Effect of Rhenium Addition on Tungsten Diffusivity in Iron-Chromium

2 downloads 0 Views 125KB Size Report
Aug 15, 2006 - The Alloying effect of Re on the diffusivity of W in Fe-15 mol%Cr based ... Diffusion, heat resistant steel, refractory element, rhenium, tungsten.
Materials Transactions, Vol. 47, No. 8 (2006) pp. 2106 to 2108 #2006 The Japan Institute of Metals

RAPID PUBLICATION

Effect of Rhenium Addition on Tungsten Diffusivity in Iron-Chromium Alloys Tomonori Kunieda1; * , Koji Yamashita1; *, Yoshinori Murata1 , Toshiyuki Koyama2 and Masahiko Morinaga1 1

Department of Materials, Physics, and Energy Engineering, Graduate School of Engineering, Nagoya University, Nagoya 464-8603, Japan 2 National Institute for Materials Science, Tsukuba 305-0047, Japan The Alloying effect of Re on the diffusivity of W in Fe-15 mol%Cr based alloys was investigated experimentally using the Fe-15Cr/Fe15Cr-5W and Fe-15Cr-1Re/Fe-15Cr-5W diffusion systems. In these systems, a single ferrite phase existed stably at 1473 K, and Fe and W atoms interdiffused at 1473 K without any attendant changes in the Cr concentration of 15 mol% in them. The measured diffusion length of W atoms was shorter in the Re-containing diffusion system than the Re-free one. Following the binary Boltzmann-Matano method, an apparent interdiffusion coefficient at 1473 K was estimated to be 7:1  1015 m2 /s in the Re-free diffusion system and 1:5  1015 m2 /s in the Recontaining diffusion system. Thus, the presence of Re in the alloy worked to suppress the atomic diffusion of W in a ferrite phase. [doi:10.2320/matertrans.47.2106] (Received March 28, 2006; Accepted July 3, 2006; Published August 15, 2006) Keywords: Diffusion, heat resistant steel, refractory element, rhenium, tungsten

1.

Introduction

The refractory elements such as Mo, W and Re play an important role in strengthening high temperature creep of high Cr heat resistant ferritic steels.1–3) In these steels, it is known that the (Fe, Cr)2 (W, Mo) type Laves phase precipitates together with M23 C6 carbide and the MX carbonitride.4,5) Recently, it has been shown that the longterm creep strength at 923 K becomes much smaller than the extrapolated strength from the short-term creep test.6) This is related mainly to the microstructure changes in the steel during creep. Recently, it has been reported that the addition of a small amount of Re is effective in improving the longterm creep strength of the W-containing steels.7–9) This implies that the presence of Re in the steel lowers the W diffusivity in some ways, because the microstructure evolution during creep is subjected more or less to the coalescence of the Laves phase, (Fe, Cr)2 (W, Mo), and M23 C6 carbide. The diffusivity of refractory elements in heat resistant steels is crucial for the basic understanding of the microstructural stability during creep.10) The purpose of this study is to investigate the alloying effect of Re on the W diffusivity in the ferritic steel. For this aim, Fe-15 mol%Cr was chosen as a base alloy for the bcc matrix phase in high Cr ferritic steels. This alloy is suitable for the present diffusion experiment, as a bcc phase is stable up to the melting temperature in it. 2.

were cut from the ingots. The homogenization and the grain growth of each specimen were accomplished by 50% cold rolling, followed by the annealing at 1473 K for 24 h. The grain size of each specimen was estimated by an optical microscope (OM) and a scanning electron microscope (SEM) equipped with an instrument for taking an electron backscattered diffraction pattern (EBSD). The grain size of each alloy was as large as about 1 mm after the annealing. The diffusion temperature, 1473 K, employed in this study was about 0.8Tm (Tm : melting temperatures of the experimental alloys). In general, it is believed that when the temperature is higher than 0.75Tm , there is little effect of interfacial diffusion on lattice diffusion in the alloy.11) A diffusion couple shown in Fig. 1 was prepared in the following way. First each side of the plate specimen was ground, and then one surface of each specimen was polished mechanically with emery papers and 0.25 mm diamond

Fe-15Cr-5W

W

Alumina fiber

Experimental Procedures

Fe-15Cr

Fe-15Cr binary alloy, Fe-15Cr-1Re and Fe-15Cr-5W ternary alloys were used in this study. Here, the alloy compositions are given in mol% units. According to the FeCr and Fe-Cr-W phase diagrams, these three alloys should consist of a bcc phase at 1473 K. The button ingots of these alloys were prepared by arc melting in a high purity argon gas atmosphere. Plate specimens with about 1 cm in thickness *Graduate

Student, Nagoya University

Fe-15Cr-1Re

W Fig. 1

Schematic illustration showing a diffusion couple.

