Creep Rupture Strength and Phase Stability of the Austenite Phase for KA-SUS304J1HTB

Abstract

Metallurgical factor causing the heat-to-heat variation in creep rupture strength have been investigated for KA-SUS304J1HTB. In the long-term, there was a maximum difference of 3.5 times in creep rupture time between the heat with low creep strength and the heat with high creep strength. In the heat with low creep rupture strength, most of the creep voids occurred at the matrix/σ phase interface. Moreover, in the heat with low creep rupture strength, the area fraction of σ phase was larger than in the heat with high creep rupture strength. Considering that the difference in the area fraction of σ phase in each heat is related to the difference in phase stability of the austenite phase, the Md value in each heat was evaluated. The Md value is the parameter correlated with phase stability. The creep rupture time of each heat was correlated with the Md value. The smaller the Md value, the longer the creep rupture time. Therefore, the heat-to-heat variation in creep rupture strength is caused by the difference in the phase stability of each heat. In other words, in the heats with low phase stability, creep rupture strength is low because a large amount of σ phase precipitates during the creep test.

1. Introduction

KA-SUS304J1HTB steel (ASME Code Case 2328), which has excellent high-temperature strength and corrosion resistance, is widely used in heat-transfer tubes in ultra-supercritical (USC) boilers. Because there are plants where this steel has been used for over 20 years since its application in USC boilers, predicting the creep rupture strength of this steel is extremely important.

KA-SUS304J1HTB steel is a material that adds appropriate amounts of Cu, Nb, and N to SUS304H steel and has higher creep rupture strength than conventional austenitic stainless steel.^1,2,3) This high strength is due to precipitation strengthening caused by the fine Cu phase and Nb carbonitrides. However, recent studies have indicated that the creep rupture strength of this steel at high temperatures and in the long term is lower than that predicted from the low-temperature and short-term regime.^4,5) Therefore, researchers have focused on the creep degradation of this steel.^{4,5,6,7,8,9,10)} However, these studies were based on the test results of one heat, and it has been reported that the creep rupture strength of conventional austenitic stainless steels varies significantly among heats.^11,12,13) Therefore, to predict the creep rupture strength of this steel, it is necessary to investigate the heat-to-heat variation in its creep rupture strength.

Hashimoto et al.¹⁴⁾ concluded that the creep rupture strength of KA-SUS304J1HTB steel depends on the amount of soluble Nb after solution treatment, the amount of Cu phase precipitated in the grain interior, and the amount of M₂₃C₆ precipitated on the grain boundaries during creep. However, it is difficult to explain the differences in the creep rupture strength of this steel due to these microstructural differences, and a detailed investigation of the relationships between the creep rupture strength and the chemical composition and microstructure is required. Therefore, in this study, creep tests were conducted using KA-SUS304J1HTB steel, and the effects of the chemical composition and microstructure on the creep rupture strength were investigated.

2. Experimental Procedures

The test materials comprised six heats (Heats A–F) of KA-SUS304J1HTB and ASME Code Case 2328; their chemical compositions are shown in Table 1. In addition, the chemical compositions of the National Institute for Materials Science (NIMS) materials^15,16) and material specification¹⁷⁾ are shown. For Heats A–E, residual analysis was conducted to investigate the amount of soluble Nb after solution treatment. The amount of soluble Nb was estimated by subtracting the amount of Nb determined via the extracted residual analysis from the total amount of Nb, as shown in Table 2. Creep rupture tests were performed at 700–750°C and 50–170 MPa. This temperature condition is the range in which a significant strength reduction occurs according to the NIMS Creep Data Sheet.¹⁵⁾

Table 1. Chemical compositions (mass%) of the test metals.

