2020 Volume 61 Issue 6 Pages 1138-1142
Under the primary conditions of a pressurized water reactor, the intergranular stress corrosion cracking susceptibility of a Ni based alloy is improved by chromium carbide precipitation in grain boundaries. The effects of chromium carbide precipitation treatments are explained by the grain boundary strengthening mechanism, the intergranular corrosion prevention mechanism, and so on. However, few studies have demonstrated these mechanisms, and the effects on intergranular stress corrosion cracking susceptibility are not completely understood. Therefore, for the purpose of demonstrating the change in the internal stress distribution with or without grain boundary carbide precipitation treatment in the 600 alloy, in situ measurements were performed in this study using the X-ray diffraction technique with the BL-28B2 beamline at the SPring-8 synchrotron radiation facility under tensile stress. The results confirmed that there is a difference in the strain distribution in grain boundaries and trans-grains. Intergranular stress corrosion cracking is suppressed by preventing stress concentration in grain boundaries owing to the carbides precipitated at the grain boundaries.
This Paper was Originally Published in Japanese in J. Japan Inst. Met. Mater. 83 (2019) 54–58.
Fig. 6 Difference equivalent stress, $\text{d}\sigma _{e}$, of a difference from the equivalent stress without the applied stress and the equivalent stress ratio, $\alpha = \frac{\sigma _{e}(GB)}{\sigma _{e}(TG)}$, as a function of the applied stress at the grain boundary (GB) and trans-grain (TG). (a) Untreated specimen and (b) thermally treated specimen.
A Ni-based alloy is applied to the steam generator (SG) tube of the pressurized water reactors (PWRs). The intergranular stress corosion cracking (IGSCC) of SG tubes is recognized as an important problem that affects the safe and stable operation of nuclear power plants.1,2) It is well known that a thermal treatment for precipitating carbides in grain boundaries, which is referred to as a TT treatment,3) is an effective measure for suppressing IGSCC generation and progress. The TT treatment with the precipitation of M23C6 and M7C3 type carbides at grain boundaries shows good performance in improving the resistance of TT-treated 600 690 alloys (TT600 and TT690) to IGSCC.3) Hence, these alloys are frequently used in high-temperature water environments.
The mechanism of the improvement of IGSCC resistance with grain boundary carbide precipitation has been discussed in the following aspects: the changes in the intergranular corrosion properties due to the sensitization by the deficiency of chromium,4) the changes in mechanical properties close to grain boundaries caused by dispersing strain concentration,5) and the strengthening of grain boundaries.6) However, as the TT690 alloy recommended for PWR use has not experienced IGSCC in PWR environments, detailed discussions have not been carried out. As a result, the mechanism of improving IGSCC resistance by grain boundary carbide precipitation has not been elucidated. In this study, the internal stress distribution of 600 alloys with and without the TT treatment was investigated using energy-dispersive X-ray diffraction microscopy7) (EXDM) measurements, which were carried out at the SPring-8 synchrotron radiation facility in Japan, to clarify the improvement in the mechanical properties of the alloys by TT treatment.
The chemical composition of samples is shown in Table 1. The samples were heated at 1150°C for 10 min and then cooled with water. A sample was used as the TT-untreated specimen. Another sample was treated at 700°C for 15 h and used as the TT-treated specimen. It has been confirmed that Cr23C6 precipitates at grain boundaries and IGSCC resistance is improved after applying this thermal treatment.8)
In these processes, the grain size was adjusted to approximately 100–200 µm. The TT-untreated and TT-treated specimens were formed into tensile specimens with a gauge length of 3 mm and a width of 1 mm, as shown in Fig. 1. Then, mechanical polishing and colloidal silica polishing were performed to obtain thin samples with a thickness of 100 µm.
Schematic view of tensile specimen.
Electron backscatter diffraction (EBSD) data are freaquently applied to evaluate the strain distribution of the fine structure of materials.9) However, as the depth detection by the EBSD method is limited only to surfaces with a depth of several nanometers, the strain distribution inside materials has not been evaluated. The EXDM method7) has been developed at SPring-8 for examining the inner strain distribution of specimens. The EXDM analyzes strain based on the lattice spacing of an X-ray irradiation position using the X-ray diffraction pattern in a transmission configuration. In this method, X-ray beam size is controlled to be smaller than the grain size because the irradiation position must be considered as a single crystal. The EXDM measurement was carried out on the BL-28B2 beamline at SPring-8. The white X-ray beam on the BL-28B2 can be formed into 10 µm × 10 µm which was sufficiently small for the crystal grain size.
The specimens were set on a small tensile tester with no load, and the tensile tester was installed on the sample stage of a multiaxis diffractometer on BL-28B2. Grain boundary image information was acquired from the transmitted Laue patterns obtained by imaging the specimens with the X-ray beam.10) Figures 2(a) and (b) show the grain boundary images of the TT-untreated and TT-treated specimens, respectively.
