. In these figures, the direction is described not as a shear stress vector but as a scalar. Noted that the shear stress induced on a slip plane of a grain will act in opposite directions for two grain boundaries surrounding one grain based on the continuum mechanics, as illustrated in grain 2 of Figure 3(c). As shown in Figure 3(c), the projected shear stress vector, ⁠, was decomposed into mutually perpendicular components, i.e., one is parallel to the grain boundary, whereas the other is vertical to the grain boundary. In the figure, the parallel component of a decomposed vector is expressed as ⁠, and the vertical one is expressed as ⁠. The stress operating at a grain boundary differs depending on the type of decomposed vectors (parallel or vertical) and their directions. Figure 4 illustrates four cases whereby the decomposed vectors generated in two grains act differently at the common grain boundary. Regarding the parallel vectors shown in Figure 4(a), a large shear stress operates at the grain boundary between adjacent grains when the directions of the two vectors are opposite. This can induce a crack initiation. However, the shear stress operating at the grain boundary is small when the directions are the same, as illustrated in the lower case of Figure 4(a). In particular, if the two vectors are the same, no shear stress occurs at the grain boundary. In the case of vertical vectors, their combinations are complicated as illustrated in Figures 4(b) through (d). For the case shown in Figure 4(b), the directions of the vectors are opposite to each other, similar to those in Figure 3(c). This causes a tensile stress at the grain boundary between the grains. A cleavage of the grain boundary may then result, that is, crack initiation. By contrast, for the case shown in Figure 4(c), although the directions of the vectors are opposite, the vectors face each other across the grain boundary. This causes a compressive stress at the grain boundary. Cracks may scarcely initiate in this case. Furthermore, for the case shown in Figure 4(d), where the directions of the vertical vectors are the same, the type of stress operating at the grain boundary differs depending on the magnitude of each vertical vector. For simplicity, only the situations where both vectors point to the right are considered, as illustrated in Figure 4(d). When the vertical vector on grain 2 is smaller than that on grain 1, a compressive stress operates at the grain boundary. However, in the opposite case, a tensile stress operates at the grain boundary, leading to crack initiation. The stress at grain boundaries is analyzed according to the aforementioned scenario.
FIGURE 4.
Schematic illustration of the combinations of decomposed stress vectors: (a) parallel to grain boundary and (b) through(d) perpendicular to grain boundary.
View largeDownload slide

Schematic illustration of the combinations of decomposed stress vectors: (a) parallel to grain boundary and (b) through(d) perpendicular to grain boundary.

FIGURE 4.
Schematic illustration of the combinations of decomposed stress vectors: (a) parallel to grain boundary and (b) through(d) perpendicular to grain boundary.
View largeDownload slide

Schematic illustration of the combinations of decomposed stress vectors: (a) parallel to grain boundary and (b) through(d) perpendicular to grain boundary.

Close modal
Figures 5(a) and (b) summarize the distribution of the type of stress operating at grain boundaries in the 10% and 20% CW specimens, respectively. This is evaluated after the SSRT in the simulated PWR environment with a DH concentration of 2.75 ppm and at 633 K. The values indicated in the horizontal and vertical axes are the resolved stress intensities, assuming that the remote tensile stress loaded on the specimen is unity. The solid marks in the figures indicate cracked grain boundaries. The figures were obtained using data based on the analysis at one site shown in Figure 3(b). As many cracked grain boundaries are present in the figures, crack initiation is statistically discussed. It is clear that in both specimens, cracks initiated at the grain boundaries where a large tensile stress operated, whereas no cracks initiated at the grain boundaries where compressive stress operated. This is rational because the tensile stress contributes to the cleaving of the grain boundary, whereas the compressive one does not. To quantitatively assess the crack susceptibility, the probability of crack initiation was evaluated with respect to each stress factor; the results are presented in Figure 6. As expected from Figure 5, the probability of crack initiation is 0 when compressive stress operates on grain boundaries (not described in Figure 6), i.e., the cleaving of the grain boundary is not triggered by the compressive stress. The probability tends to increase with increasing tensile factors, whereas the probability of crack initiation induced by the shear stress decreases for shear factors ranging from 0.6 to 1. These analyses indicate that crack initiation can be triggered mainly by the tensile stress.
FIGURE 5.
Correlation of various stresses operating at the grain boundary: (a) 10% CW specimen and (b) 20% CW specimen. Solid marks indicate the grain boundaries where the crack initiated. The numbers indicated in (b) correspond to the ones in Figures 2(c) and 8; 0 to 7 are cracked, and 8 to 14 are not cracked.
View largeDownload slide