Heat/Standard		ST temperature (°C)	C	Si	Mn	P	S	Ni	Cr	N	Cu	Nb	Fe
Heat A		1150	0.08	0.22	0.78	0.029	0	8.84	18.35	0.11	3.01	0.49	Bal.
Heat B		1150	0.09	0.20	0.79	0.030	0.001	8.72	18.35	0.11	2.95	0.42	Bal.
Heat C		Unknown	0.078	0.17	0.61	0.02	0.0007	8.75	18.35	0.096	2.89	0.45	Bal.
Heat D		1150	0.08	0.22	0.86	0.030	0	8.95	18.27	0.11	3.02	0.48	Bal.
Heat E		1150	0.08	0.26	0.80	0.033	0.002	8.78	18.78	0.10	2.96	0.50	Bal.
Heat F		1150	0.08	0.20	0.79	0.029	0	8.82	18.42	0.11	2.98	0.50	Bal.
ABQ (NIMS CDS No. M-11)		1150	0.08	0.24	0.78	0.030	0.001	8.79	18.68	0.102	2.92	0.48	Bal.
KA-SUS304J1HTB	Max.	–	0.13	0.30	1.00	0.040	0.010	10.50	19.00	0.12	3.50	0.60	Bal.
	Min.	1040	0.07	–	–	–	–	7.50	17.00	0.05	2.50	0.30

Table 2. Nb in extracted residue and soluble Nb.

	Nb in chemical composition (mass%)	Nb in extracted residue (mass%)	Sol. Nb (mass%)
Heat A	0.49	0.37	0.12
Heat B	0.42	0.35	0.07
Heat C	0.45	0.33	0.12
Heat D	0.48	0.38	0.10
Heat E	0.50	0.36	0.14

The microstructures of the parallel part of the test sample after creep rupture were examined using optical microscopy (OM) and scanning electron microscopy (SEM). The samples for observation were prepared as follows. First, the test sample was cut so that the parallel part was parallel to the stress axis. Wet polishing was then performed using emery paper (#2000), followed by mirror polishing with an oxide polishing suspension. Subsequently, electrolytic etching was performed with hydrochloric acid–ethanol, and OM observations were performed. SEM was performed under mirror-polished conditions. The electron channeling contrast imaging (ECCI) method was used for SEM observation. This is a technique for observing defects, sub-boundaries, etc. in materials using the electron channeling contrast.^18,19) ECCI was performed at an accelerating voltage of 20 kV and a working distance of 4–5 mm. In addition, the area fraction of the σ phase at the grain boundary and the particle size of Cu phase in the grain interior were quantified by image analysis. Images of the σ phase and Cu phase were captured at magnifications of 1000× and 30000×, with three fields of view each. The average values of the area fraction and particle size obtained in each field of view were taken as the area fraction of the σ phase and the particle size of Cu phase, respectively.

3. Results and Discussion

3.1. Initial Microstructures and Soluble Nb

The microstructures before the creep tests are shown in Fig. 1. For all the heats, the grain size was approximately 20 μm. Coarse insoluble NbX was also observed.⁶⁾

Fig. 1. Initial microstructures before the creep test. (a) Heat A, (b) Heat B, (c) Heat C, (d) Heat D, (e) Heat E, (f) Heat F.

In austenitic stainless steels containing Nb, it has been reported that the creep rupture strength increases with the amount of soluble Nb in the matrix.²⁰⁾ Thus, the amount of soluble Nb in each heat was determined. The results of the residual analysis for Heats A–E, along with the amounts of soluble Nb estimated using these results, are presented in Table 2. Although there were no significant differences in the results of the extracted residual analysis among the heats, when the amount of soluble Nb was estimated from the amount of Nb contained in the material, the amount of soluble Nb in Heat E was the largest (0.14 mass%), and that in Heat B was the smallest (0.07 mass%).

3.2. Differences in Creep Rupture Strength Among Heats

The creep rupture strength for each heat at 700 and 750°C is shown in Fig. 2. The creep rupture strength varied among the heats, and at both temperatures, the creep rupture strength of Heat B was the highest in the long term, whereas that of Heat E was the lowest. At 750°C, the creep rupture times of Heats B and E were almost the same in the short-term regime (<1000 h), whereas the difference in creep rupture time increased to a factor of 3.5 in the long-term regime (>1000 h). This difference in strength is inconsistent with the amounts of soluble Nb presented in Table 2, and the difference in the creep rupture strength of this steel cannot be explained by the amount of soluble Nb alone. Therefore, to investigate factors other than soluble Nb, the microstructures after creep tests were compared.

Fig. 2. Creep rupture strength of KA-SUS304J1HTB.