Effect of Rhenium Addition on Tungsten Diffusivity in Iron-Chromium Alloys

2107

0.06

(a)

1.01

Mole fraction (W)

Alumina fiber

0.05

1.00

0.04

0.99

0.03

Fe+Cr W

0.02

0.98 0.97

0.01

0.96

0.00

0.95

-0.01 -1500 -1000

(b)

-500

0

Distance, l /

500

1000

Mole fraction (Fe+Cr)

(a)

0.94 1500

m 1.01

0.06

20

m

Fig. 2 Typical SEM microstructures showing the initial interface (alumina fiber) in (a) Fe-15Cr/Fe-15Cr-5W and (b) Fe-15Cr-Re/Fe-15Cr-5W diffusion system, both annealed at 1473 K for 100 h.

Mole fraction (W)

Alumina fiber

0.05

1.00

0.04

0.99

0.03

3.

Results and Discussion

3.1 Concentration profiles and diffusion path It was confirmed from the optical microscopic observation that the as-annealed specimens contained only the ferrite (bcc) single phase without any precipitates in them. Figure 2 shows typical SEM images taken from the diffusion couple annealed at 1473 K for 100 h. In this figure, the alumina fiber is seen as a dark circle image, and its position is on the initial interface between the diffusion couple. The concentration profiles of W and (Fe þ Cr) are shown in Fig. 3(a) for Fe15Cr/Fe-15Cr-5W and in Fig. 3(b) for Fe-15Cr-1Re/

W

0.02

0.98 0.97

0.01

0.96

0.00

0.95

-0.01 -1500

slurry. Subsequently, the diffusion couple of Fe-15Cr/Fe15Cr-5W and Fe-15Cr-1Re/Fe-15Cr-5W was assembled as shown Fig. 1. There were the two diffusion systems in this assembly. Alumina fibers were sandwiched between the plate specimens as a marker of the initial interface. This couple was held tightly with a molybdenum holder. Then, it was encapsulated in a quartz tube with argon gas and annealed at 1473 K for 100 h. After the annealing, the cross section of the diffusion couple was examined by the scanning electron microscopy (SEM) and the energy dispersive X-ray spectroscopy (EDX) to measure concentration profiles across the diffusion interface. The origin of the concentration profile was set at the position of the alumina fiber.

Fe+Cr+Re

-1000

-500 0 500 Distance, l / m

1000

Mole fraction (Fe+Cr+Re)

(b)

0.94 1500

Fig. 3 Concentration profiles of W and (Fe þ Cr) obtained from (a) Fe15Cr/Fe-15Cr-5W diffusion system and (b) Fe-15Cr-1Re/Fe-15Cr-5W diffusion system, both annealed at 1473 K for 100 h.

Fe-15Cr-5W. These concentration profiles exhibited the S curves without any discontinuity, indicating that the diffusion occurred in the single ferrite phase without precipitating any phases in the alloy. From Fig. 3, it was found that W atoms diffused into the region of about 800 mm apart from the interface in the Re-containing system. Such a diffusion region of W atoms was further extended over above 1000 mm in the Re-free system. This result indicates that W atoms diffuse more easily in the Re-free alloy than in the Recontaining alloy. The diffusion path of each diffusion system was obtained by plotting the data shown in Fig. 3 in the Gibbs triangle. The result is shown in Fig. 4. The concentrations of Cr did not change during annealing in both the diffusion systems. A very small difference in the diffusion paths between the two systems was attributable to the existence of 1 mol% Re in the Fe-15Cr-1Re/Fe-15Cr-5W, but the absence in the Fe-15Cr/ Fe-15Cr-5W. This result implies that the presence of Cr scarcely affects the W diffusion in the Fe-15Cr based diffusion couples. In other words, W atoms diffuse in the

2108

T. Kunieda, K. Yamashita, Y. Murata, T. Koyama and M. Morinaga

Re-free Re-containing

alloys. The reason why the presence of Re retards the W diffusion in Fe-15Cr alloys is not clear at the moment, but there are two possibilities. One is the existence of attractive force between W and Re due to the trend of the formation of an intermetallic compound between them. The other is the vacancy trapping effect of Re in the alloy.