Correlation of various stresses operating at the grain boundary: (a) 10% CW specimen and (b) 20% CW specimen. Solid marks indicate the grain boundaries where the crack initiated. The numbers indicated in (b) correspond to the ones in Figures 2(c) and 8; 0 to 7 are cracked, and 8 to 14 are not cracked.

FIGURE 5.
Correlation of various stresses operating at the grain boundary: (a) 10% CW specimen and (b) 20% CW specimen. Solid marks indicate the grain boundaries where the crack initiated. The numbers indicated in (b) correspond to the ones in Figures 2(c) and 8; 0 to 7 are cracked, and 8 to 14 are not cracked.
View largeDownload slide

Correlation of various stresses operating at the grain boundary: (a) 10% CW specimen and (b) 20% CW specimen. Solid marks indicate the grain boundaries where the crack initiated. The numbers indicated in (b) correspond to the ones in Figures 2(c) and 8; 0 to 7 are cracked, and 8 to 14 are not cracked.

Close modal
FIGURE 6.
Distributions on the probability of crack initiation against different factors operating at grain boundary for (a) and (b) 10% CW specimen and (c) and (d) 20% CW specimen. Error bars indicate the width of the standard deviation.
View largeDownload slide

Distributions on the probability of crack initiation against different factors operating at grain boundary for (a) and (b) 10% CW specimen and (c) and (d) 20% CW specimen. Error bars indicate the width of the standard deviation.

FIGURE 6.
Distributions on the probability of crack initiation against different factors operating at grain boundary for (a) and (b) 10% CW specimen and (c) and (d) 20% CW specimen. Error bars indicate the width of the standard deviation.
View largeDownload slide

Distributions on the probability of crack initiation against different factors operating at grain boundary for (a) and (b) 10% CW specimen and (c) and (d) 20% CW specimen. Error bars indicate the width of the standard deviation.

Close modal
In the present study, the initiation of IGSCC on Alloy 600 in a simulated PWR water environment with a DH of 2.75 ppm and at 633 K is characterized. It is assumed that the tensile plastic deformation during the SSRT induces stress at the grain boundary, resulting in its cleavage. The induced stress at a grain boundary is resolved into local tensile, compressive, or shear stresses. It is confirmed that among these types of stresses, the tensile stress mainly contributes to the cleaving of grain boundary, whereas the compressive stress never causes cleaving as described in Figure 5. Furthermore, as shown in Figure 6, the probability of crack initiation increases with increasing tensile stress. However, the relation between shear stress at the grain boundary and the probability of crack initiation is not clear. These results indicate that tensile stress is more effective than shear stress in the cleaving of the grain boundary. Conversely, the decreased probability at a higher range of shear factors might be attributed to the small tensile stresses operating at the grain boundaries. This is because shear stress tends to decrease with increasing tensile stress as shown in Figure 5. In order to discuss the grain boundary cleavage under shear deformation in grains, the motion of dislocations should be considered. Dislocation pile-up and annihilation at grain boundary due to shear deformation induce defects that might initiate grain boundary cracking. Therefore, microcharacterizations of slip deformation are required for further discussion. However, slip bands were not always visible at all grains examined: therefore, the actual slip deformation cannot be discussed in the present study. The distributions of the tensile and shear stresses described in Figure 5 were summarized for one location of each specimen. The cracked boundaries obtained from five different locations for each specimen are plotted as stress distributions in Figure 7. It was clarified that some Σ3 boundaries cracked when a relatively large tensile stress was operating. As apparent from Figures 7(a) and (b), more Σ3 boundaries cracked than the other CSL boundaries although the probability of crack initiation on Σ3 boundaries was reported to be low compared with the other CSL boundaries.45,53,-54  This is attributed to the aforementioned findings that the number of CSL boundaries differs. As presented in Figure 1, most grain boundaries of specimens used in the present study were characterized as the Σ3 boundary. Therefore, it may be rational to conclude that some Σ3 boundaries cracked only when a larger tensile stress was operated.
FIGURE 7.
Stress distribution at cracked (a) and (b) CSL boundaries and (c) and (d) random boundaries obtained for (a) and (c) 10% CW and (b) and (d) 20% CW specimens.
View largeDownload slide