3.3. Microstructural Changes After Creep Tests

3.3.1. Microstructural Changes in Short-Term Regime (<1000 h)

To clarify why the creep rupture times of Heats B and E were almost the same in the short-term regime (<1000 h), microstructural observations of the creep ruptured samples were conducted. SEM observation results for the grain interior of samples ruptured at 750°C and 120 MPa are shown in Fig. 3. Figures 3(a) and 3(b) show the ECCI results for the grain interiors of Heats B and E, respectively, and Figs. 3(c) and 3(d) show corresponding black-and-white inverted images, respectively. The fine white particles observed via ECCI were Cu phase, and the linearly observed ones were dislocations⁸⁾ in Figs. 3(a) and 3(b). The area fraction and particle size of the Cu phase, which were quantified using the SEM images, are shown in Figs. 3(c) and 3(d). ECCI revealed that the fine Cu phases in both Heats B and E were pinning dislocations. It is considered that the precipitation strengthening due to the Cu phase in Heats B and E was almost the same because there were no significant differences in the area fraction and the particle size of the Cu phase.

Fig. 3. ECC images of the grain interior ruptured at 750°C and 120 MPa. (a), (c) Heat B tr=530.4 h, (b), (d) Heat E tr=429.9 h.

The SEM observation results near the grain boundaries of Heats B and E are shown in Fig. 4. Figures 4(c) and 4(d) are magnified images of Figs. 4(a) and 4(b), respectively. The black and white precipitates observed at the grain boundaries in Figs. 4(c) and 4(d) were the M₂₃C₆ and Cu phases, respectively. In contrast, the fine and coarse white precipitates in the grain interior were the Cu and NbX phases, respectively. At some grain boundaries of Heat E, the σ phase was observed. In both heats, subgrains developed near the grain boundaries, and almost no dislocations were observed inside the subgrains.

Fig. 4. ECC images of the grain boundary ruptured at 750°C and 120 MPa. (a), (c) Heat B tr=530.4 h, (b), (d) Heat E tr=429.9 h.

From the results in Figs. 3 and 4, although both Heats B and E were strengthened by the Cu phase in the grain interior, recovery structures such as subgrains were preferentially formed near the grain boundaries. This microstructural change is consistent with the microstructural changes observed during the short-term creep degradation of this steel.⁷⁾ Therefore, the creep rupture times of Heats B and E were almost the same in the short-term tests, as the subgrains that caused creep degradation were formed near the grain boundaries in both heats.

3.3.2. Microstructural Changes in Long-Term Regime (>1000 h)

Next, the microstructures of samples ruptured at 750°C and 70 MPa were examined, and a difference in the creep rupture times of Heats B and E was confirmed. The OM observation results for the creep ruptured samples are shown in Fig. 5. The precipitates indicated by the arrows in the figure represent the σ phase. In Heat E, a large amount of σ phase was observed, even though it was tested for approximately 1/3 of the time of Heat B. Additionally, it was found that many of the creep voids observed in Heat E were generated at the matrix/σ phase interface. We conducted internal pressure creep tests on KA-SUS304J1HTB and clarified that fracture originated at the matrix/σ phase interface.⁸⁾ Therefore, it is possible that there was a difference in the creep rupture times of Heats B and E due to the difference in the amount of σ phase. However, in the microstructural observation results shown in Fig. 5, σ phase precipitation rate cannot be compared, because the creep rupture times of Heats B and E differed. Therefore, the test stress is different, but the microstructures were compared using samples ruptured at approximately 3000 h.

Fig. 5. OM images of samples ruptured at 750°C and 70 MPa. (a) Heat B tr = 9553.4 h, (b) Heat E tr = 3272.5 h.

The SEM observation results for Heats B and E, which ruptured at approximately 3000 h, are shown in Fig. 6. The precipitates indicated by the arrows in the figure represent the σ phase. Despite the test time being almost the same, the amount of σ phase in Heat E exceeded that in Heat B. Additionally, as shown in Fig. 5, creep voids were generated at the matrix/σ phase interface.

Fig. 6. ECC images of samples ruptured at approximately 3000 h. (a) Heat B tr = 4443.5 h, (b) Heat E tr = 3272.5 h.