Grain boundary images visualized from the specimen. The images includes the measurement positions. Red lines are the grain boundaries. Red points indicate measurement points. Grain boundary images of (a) TT-untreated 600 alloy and (b) TT-treated 600 alloy.
The EXDM measurements were performed according to the following procedure. Measurement points were selected with reference to the obtained grain boundary images. The measurement points indicated by the red dots in Fig. 2(a) and (b) were extracted from the grain boundary and trans-grain of each specimen. Sixty points were selected for the TT-untreated specimen, among which 36 points were around the grain boundary and 24 points were in the trans-grain. In the case of the TT-treated specimen, fifty points were selected. Twenty five out of this 56 points were around the grain boundary and the other 25 points were in the trans-grain. Each measurement points was irradiated with the white X-rays beam to obtain transmitted Laue images. Twenty strong spots were selected in each transmitted Laue image. Thereafter, a diffraction pattern was obtained from each spot using a semiconductor detector. The same measurements were performed by applied stress within the elastic strain range determined by stress-strain curves, as shown in Fig. 3. Figure 3 shows the stress-strain curve of the sample prepared via TT treatment from 0 to 15 h. In the strain measurement by the measurmens of the EXDM, the stresses applied in the tensile direction were 64 MPa and 150 MPa for the TT-untreated specimen and 50 MPa and 150 MPa for the TT-treated specimen.
Time dependence of the stress-strain curves for thermally treated 600 alloy. Numerical values represent the time required for thermal treatment.
The stress in each sample at the X-ray irradiation position was analyzed according to the following procedure: To analyze the elastic strain, the lattice spacings of three or more independent lattice planes were calculated using the X-ray diffraction peaks obtained at each irradiation position. The strain-free lattice plane spacing was used as the average value of the lattice constants obtained from all measured data without applied stress. Tensor analysis was performed by assuming that the stress parallel to the beam direction, that is, in the direction perpendicular to the specimen surface, was released because the grain size and specimen thickness were approximately equal. Here, the elastic constant of Ni obtained from literature was employed as the elastic constant in the tensor analysis; c11 = 25.08 × 1010 Pa, c12 = 15.00 × 1010, c44 = 12.35 × 1010.11)
Figures 4(a), (b), and (c) show the principal stress conditions at each point of the TT-untreated specimen at applied loads of 0 MPa, 64 MPa, and 150 MPa, respectively. The red points indicated the measurement points, the red arrows indicated tensile stress and the blue arrows indicate compressive stress. Here, the stress at the red points without red and blue arrows could not analyzed because the information about the three independent lattice plane spaces could not be obtained. The 31 points for grain boundaries and the 19 points for interior of grains were analyzed consistently during the experiments.
Principal stresses under applied tensile stresses acting in horizontal direction on the untreated specimen. Red and blue arrows indicate tensile and compressive stress, respectively. The points represented by only red dots cannot be analyzed. (a) No applied stress, (b) 64 MPa, and (c) 150 MPa.
As shown in Fig. 4(a), stress was observed at the measurement points even under no load. As a lattice number of distorted Laue spots were observed in the acquired transmission Laue images, it was probable that the residual stress formed at the time of sample preparation was detected. Figures 4(b) and (c) show that the main stress direction changed by applying applied tensile stress in the horizontal direction. Comparing the stress conditions between grain boundaries and trans-grains, the stress around the grain boundaries showed an increasing trend. Table 2 shows the average of the maximum principal stress at each applied stress and the ratio of the tensile stress and maximum principal stress at the grain boundaries and trans-grains. The measured stress increased with the applied stress, and the number of the points indicated that the maximum principal stress of tension in the grain increased significantly. The principal stress increased considerably at the grain boundaries and trans-grain.
Figures 5(a), (b), and (c) shows the principal stress at each point of the TT-treated specimen at applied loads 0 MPa, 50 MPa, and 150 MPa, respectively. The points indicated by only red circles are the measurement points that could not be analyzed using the data obtained in the experiments. The number of points that could not be analyzed was larger for the TT-untreated specimen compared with the TT-untrearted specimen. As inclusions precipitated with the grain boundaries, the diffraction patterns from the inclusions and base material were mixed, and the three independent lattice spacings required for the strain analysis could not be obtained. The 14 points for grain boundaries and the 17 points for interior of grains were analyzed consistently during the experiments. Stress was observed at each analysis point even without applied load, as shown in Fig. 5(a). It was considered that the residual stress formed during sample preparation was detected. As shown in Fig. 5(b) and (c), the principal stress direction changed according to the stress distribution in the specimen depending on the applied tensile stress in the horizontal direction. Table 3 shows the average of the maximum principal stress at each analysis points and the maximum principal stress at the grain boundaries and trans-grains. The number of points indicated that tensile stress increased with the external force in the horizontal direction. Even though the maximum principal stress increased with the applied stress, the maximum principal stress at the grain boundaries and trans-grains was lower compared with the TT-untreated specimen.