Stress distribution at cracked (a) and (b) CSL boundaries and (c) and (d) random boundaries obtained for (a) and (c) 10% CW and (b) and (d) 20% CW specimens.

FIGURE 7.
Stress distribution at cracked (a) and (b) CSL boundaries and (c) and (d) random boundaries obtained for (a) and (c) 10% CW and (b) and (d) 20% CW specimens.
View largeDownload slide

Stress distribution at cracked (a) and (b) CSL boundaries and (c) and (d) random boundaries obtained for (a) and (c) 10% CW and (b) and (d) 20% CW specimens.

Close modal
In the case of random boundaries, the number of crack initiation exhibits different trends depending on the degree of CW as shown in Figures 7(c) and (d). The number of cracks for the 10% CW specimen was the largest at around 40° and the number decreased with decreasing the misorientation angle to 20°, as well as increasing the angle to 60°. Chen, et al., described that the grain boundary energy of the extended Read-Shockley model reached a maximum at a misorientation angle of 45°.55  They reported that cracks mainly occurred at grain boundaries with a misorientation angle ranging from 30° to 45° because the grain boundaries with large grain boundary energy are more likely to crack. This is approximately in accordance with the present results obtained for the 10% CW specimen. In addition, Figure 7(c) indicates that the crack initiation occurs even at grain boundaries out of the misorientation angle range indicated by Chen, et al., when a larger tensile stress operates at the boundary. Conversely, as presented in Figure 7(d), the number of crack initiation for the 20% CW specimen was the largest at 50°. Although this sounds similar to the case of the 10% CW specimen, the misorientation angle dependence does not necessarily seem obvious for the 20% CW specimen. In general, CW introduces residual stress and strain in materials; therefore, more strains are introduced in 20% CW specimens than in 10% CW specimens. Figure 8 shows a kernel average misorientation (KAM) map for the 20% CW specimen, the corresponding SEM image of which was presented in Figure 2(c). In the figure, the arrows numbered 1 to 7 represent the locations of the cracked random boundaries, whereas those from 8 to 14 represent the locations of random boundaries without cracking (the numbers are also indicated in Figure 5[b]). Figure 8 further shows the locally concentrated strain along the grain boundaries and demonstrates that cracks initiate at the grain boundaries where a strong strain is present. This may imply that the strains introduced by CW-induced crack initiation, whereas no cracks were observed at the grain boundaries without the concentrated strains, even if a large tensile or shear stress was operating. Therefore, it is confirmed that the crack initiation in 20% CW specimens is triggered not only by the stress operating at the grain boundary but also by the strain concentrated at the grain boundary. It is well-known that the local strains observed in the KAM map are directly linked to the local flow stress. Therefore, according to the aforementioned discussion, the grain boundaries where strong flow stress was generated roughly correspond to those where the calculated stresses are large. This indicates that the stress analysis proposed in the present work reflects the deformation behavior of the alloy examined although the calculations were performed under several assumptions including homogeneous stress distribution within a single grain. As these assumptions might cause the result that the stresses operating at grain boundaries do not necessarily result in the crack initiation.
FIGURE 8.
KAM obtained for 20% CW specimen after performing SSRT with the DH of 2.75 ppm, corresponding to the identical area shown in Figures 2(c) and (d). The numbers correspond to the numbers in Figures 2(c) and 5(b). Note that locations numbered as 0 to 7 cracked, and 8 to 14 not cracked.
View largeDownload slide