The SEM observation results for the grain boundaries and grain interiors are shown in Fig. 7. In Heat E, as shown in Fig. 7(b), the amount of M₂₃C₆ observed at the grain boundaries was smaller than that in Heat B, as shown in Fig. 7(a). It has been reported that M₂₃C₆ disappears upon precipitation and coarsening of the σ phase in KA-SUS304J1HTB.^4,5) Therefore, in Heat E, where the σ phase was observed at the grain boundaries, we consider that the amount of M₂₃C₆ was smaller than that in Heat B. Figures 7(c) and 7(d) show SEM images of the grain interiors of Heats B and E, respectively. The figures also show the area fraction and particle size of the Cu phase quantified using the SEM images. As shown in Fig. 3, the area fraction of the Cu phase was approximately 2% for both heats. In addition, there was no significant difference in the particle size of the Cu phase. Thus, while there was little difference in the microstructures in the grain interiors of Heats B and E, the amount of σ phase near the grain boundaries was larger in Heat E than in Heat B. Therefore, to clarify the relationship between the amount of σ phase and the creep rupture strength, the area fractions of the σ phase for Heats B, C, and E ruptured at 750°C were quantified. The results are shown in Fig. 8. In all the heats, the area fraction of the σ phase increased with the rupture time. The precipitation rate of the σ phase and the area fraction decreased in the order of Heat E, Heat C, and Heat B. This trend is consistent with the differences in creep rupture strength shown in Fig. 2. Therefore, it is considered that the differences in creep rupture strength among the heats were related not only to the amount of σ phase but also to its precipitation rate.

Fig. 7. ECC images of samples ruptured at approximately 3000 h. (a), (c) Heat B tr=4443.5 h, (b), (d) Heat E tr=3272.5 h.

Fig. 8. Changes in the area fraction of the σ phase during creep exposure.

3.4. Creep Rupture Strength and Phase Stability of Austenite Phase

As described in section 3.3, the precipitation rate of the σ phase and the area fraction of the σ phase varied among the heats. It is known that the precipitation rate of the σ phase depends on the amount of soluble elements in the matrix.^21,22,23,24) Additionally, the difference in the amount of σ phase is due to the phase stability of the austenite phase. Therefore, the stability of the austenite phase for each heat was estimated using the compositional average Md (Md) proposed by Morinaga et al.^{25,26,27,28,29,30,31)} Morinaga et al. applied Md to austenitic steel and high Cr ferritic steel and predicted whether the σ and δ ferrite phases would form in an equilibrium state. According to their reports, in austenitic steel and high Cr ferritic steel with large Md values, the σ and δ ferrite phases precipitated, and the amount of precipitation tended to increase.^28,30) A larger Md value corresponds to lower phase stability of the alloy. The Md value can be calculated as follows:

  

$$\overline{Md}=\sum X_i\cdot \left(Md\right)i$$

where X_i represents the atomic fraction (at%) of component i in the alloy, and (Md)_i represents the d-orbital energy level (eV) of component i. The Md value of each element was calculated using the values for the austenite phase.³¹⁾ A point to note when calculating the Md value is that for X_i, only the amount of soluble alloy elements in the matrix should be used, and the amount of alloy elements contained in the precipitates must be considered.²⁵⁾ Therefore, it is necessary to determine X_i assuming that precipitation has already been completed for the carbide, nitride, and Cu phases. Long-term test results indicated that M₂₃C₆, Cu phase, NbX, and NbCrN exist as precipitates in KA-SUS304J1HTB.⁵⁾ In addition, Okada et al.³²⁾ performed an extracted residual analysis using samples of KA-SUS304J1HTB after long-term aging tests and quantitatively determined the amount of alloy elements that existed as precipitates. The results are presented in Table 3. For example, 1.5 mass% of the Cr mass existed as a precipitate. Therefore, when calculating the Md value, the amount of Cr shown in Table 1 was reduced by 1.5 and then converted to at%. For Cu and C, the solubility limits reported in the literature were used, because the amount that exists as precipitates is unknown.^33,34) The relationship between the calculated Md value and creep rupture time obtained using the above method is shown in Fig. 9. Figures 9(a) and 9(b) show the creep data at 750 and 700°C, respectively. The figures also present data from NIMS and Hashimoto et al. In short-term tests where the σ phase hardly precipitated, the creep rupture time was independent of the Md value. In contrast, in long-term tests where the σ phase precipitated, the creep rupture time increased as the Md value decreased. This implies that for each heat, the creep rupture time in the long-term regime was correlated with the phase stability of the austenite phase. Thus, it is considered that the creep rupture strength depends on the amount of σ phase.

