Precipitation behavior of Laves phase in 10%Cr steel X12CrMoWVNbN10-1-1 during short-term creep exposure
Huiran Cui,Feng Sun,Ke Chen,Lanting Zhang,Rongchun Wan,Aidang Shan,Jiansheng Wu
DOI: https://doi.org/10.1016/j.msea.2010.08.013
2010-01-01
Abstract:Research highlights ▶ Laves phase precipitates in steel X12 during short-term exposure of 204 h at 650 °C and 400 h at 600 °C. ▶ Laves phase has stacking faults in it and the SAED patterns of it are characteristic of streaked. ▶ There are two mechanisms of the nucleation and growth for Laves phase present in steel X12. ▶ The size and the number of Laves phase are strongly dependent on temperature, but slightly on stress. ▶ The formation of Laves phase in steel X12 is a thermodynamically possible process. Abstract Precipitation behavior of Laves phase in steel X12CrMoWVNbN10-1-1 during short-term (≤400 h) creep exposure was investigated using a transmission electron microscope (TEM) and an ultra high resolution thermal field emission scanning electron microscope (UHR-FESEM). The results demonstrate that Laves phase precipitates are detected in the creep specimens exposed for such a short-term of 400 h at 600 °C and 204 h at 650 °C. Stacking faults are observed in them and the selected area electron diffraction patterns of the phase are characteristic of streaked. There are two mechanisms of the nucleation and growth for Laves phase, one of which is that Laves phases occur alone on martensite lath boundaries and they are coherent with a grain but grow into the adjacent grain which has no rational orientation relationship with them. The other one is that Laves phases are formed in the regions adjacent to M 23 C 6 particles, and then they grow at the expense of the Cr-rich M 23 C 6 carbides in close vicinity. It is established that the former one is the dominant formation mechanism for Laves phase in steel X12CrMoWVNbN10-1-1. Laves phases prefer to locate on the prior austenite grain boundaries and the martensite lath boundaries. The mean diameter and the number density of the phase show significant change with temperature, but slightly change with stress. Moreover, it is proved that the formation of Laves phase in this steel is a thermodynamically possible process. The nucleation process has not finished after heat-treatment and the number of nuclei for Laves phase formed in the steel decreases with raising the temperature. Keywords Laves phase Precipitation behavior X12CrMoWVNbN10-1-1 10% Cr steels Short-term creep Mean diameter 1 Introduction X12CrMoWVNbN10-1-1 (COST steel E [1] , hereby, it will be referred as X12) is recognized as a mature material that is widely used for rotor steel of ultra-supercritical steam turbines. It was modified from 9Cr–1Mo–V–Nb steel (ASME SA213-T91 [2] ) by adding W to increase creep strength and increasing Cr content for better oxidation resistance. The steel is characterized by tempered martensite matrix and strengthened by solution hardening of such elements as Mo, W and by precipitation hardening of such phases as M 23 C 6 (M: Cr, Fe, Mo, W, etc.) type carbides and MX (M: V, Nb, etc. and X: C, N) type carbonitrides [3–9] . The degradation of creep properties of the steel during long-term exposure is generally accompanied by ripening of M 23 C 6 carbides and precipitation of new phases such as Laves phase and Z phase. Precipitation of Laves phase is unavoidable for W and Mo alloyed X12 steel during ageing or creep exposure at temperatures around 600 °C because it is thermodynamically stable. The effect of the Laves phase precipitation on long-term creep strength has been debated for long. Laves phase is generally considered detrimental for the creep strength of 9–12% Cr steels because the precipitation of Laves phase promotes the depletion of Mo and W from solid solution [10–12] . On the other hand, they can contribute to the increase of the creep strength by precipitation hardening under certain circumstances [13] . Recently, it is generally agreed that the fine Laves phase may contribute to the creep strength [13–16] . In this case, the coarsening of Laves phase should be controlled during high-temperature exposure to maintain a good creep resistance. So far, the precipitation behavior of Laves phase in these steels including steel X12 has been extensively studied for long-term (>1000 h) creep exposure [3–20] , but very limitedly for short-term (200–400 h) creep, showing the behavior at the early stage of creep. However, it is difficult to control the coarsening of Laves phase without complete understanding on its behavior at the early stage. The aim of the present work is to investigate the precipitation behavior of Laves phase in X12 during short-time (≤400 h) creep, focusing on its microstructure evolution, the mechanism of its nucleation and growth and quantification of the particles. The role of stress and temperature on the precipitation of Laves phase is discussed. 