KAM obtained for 20% CW specimen after performing SSRT with the DH of 2.75 ppm, corresponding to the identical area shown in Figures 2(c) and (d). The numbers correspond to the numbers in Figures 2(c) and 5(b). Note that locations numbered as 0 to 7 cracked, and 8 to 14 not cracked.

FIGURE 8.
KAM obtained for 20% CW specimen after performing SSRT with the DH of 2.75 ppm, corresponding to the identical area shown in Figures 2(c) and (d). The numbers correspond to the numbers in Figures 2(c) and 5(b). Note that locations numbered as 0 to 7 cracked, and 8 to 14 not cracked.
View largeDownload slide

KAM obtained for 20% CW specimen after performing SSRT with the DH of 2.75 ppm, corresponding to the identical area shown in Figures 2(c) and (d). The numbers correspond to the numbers in Figures 2(c) and 5(b). Note that locations numbered as 0 to 7 cracked, and 8 to 14 not cracked.

Close modal

It has been widely accepted that IGSCC of Alloy 600 in the PWR primary water environment proceeds as internal oxidation along the grain boundary.8,10-11,18,56-57  A stress operating at a grain boundary can break the oxide at the grain boundary, resulting in the subsequent exposure of the underlying grain boundary to the environment. This induces crack initiation along the grain boundary. Furthermore, the grain boundary characteristics are reported to affect intergranular oxidation; oxidation proceeded at random grain boundaries, whereas it was suppressed at the CSL boundaries.58  This difference in the degree of oxidation may be one reason for the probability distribution of crack initiation among various grain boundary characteristics other than the variation in stress operating at grain boundaries.

Based on the simple stress analysis introduced for the first time in this study, intergranular cracking can be categorized into two groups. In one group, the crack initiation depends on the stress distribution as well as the misorientation angle around the grain boundary; in the other case, the crack initiation depended not only on the stress distribution but also on the strain concentrated around the grain boundary.

  • In the present study, the authors examined the crack initiation of Alloy 600 in a simulated PWR primary water environment with a DH concentration of 2.75 ppm and at 633 K. Tensile specimens subjected to CW with reduction rates of 10% or 20% in advance were elongated up to a tensile strain of 10% in a SSRT, and the strain was maintained for 50 h. The surface of the specimen was microcrystallographically examined.

  • Many IGSCCs were observed on the surface of the specimen after SSRT and were characterized using an FE-SEM/EBSD. The stresses operating at the grain boundaries between two adjacent grains during the SSRTs were analyzed to resolve tensile, compressive, and shear stresses, assuming that the primary slip system was active, and that slip deformation in two adjacent grains generated stress at their common grain boundary. It is confirmed that IGSCC preferentially occurred at the grain boundary with a larger tensile stress, which contributed to the cleaving of the grain boundary. No cracks were observed at the grain boundaries with compressive stress. Furthermore, crack initiation also depends on the degree of strain concentration, which was characterized by KAM at the grain boundary for the 20% CW specimen.

A part of this study was performed as the project on “Enhancement of Ageing Management and Maintenance of Nuclear Power Station” sponsored by the Nuclear and Industry Safety Agency, Minister of Economy, Trade and Industry. The authors also appreciate Professor Masahito Mochizuki and Professor Yoshiki Mikami in Osaka University for a helpful discussion.