Table 3. Solid solubility of each element in the matrix.

	Element in extracted residue (mass%)	Solid solubility of each element in matrix. (mass%)	Ref.
Cr	1.5	Cr-1.5	32)
N	0.06	N-0.06	32)
Nb	0.4	Nb-0.4	32)
Fe	0.2	Fe-0.2	32)
Cu	–	3.186–3093/T	33)
C	–	5.4×103exp(−13400/T)+0.007	34)

Fig. 9. Relationship between time to rupture and Md. (a) 750°C, (b) 700°C.

3.5. Effect of Alloy Elements on Precipitation Rate of σ Phase

In the previous section, we discussed the relationship between the creep rupture strength and the amount of σ phase. However, the microstructural observations suggest that the creep rupture strength of each heat depended on not only the amount of σ phase but also the precipitation rate of the σ phase. Therefore, the effect of alloy elements on the precipitation rate of the σ phase was examined. In austenitic steel, the large Md value region becomes a two-phase stable region of the γ phase and σ phase.²⁹⁾ Furthermore, a larger Md value corresponds to a larger amount of σ phase.²⁸⁾ Therefore, in austenitic steel with a large Md value, the supersaturation of each alloy element that dissolves in the austenite phase before the σ phase precipitates is high. In general, as the amount of σ phase precipitation-promoting elements dissolved in the matrix increases, the precipitation rate of the σ phase increases.^21,22,23,24) The reason for this is illustrated schematically in Fig. 10.

Fig. 10. Change in the driving force for precipitation with increasing a soluble element in matrix.

Here, element i is considered as a σ phase precipitation-promoting element. When the amount of soluble element i in the matrix is X_i, the driving force for σ phase precipitation can be expressed as ΔG^σ at X_i = $X_i^{\mathrm{I}}$.³⁵⁾ When the amount of soluble element i in the matrix increases to $X_i^{\mathrm{I}\mathrm{I}}$ (>$X_i^{\mathrm{I}}$), the driving force for σ phase precipitation becomes ΔG^σ at X_i = $X_i^{\mathrm{I}\mathrm{I}}$; thus, the driving force for σ phase precipitation increases, and the precipitation rate of the σ phase is expected to increase. Hence, the precipitation rate of the σ phase depends on the type and amount of soluble elements in the matrix, and it is inferred that the precipitation rate of the σ phase differed owing to the difference in heat in Fig. 8. This difference in the precipitation rate also promotes microstructural changes, such as the formation of creep voids, which are considered to have affected the creep rupture strength.

Thus, the factor that caused the heat-to-heat variation in the creep rupture strength of KA-SUS304J1HTB in the long-term regime was the difference in phase stability among the heats. Accordingly, it is considered that in the heat where the phase stability of the austenite phase was low, the σ phase precipitation during the creep test was fast, and the amount of σ phase is large; thus, creep voids occurred at the matrix/σ phase interface, leading to early rupture.

4. Conclusions

To clarify the factors that cause heat-to-heat variation in the creep rupture strength of KA-SUS304J1HTB, the relationships among the creep rupture strength, chemical composition, and microstructure were investigated. The results are summarized below.

(1) The creep rupture strength varied among the heats, and at both 700 and 750°C, the creep rupture strength of Heat B was the highest in the long term, whereas that of Heat E was the lowest. At 750°C, the creep rupture times of Heats B and E were almost the same in the short-term regime (<1000 h), whereas the difference in creep rupture time increased to a factor of 3.5 in the long-term regime (>1000 h).

(2) The microstructures of Heats B and E were compared in the short-term regime at 750°C, where the creep rupture times were almost the same. Although both Heats B and E were strengthened by the Cu phase in the grain interior, subgrains that caused creep degradation were preferentially formed near the grain boundaries.