20

Cr ol) (m

4.

The alloying effect of Re on the W diffusivity was examined experimentally using both the Fe-15Cr/Fe-15Cr5W and the Fe-15Cr-1Re/Fe-15Cr-5W diffusion systems. The apparent interdiffusion coefficient for the Re-containing alloy was found to be about one fifth of that for the Re-free alloy. It was concluded that the existence of Re retarded significantly the W diffusion in Fe-15 mol%Cr based alloy.

10

0

10

20

Acknowledgments

W (mol) Fig. 4 Diffusion paths in Fe-15Cr/Fe-15Cr-5W and Fe-15Cr-1Re/Fe15Cr-5W diffusion systems.

Table 1 Apparent interdiffusion coefficients obtained from Fe-15Cr/Fe15Cr-5W and Fe-15Cr-1Re/Fe-15Cr-5W diffusion systems. diffusion system Fe-15Cr/Fe-15Cr-5W Fe-15Cr-1Re/Fe-15Cr-5W

Conclusion

diffusion coefficient (m2 s1 ) 7:1  1015 15

1:5  10

Fe-Cr (-Re) alloy without distinguishing between Fe and Cr, owing mainly to the equi-composition, 15 mol%, of Cr between the two diffusion systems. 3.2 Interdiffusion coefficients In order to compare the W diffusivity quantitatively in each diffusion system, interdiffusion coefficients were estimated by assuming that the diffusion occurs in the (Fe, Cr)-W pseudo-binary system. The apparent interdiffusion coefficient was calculated using the binary Boltzmann-Matano method.12) The calculated interdiffusion coefficients at 1473 K in the two diffusion systems are shown in Table 1. The interdiffusion coefficient was about 5 times larger in the Re-free alloy than in the Re-containing alloy. This result indicates that the existence of Re suppresses the diffusion of W atoms in Fe-Cr

The authors would like to thank Associate Prof. Numakura in Kyoto University for his valuable discussion. Also, this work was supported in part by the Grant-in-Aid for Scientific Research of Japan Society for Promotion of Science (JSPS), Japan. REFERENCES 1) T. Fujita: Tetsu-to-hagane 76 (1990) 1053–1059. 2) H. Naoi, M. Oogami, Y. Hasegawa, H. Mimura and T. Fujita: Advanced Heat Resistant Steels for Power Plant Generation, eds. R. Viswanathan and J. Nutting, (Institute of Materials, London, 1999) 257–269. 3) Y. Murata, K. Kawamura, M. Kamiya, M. Morinaga, R. Hashizume, K. Miki, T. Azuma and T. Ishiguro: ISIJ Int. 42 (2002) 1591–1593. 4) L. M. Lundin: Scr. Mater. 34 (1996) 741–747. 5) Y. Murata, M. Morinaga, R. Hashizume, K. Takami, T. Azuma, Y. Tanaka and T. Ishiguro: Mater Sci Eng A 282 (2000) 251–261. 6) Y. Hasegawa, T. Muraki and M. Oogami: J. Soc. Mater. Sci., Jpn. 52 (2003) 843–850. 7) M. Morinaga, R. Hashizume and Y. Murata: Materials for Advanced Power Engineering, 1994, ed. by D. Coutsouradis et al., (Kluwer Academic Publishers, Dordrecht, 1994) 319–328. 8) Y. Tsuda, M. Yamada, R. Ishii and O. Watanabe: Advances in Turbine Materials, Design and Manufacturing, ed. by A. Strang et al., (The Institute of Materials, London, 1997) 283–295. 9) F. Masuyama and N. Komai: Materials for Advanced Power Engineering, 1998, ed. by J. Lecomte-Beckers et al., (Forchungszentrum Julich, 1998) 269–276. 10) K. Maruyama and H. Nakazima: Kouonkyoudo-no-zairyokagaku, Japan (1997) pp. 172–173. 11) D. A. Porter and K. E. Easterling: Phase Transformations in Metals and Alloys, (Van Nostrand Reinhold, England, 1981) p. 102. 12) C. Matano: Japanese J. Phy. 8 (1933) 109–113.