1.
Cattant
 
F.
,
Crusset
 
D.
,
Féron
 
D.
,
Mater. Today
11
(
2008
):
p
.
32
–
37
.
Google Scholar
Crossref
Search ADS
 
2.
Hwang
 
S.S.
,
Lim
 
Y.S.
,
Kim
 
S.W.
,
Kim
 
D.J.
,
Kim
 
H.P.
,
Nucl. Eng. Technol.
45
(
2013
):
p
.
73
–
80
.
Google Scholar
Crossref
Search ADS
 
3.
Scenini
 
F.
,
Newman
 
R.C.
,
Cottis
 
R.A.
,
Jacko
 
R.J.
,
Corrosion
64
(
2008
):
p
.
824
–
835
.
Google Scholar
Crossref
Search ADS
 
4.
Allen
 
T.
,
Busby
 
J.
,
Meyer
 
M.
,
Petti
 
D.
,
Mater. Today
13
(
2010
):
p
.
14
–
23
.
Google Scholar
Crossref
Search ADS
 
5.
Zinkle
 
S.J.
,
Was
 
G.S.
,
Acta Mater.
61
(
2013
):
p
.
735
–
758
.
Google Scholar
Crossref
Search ADS
 
6.
Airey
 
G.P.
, “
Optimization of Metallurgical Variables to Improve the Stress Corrosion Resistance of Inconel 600
,”
EPRI report, NP-1354
,
1980
.
Google Scholar
 
7.
Bruemmer
 
S.M.
,
Was
 
G.S.
,
J. Nucl. Mater.
216
(
1994
):
p
.
348
–
363
.
Google Scholar
Crossref
Search ADS
 
8.
Arioka
 
K.
,
Corrosion
71
(
2015
):
p
.
403
–
419
.
Google Scholar
Crossref
Search ADS
 
9.
Zhai
 
Z.
,
Toloczko
 
M.B.
,
Olszta
 
M. J.
,
Bruemmer
 
S.M.
,
Corros. Sci.
123
(
2017
):
p
.
76
–
87
.
Google Scholar
Crossref
Search ADS
 
10.
Scott
 
P.M.
,
Combrade
 
P.
,
J. Nucl. Mater.
524
(
2019
):
p
.
340
–
375
.
Google Scholar
Crossref
Search ADS
 
11.
Scott
 
P.M.
, “
An Overview of Internal Oxidation as a Possible Explanation of Intergranular Stress Corrosion Cracking of Alloy 600 in PWRS
,”
in
Proc. 9th Int. Conf. on Environmental Degradation of Materials in Nuclear Power Systems—Water Reactors
,
eds.
Ford
 
F.P.
,
Bruemmer
 
S.M.
,
Was
 
G.S.
 (
Pittsburgh, PA
:
The Minerals, Metals, and Materials Society
,
1999
),
p
.
3
–
14
.
Google Scholar
Crossref
Search ADS
 
12.
Gendron
 
T.S.
,
Scott
 
P.M.
,
Bruemmer
 
S.M.
,
Thomas
 
L.E.
, “
Internal Oxidation as a Mechanism for Steam Generator Tube Degradation
,”
in
Proc. Third Int. Steam Gener. Heat Exch. Conf.
(
Toronto, Canada
:
Canadian Nuclear Society
,
1998
),
p
.
389
–
406
.
Google Scholar
 
13.
Lecoester
 
F.
,
Chêne
 
J.
,
Noel
 
D.
,
Mater. Sci. Eng. A
262
(
1999
):
p
.
173
–
183
.
Google Scholar
Crossref
Search ADS
 
14.
Chêne
 
J.
,
Brass
 
A.M.
,
Metall. Mater. Trans. A
35
(
2004
):
p
.
457
–
464
.
Google Scholar
Crossref
Search ADS
 
15.
Hicks
 
P.D.
,
Altstetter
 
C.J.
,
Metall. Trans. A
21
(
1990
):
p
.
365
–
372
.
Google Scholar
Crossref
Search ADS
 
16.
Hicks
 
P.D.
,
Altstetter
 
C.J.
,
Metall. Trans. A
23
(
1992
):
p
.
237
–
249
.
Google Scholar
Crossref
Search ADS
 