(3) The microstructures of Heats B and E after approximately 3000 h of testing were compared in the long-term regime at 750°C, and a difference in the creep rupture time was confirmed. Thus, the precipitation strengthening due to the Cu phase in the grain interiors was almost the same for these heats. However, the amount of σ phase in Heat E exceeded that in Heat B near the grain boundaries, and creep voids mainly occurred at the matrix/σ phase interface.

(4) Quantifying the area fraction of the σ phase for samples ruptured at 750°C in Heats B, C, and E revealed that the precipitation rate of the σ phase and the area fraction decreased in the order of Heat E, Heat C, and Heat B. This trend was consistent with the differences in creep rupture strength.

(5) Using the Md value, the phase stability of the austenite phase for each heat was evaluated. The Md values differed among the heats, corresponding to the differences in the amount of σ phase. Additionally, in the long-term regime where the σ phase precipitated, there was a correlation between the creep rupture time and the phase stability of the austenite phase for each heat.

Thus, the main factor that causes the heat-to-heat variation in the creep rupture strength of KA-SUS304J1HTB is the difference in the phase stability of each alloy. In heats with low stability of the austenite phase, the σ phase precipitation during the creep test was fast, and the amount of σ phase was large; thus, creep voids occurred at the matrix/σ phase interface, reducing the creep rupture strength.