Principal stresses under applied tensile stresses acting in horizontal direction on the thermally treated specimen. Red and blue arrows indicate tensile stress and compressive stress, respectively. The points represented by only red dots cannot be analyzed. (a) No applied stress, (b) 50 MPa and (c) 150 MPa.
The difference in the internal stress distributions of the TT-treated and TT-untreated specimens under the applied tensile stress was measured by the EXDM. According to the changes in principal stress obtained through tensor analysis, the following results were confirmed: In the TT-untreated specimen, the stress in the grain boundary increased and the maximum principal stress increased with strain. In the case of the TT-treated specimen, the principle stress changed negligibly compared with the TT-untreated specimen. The magnitude and direction of the principal stress at the each measurement point changed according to the applied stress, as shown in Fig. 4 and 5. These behaviors were related to the individual shapes of the crystal grains and the difference between each three-dimensional crystal grain arrangement and connectivity. Thus it was considered that the direction of a stress changed in a complex manner. To compare the stress between the grain boundaries and trans-grains, the principal stress, which was a complex vector with directionality, was converted into equivalent stress, which was a scalar value. The von Mises equivalent stress at each point was calculated from the eq. 1.
| \begin{equation} \sigma_{e} = \sqrt{\frac{1}{2}((\sigma_{1} - \sigma_{2})^{2} + \sigma_{1}^{2} + \sigma_{2}^{2})} \end{equation} | (1) |
Here, $\sigma _{e}$ is the equivalent stress at each point and $\sigma _{1}$ and $\sigma _{2}$ are the maximum and minimum principal stresses at each point, respectively. The average values of the equivalent stress at the grain boundaries and the trans-grains, $\sigma _{e}(GB)$, $\sigma _{e}(TG)$, were calculated. The ratio of these values, $\alpha $, was estimated as shown in eq. 2. Figure 6 shows the changes in the equivalent stress and the $\alpha $ dependent on the applied stress based on the absence of applied stress on the TT-untreated and TT-treated specimens, respectively.
| \begin{equation} \alpha = \frac{\sigma_{e}(GB)}{\sigma_{e}(TG)} \end{equation} | (2) |
Difference equivalent stress, $\text{d}\sigma _{e}$, of a difference from the equivalent stress without the applied stress and the equivalent stress ratio, $\alpha = \frac{\sigma _{e}(GB)}{\sigma _{e}(TG)}$, as a function of the applied stress at the grain boundary (GB) and trans-grain (TG). (a) Untreated specimen and (b) thermally treated specimen.
As shown in Fig. 6, the internal stress of both specimens increased with the applied stress. These tendencies varied depending on thermal treatment and the analysis positions of the specimens. The changes in the equivalent stress of the TT-untreated specimen with applied stress was larger around the grain boundary compared to the trans-grain. In contrast, the change in the equivalent stress of the TT-treated specimen with applied stress was larger in the trans-grain compared with the grain boundary. A significant difference was observed in the equivalent stress in the grain boundary between the TT-treated and TT-untreated specimens. The equivalent stress of the TT-untreated specimen was larger than that of the TT-treated specimen, and the stress of the TT-untreated specimen was concentrated in the grain boundary. This can be clearly seen in the behavior of $\alpha $. In the TT-untreated specimen, $\alpha $ increased with applied stress, whereas it remained almost constant when applied stress was applied to the TT-treated specimen. That is, even though stress was concentrated close to the grain boundary in the TT-untreated specimen, the stress concentration was relieved by the TT treatment. As a result of the relaxation of stress around the grain boundary, IGSCC sensitivity decreased in the TT-treated specimen.
Bruemmer et al.,5) Miglin et al.,12) and Kawamura et al.13) pointed out that the stress concentration close to the grain boundary might be suppressed by dividing the inclusions of the grain boundary. The difference in the stress behavior close to the grain boundary with and without the TT treatment provided a clear experimental fact against the prediction that the TT treatment would disrupt stress propagation at the grain boundary.
The change in internal stress under applied stress was examined via EXD to clarify the effect of TT treatment on mechanical properties. The following conclusions were obtained:
The effect of the TT treatment is that the stress due to bias a stress toward the grain boundaries of TT-untreated material is separated from the progression of dislocations caused by stress in the presence of grain boundary inclusions. This is consistent with the idea of stress homogenization in the structure.
This was the first time that the stress concentration closed to the grain boundary was alleviated by the TT treatment, and this was proved through the in-situ stress distribution measurement.
The EXDM measurement was carried out on the SPring-8 BL28B2 beamline as a joint use research project (2012B1207, 2013A1160, 2016A1543) of the Japan Synchrotron Radiation Research Institute (JASRI).