17.
Liu
 
L.
,
Zhai
 
C.
,
Lu
 
C.
,
Ding
 
W.
,
Hirose
 
A.
,
Kobayashi
 
K.F.
,
Corros. Sci.
47
(
2005
):
p
.
355
–
367
.
Google Scholar
Crossref
Search ADS
 
18.
Andresen
 
P.L.
,
Peter Ford
 
F.
,
Mater. Sci. Eng.
103
(
1988
):
p
.
167
–
184
.
Google Scholar
Crossref
Search ADS
 
19.
Hall
 
M.M.
,
Corros. Sci.
51
(
2009
):
p
.
1103
–
1106
.
Google Scholar
Crossref
Search ADS
 
20.
Hall
 
M.M.
,
Corros. Sci.
51
(
2009
):
p
.
225
–
233
.
Google Scholar
Crossref
Search ADS
 
21.
Clarke
 
W.L.
,
Gordon
 
G.M.
,
Corrosion
29
(
1973
):
p
.
1
–
12
.
Google Scholar
Crossref
Search ADS
 
22.
Bruemmer
 
S.M.
,
Charlot
 
L.A.
,
Henager
 
C.H.
,
Corrosion
44
(
1988
):
p
.
782
–
788
.
Google Scholar
Crossref
Search ADS
 
23.
Lim
 
Y.S.
,
Kim
 
H.P.
,
Hwang
 
S.S.
,
J. Nucl. Mater.
440
(
2013
):
p
.
46
–
54
.
Google Scholar
Crossref
Search ADS
 
24.
Magnin
 
T.
,
Noël
 
D.
,
Rios
 
R.
,
Mater. Sci. Eng. A
177
(
1994
):
p
.
11
–
14
.
Google Scholar
Crossref
Search ADS
 
25.
Moshier
 
W.C.
,
Brown
 
C.M.
,
Corrosion
56
(
2000
):
p
.
307
–
320
.
Google Scholar
Crossref
Search ADS
 
26.
Yamazaki
 
S.
,
Lu
 
Z.
,
Ito
 
Y.
,
Takeda
 
T.
,
Shoji
 
T.
,
Corros. Sci.
50
(
2008
):
p
.
835
–
846
.
Google Scholar
Crossref
Search ADS
 
27.
Hou
 
J.
,
Peng
 
Q.J.
,
Lu
 
Z.P.
,
Shoji
 
T.
,
Corros. Sci.
53
(
2011
):
p
.
1137
–
1142
.
Google Scholar
Crossref
Search ADS
 
28.
Takiguchi
 
H.
,
Ullberg
 
M.
,
Uchida
 
S.
,
J. Nucl. Sci. Technol.
41
(
2004
):
p
.
601
–
609
.
Google Scholar
Crossref
Search ADS
 
29.
Nakagawa
 
T.
,
Totsuka
 
N.
,
Terachi
 
T.
,
Nakajima
 
N.
,
J. Nucl. Sci. Technol.
40
(
2003
):
p
.
39
–
43
.
Google Scholar
Crossref
Search ADS
 
30.
Morton
 
D.S.
,
Attanasio
 
S.
,
Richey
 
E.
,
Young
 
G.
, “
In Search of the True Temperature and Stress Intensity Factor Dependencies for PWSCC
,”
in
Proc. Twelfth Int. Conf. Environ. Degrad. Mater. Nucl. Power Syst.—Water React.
(
Salt Lake City, UT
:
The Minerals, Metals, and Materials Society
,
2005
),
p
.
977
–
988
.
Google Scholar
 
31.
Parkins
 
R.N.
,
JOM
44
(
1992
):
p
.
12
–
19
.
Google Scholar
Crossref
Search ADS
 
32.
Shibata
 
T.
,
Iron Steel Inst. Jpn.
31
(
1991
):
p
.
115
–
121
.
Google Scholar
Crossref
Search ADS
 
33.
Palumbo
 
G.
,
King
 
P.J.
,
Aust
 
K.T.
,
Erb
 
U.
,
Lichtenberger
 
P.C.
,
Scr. Metall. Mater.
25
(
1991
):
p
.
1775
–
1780
.
Google Scholar
Crossref
Search ADS
 