References

• 1)   H.  Semba,  Y.  Sawaragi,  K.  Ogawa,  A.  Natori and  T.  Kan: Materia Jpn., 41 (2002), 120 (in Japanese). https://doi.org/10.2320/materia.41.120
• 2)   H.  Semba,  H.  Okada,  K.  Ogawa,  A.  Iseda and  Y.  Yamadera: Materia Jpn., 46 (2007), 99 (in Japanese). https://doi.org/10.2320/materia.46.99
• 3)   F.  Masuyama: ISIJ Int., 41 (2001), 612. https://doi.org/10.2355/isijinternational.41.612
• 4)   K.  Kimura,  M.  Murata,  K.  Kamihira and  S.  Kobayashi: Proc. 10th Asia-Pacific Conference on Fracture and Strength. (APCFS 2016), (Toyama), The Japan Society of Mechanical Engineers, Toyama, (2016), 265.
• 5)   K.  Sawada,  K.  Sekido,  M.  Murata,  K.  Kamihira and  K.  Kimura: Mater. Charact., 141 (2018), 279. https://doi.org/10.1016/j.matchar.2018.04.055
• 6)   M.  Mitsuhara,  K.  Ikeda,  H.  Nakashima,  H.  Hayakawa,  J.  Kusumoto and  A.  Kanaya: Tetsu-to-Hagané, 94 (2008), 299 (in Japanese). https://doi.org/10.2355/tetsutohagane.94.299
• 7)   K.  Kubushiro,  Y.  Shioda and  K.  Nomura: J. Soc. Mater. Sci. Jpn., 66 (2017), 179 (in Japanese). https://doi.org/10.2472/jsms.66.179
• 8)   T.  Kimura,  Y.  Shioda and  K.  Kubushiro: J. Soc. Mater. Sci. Jpn., 68 (2019), 142 (in Japanese). https://doi.org/10.2472/jsms.68.142
• 9)   T.  Hatakeyama,  K.  Sawada,  T.  Hara,  K.  Sekido and  K.  Kimura: Scr. Mater., 200 (2021), 113904. https://doi.org/10.1016/j.scriptamat.2021.113904
• 10)   K.  Sawada,  T.  Hatakeyama,  K.  Sekido and  K.  Kimura: Mater. Charact., 178 (2021), 111286. https://doi.org/10.1016/j.matchar.2021.111286
• 11)   F.  Abe: Proc. ASME 2014 Pressure Vessels & Piping Conference. (PVP 2014), (Anaheim), PVP2014-28158. 10.1115/PVP2014-28158
• 12)   F.  Abe,  H.  Tanaka and  M.  Murata: Mater. at High Temp., 33 (2016), 626. https://doi.org/10.1080/09603409.2016.1206740
• 13)   F.  Abe: Metall. Metall. Mater. Trans. A, 47 (2016), 4437. https://doi.org/10.1007/s11661-016-3587-3
• 14)   K.  Hashimoto,  K.  Arisue,  N.  Saito,  N.  Komai,  E.  Nagashima,  K.  Tominaga and  M.  Fujita: Proc. Joint EPRI-123HiMAT International Conference on Advances in High Temperature Materials, (Nagasaki), (2019), 655.
• 15)  NIMS Creep Data Sheet, Metallographic Atlas of Long-term Crept Materials, No. M-11, National Institute for Materials Science, Japan, (2016).
• 16)  NIMS Creep Data Sheet, No. 56A, National Institute for Materials Science, Japan, (2018).
• 17)  The Material Specification in the Technical Regulation of Equipment for Thermal Power Plant, (2016), 137.
• 18)   Y.  Shioda,  N.  Saito,  T.  Tanoue,  K.  Nomura and  K.  Kubushiro: J. Soc. Mater. Sci. Jpn., 70 (2021), 177 (in Japanese). https://doi.org/10.2472/jsms.70.177
• 19)   Y.  Shioda,  H.  Nakagawa,  K.  Nomura,  N.  Yamazaki and  K.  Kubushiro: Proc. Joint EPRI-123HiMAT International Conference on Advances in High Temperature Materials, (Nagasaki), (2019), 592.
• 20)   T.  Yukitoshi,  K.  Yoshikawa and  H.  Teranishi: Tetsu-to-Hagané, 70 (1984), 1962 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.70.15_1962
• 21)   T.  Sourami: Mater. Sci. Technol., 17 (2001), 1. https://doi.org/10.1179/026708301101508972
• 22)   K.  Ogawa and  T.  Osuki: Q. J. Jpn. Weld. Soc., 33 (2015), 55 (in Japanese). https://doi.org/10.2207/qjjws.33.55
• 23)   K.  Ogawa and  T.  Osuki: Q. J. Jpn. Weld. Soc., 33 (2015), 62 (in Japanese). https://doi.org/10.2207/qjjws.33.62
• 24)   Y.  Okazaki,  K.  Miyahara,  Y.  Hosoi,  M.  Tanino and  H.  Komatsu: J. Jpn. Inst. Met., 53 (1989), 512 (in Japanese). https://doi.org/10.2320/jinstmet1952.53.5_512
• 25)   M.  Morinaga,  N.  Yukawa and  H.  Adachi: Tetsu-to-Hagané, 71 (1985), 1441 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.71.11_1441
• 26)   M.  Morinaga,  N.  Yukawa,  H.  Ezaki and  H.  Adachi: Philos. Mag. A, 51 (1985), 223. https://doi.org/10.1080/01418610.1985.12069159
• 27)   M.  Morinaga,  N.  Yukawa,  H.  Ezaki and  H.  Adachi: Philos. Mag. A, 51 (1985), 247. https://doi.org/10.1080/01418610.1985.12069160
• 28)   H.  Ezaki,  M.  Morinaga and  N.  Yukawa: Philos. Mag. A, 53 (1986), 709. https://doi.org/10.1080/01418618608242867
• 29)   Y.  Murata,  M.  Morinaga,  N.  Yukawa,  S.  Ukai,  S.  Nomura,  T.  Okuda and  M.  Harada: J. At. Energy Soc. Jpn., 41 (1999), 1244. https://doi.org/10.3327/jaesj.41.1244
• 30)   H.  Ezaki,  M.  Morinaga,  K.  Kusunoki and  Y.  Tsuchida: Tetsu-to-Hagané, 78 (1992), 1377 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.78.8_1377
• 31)   M.  Morinaga: A Quantum Approach to Alloy Design, Elsevier, (2018), 56.
• 32)   H.  Okada: Ph.D. thesis, Tohoku University, (2012), 147.
• 33)   S.  Liu,  X.  Rong,  H.  Guo,  R.D.K.  Misra,  X.  Jin and  C.  Shang: Mater. Sci. Eng. A, 825 (2021), 141783. https://doi.org/10.1016/j.msea.2021.141783
• 34)   H.  Arai: Tetsu-to-Hagané, 56 (1970), 44 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.56.1_44
• 35)   K.  Sakuma: J. Jpn. Inst. Met., 20 (1981), 247 (in Japanese). https://doi.org/10.2320/materia1962.20.247