34.
Aust
 
K.T.
,
Erb
 
U.
,
Palumbo
 
G.
,
Mater. Sci. Eng. A
176
(
1994
):
p
.
329
–
334
.
Google Scholar
Crossref
Search ADS
 
35.
Alexandreanu
 
B.
,
Was
 
G.S.
,
Corrosion
59
(
2003
):
p
.
705
–
720
.
Google Scholar
Crossref
Search ADS
 
36.
Gertsman
 
V.Y.
,
Bruemmer
 
S.M.
,
Acta Mater.
49
(
2001
):
p
.
1589
–
1598
.
Google Scholar
Crossref
Search ADS
 
37.
Crawford
 
D.C.
,
Was
 
G.S.
,
Metall. Trans. A
23
(
1992
):
p
.
1195
–
1206
.
Google Scholar
Crossref
Search ADS
 
38.
Bertali
 
G.
,
Scenini
 
F.
,
Burke
 
M.G.
,
Corros. Sci.
111
(
2016
):
p
.
494
–
507
.
Google Scholar
Crossref
Search ADS
 
39.
Bertali
 
G.
,
Scenini
 
F.
,
Burke
 
M.G.
,
Corros. Sci.
100
(
2015
):
p
.
474
–
483
.
Google Scholar
Crossref
Search ADS
 
40.
O’Brien
 
C.J.
,
Foiles
 
S.M.
,
Philos. Mag.
96
(
2016
):
p
.
1463
–
1484
.
Google Scholar
Crossref
Search ADS
 
41.
Seita
 
M.
,
Hanson
 
J.P.
,
Gradečak
 
S.
,
Demkowicz
 
M.J.
,
Nat. Commun.
6
(
2015
):
p. 6164
.
Google Scholar
Crossref
Search ADS
 
42.
Bechtle
 
S.
,
Kumar
 
M.
,
Somerday
 
B.P.
,
Launey
 
M.E.
,
Ritchie
 
R.O.
,
Acta Mater.
57
(
2009
):
p
.
4148
–
4157
.
Google Scholar
Crossref
Search ADS
 
43.
Seita
 
M.
,
Hanson
 
J.P.
,
Gradečak
 
S.
,
Demkowicz
 
M.J.
,
J. Mater. Sci.
52
(
2017
):
p
.
2763
–
2779
.
Google Scholar
Crossref
Search ADS
 
44.
Fujimoto
 
S.
,
Mochizuki
 
M.
,
Morita
 
Y.
,
Mikami
 
Y.
,
Tsuchiya
 
H.
,
Nishimoto
 
K.
, “
Crystallographical Characterization of Initiation of Intergranular Stress Corrosion Cracking of Alloy 600 in PWR Environment
,”
in
Proc. Fifteenth Int. Conf. Environ. Degrad. Mater. Nucl. Power Syst.—Water React.
(
Colorado Springs, CO
:
The Minerals, Metals, and Materials Society
,
2011
),
p
.
1685
–
1696
.
Google Scholar
 
45.
Jung
 
K.T.
,
Tsuchiya
 
H.
,
Fujimoto
 
S.
,
Mater. Tarns.
61
(
2020
):
p
.
1339
–
1345
.
Google Scholar
Crossref
Search ADS
 
46.
Jung
 
K.-T.
,
Ogawa
 
T.
,
Morita
 
Y.
,
Fujimoto
 
S.
, “
Grain Boundary Characterization for Initial Stage of IGSCC of Alloy 600 in PWR Primary Water Environment
,”
in
Proc. Seventeenth Int. Conf. Environ. Degrad. Mater. Nucl. Power Syst.—Water React.
(
Ottawa, Canada
:
Canadian Nuclear Society
,
2015
),
p
.
895
–
903
.
Google Scholar
 
47.
Andresen
 
P.L.
,
Hickling
 
J.
,
Ahluwalia
 
A.
,
Wilson
 
J.
,
Corrosion
64
(
2008
):
p
.
707
–
720
.
Google Scholar
Crossref
Search ADS
 
48.
Etien
 
R.A.
,
Richey
 
E.
,
Morton
 
D.S.
,
Eager
 
J.
, “
SCC Initiation Testing of Alloy 600 in High Temperature Water
,”
in
Proc. 15th Int. Conf. Environ. Degrad. Mater. Nucl. Power Syst.—Water React.
(
Colorado Springs, CO
:
The Minerals, Metals, and Materials Society
,
2011
),
p
.
2274
–
2284
.
Google Scholar
 
49.
Moss
 
T.
,
Was
 
G.S.
,
J. Electrochem. Soc.
162
(
2015
):
p
.
C35
–
C41
.
Google Scholar
Crossref
Search ADS
 
50.
Rickey
 
E.
,
Morton
 
D.S.
,
Etien
 
R.A.
, “
SCC Initiation Testing of Alloy 600in High Temperature Water
,”
in
Proc. 13rd Int. Conf. Environ. Degrad. Mater. Nucl. Power Syst.—Water React.
(
Whistler, Canada
:
Canadian Nuclear Society
,
2007
),
p
.
831
–
850
.
Google Scholar
 
51.
Fujimoto
 
S.
,
Kim
 
W.-S.
,
Sato
 
M.
,
Son
 
J.-S.
,
Machida
 
M.
,
Jung
 
K.-T.
,
Tsuchiya
 
H.
,
J. Solid State Electrochem.
19
(
2015
):
p
.
3521
–
3531
.
Google Scholar
Crossref
Search ADS
 
52.
Taylor
 
G.I.
,
J. Inst. Met.
62
(
1938
):
p
.
307
–
324
.
Google Scholar
 
53.
Lu
 
Z.-P.
,
Shoji
 
T.
,
Yamazaki
 
S.
,
Ogawa
 
K.
,
Corros. Sci.
(
2013
):
p
.
211
–
228
.
54.
Sakaguchi
 
N.
,
Ohguchi
 
Y.
,
Shibayama
 
T.
,
Watanabe
 
S.
,
Kinoshita
 
H.
,
J. Nucl. Mater.
432
(
2013
):
p
.
23
–
27
.
Google Scholar
Crossref
Search ADS
 
55.
Chen
 
J.Q.
,
Lu
 
H.
,
Cui
 
W.
,
Chen
 
J.M.
,
Huang
 
Y.F.
,
Mater. Sci. Technol.
30
(
2014
):
p
.
1189
–
1196
.
Google Scholar
Crossref
Search ADS
 
56.
Arioka
 
K.
,
Yamada
 
T.
,
Miyamoto
 
T.
,
Aoki
 
M.
,
Corrosion
70
(
2014
):
p
.
695
–
707
.
Google Scholar
Crossref
Search ADS
 
57.
Scott
 
P.M.
,
Le Calvar
 
M.
, “
Some Possible Mechanisms of Intergranular Stress Corrosion Cracking of Alloy 600 in PWR Primary Water
,”
in
Proc. 6th Int. Conf. on Environmental Degradation of Materials in Nuclear Power Systems—Water Reactors
,
ed.
,
Simonen
 
E.P.
,
held 1993
(
Warrendale, PA
:
The Minerals, Metals, and Materials Society
,
1993
),
p
.
657
–
665
.
Google Scholar
 
58.
Lim
 
Y.-S.
,
Kim
 
S.-W.
,
Hwang
 
S.-S.
,
Kim
 
H.P.
,
Jang
 
C.
,
Corros. Sci.
108
(
2016
):
p
.
125
–
133
.
Google Scholar
Crossref
Search ADS
 
© 2022, AMPP
2022