2 Materials and experimental methods In present study, the testing material was cut from a steam turbine rotor made of a forged steel X12 produced by SAAR in Germany. The material was austenitized at 1050 °C for 21.5 h and then cooled in oil, followed by tempering at 570 °C for 21 h in the first step and at 690 °C for 23 h in the second step. The nominal composition is listed in Table 1 . Particle-size measurements on Laves phase were carried out on samples crept at 600 °C/650 °C for 400 h under different stresses ( Table 2 ). All measurements were made on the gauge area. Laves phase in the steel can be successfully identified against the matrix as bright precipitates via scanning electron microscope under backscattered electron (SEM-BSE) mode due to the large Z contrast of 28% between Laves phase and the matrix in the steel X12 [15] . The technique of quantification of Laves phase in various samples is the same as the method reported in Ref 16. Samples were prepared by grinding, polishing, and slight etching in Villella's etching solution. SEM was performed at 15 kV on an ultra high resolution thermal field emission scanning electron microscope (UHR-FESEM) JSM-7600F at a displayed magnification of 25,000×. 20 micrographs per specimen were recorded. Both secondary electron (SE) and backscattered electron (BSE) images consisting of 1280 × 962 pixels were taken from each area. The Image-Pro Plus software [17] was used for image analysis. To ensure a minimal systematic error [15,16] , the quantification of the particle-size smaller than approximately 10 pixels was deleted from the results. Hence, Laves phase particles with a diameter smaller than 40 nm are not included in this study. The area fraction and the number of Laves phase particles were evaluated on the basis of information from the mentioned 20 areas for each specimen. The detailed microstructure of Laves phase was observed by using a conventional transmission electron microscope (JSM-2010) operating at 200 kV. 3 Results and discussion In the as-received condition, there are only M 23 C 6 and MX present in steel X12, and no Laves phase has been detected. However, in addition to M 23 C 6 and MX, Laves phase has been found in the creep specimens after exposure at 650 °C for 204.7 h and at 600 °C for 400 h, as shown in Figs. 1 and 2 . From the bright-field images ( Figs. 1(a) and 2(a) ) and the dark-field images ( Figs. 1(b) and 2(b) ), it can be observed that Laves phase precipitates have irregular shape and locate along the martensite lath boundaries or subgrain boundaries. The stacking faults within the Laves phases could be observed in both tested specimens. The particles of Laves phase formed during creep at 650 °C are larger than those precipitated at 600 °C. In Fig. 1 (a), B, C and D precipitates are also identified as Laves phase. They are clustering along the lath boundary, which is in agreement with the previous report that the distribution of the Laves phase is very heterogeneous and often in a clustering state [18] . Analysis of the selected area electron diffraction (SAED) pattern ( Fig. 1 (c)) shows that the crystal structure of Laves phase is a hexagonal structure with the lattice constants of a = 4.75 Å and c = 7.73 Å, in agreement with those of Laves phase in a 12Cr–2W heat-resistant steels [18] . Characteristic streaks on the SAED pattern are displayed in Fig. 1 (c) for particle A, similar to the result shown previously in 9–12% Cr steels [18,19] . This feature allows easy distinction of Laves phase from M 23 C 6 . Both particles prefer to precipitate in close vicinity, preferentially on subgrain boundaries [19] , as shown in Fig. 2 (a). In addition, Fig. 1 (d) compares the energy-dispersive spectrum (EDS) of particle A and the matrix. It is obvious that Laves phase in the experimental steel is of (Fe, Cr) 2 (W, Mo) type and the particle is enriched in W and Mo. With regard to the nucleation and growth of Laves phase in steel X12, in current study, two mechanisms have been found, one mechanism is that the nucleation of Laves phase in the matrix occurs on martensite lath boundaries or subgrain boundaries alone (with no M 23 C 6 carbides in the vicinity), Laves phases are coherent with a grain but grow into the adjacent grain with which they do not have a rational orientation relationship, because of coherent interfaces having a lower mobility than incoherent interfaces, as shown in Fig. 1 (a), and similar results have been found in a 12Cr–2W power plant steel [18] and steel P92 [19] . The other one is that Laves phase precipitates in the chromium-depleted regions adjacent to M 23 C 6 particles, as displayed in Fig. 2 (a). It is because such regions have relatively high W and Mo content, which are beneficial to the formation of Laves phases. Then Laves phases grow at the expense of the Cr-rich M 23 C 6 carbides in the vicinity due to the rearrangement of the alloy elements (Cr, Mo, W) [20,21] . Similar observations have been made for steel T9 [22] . It is interesting to note that most of Laves phases precipitating in steel X12 during short-term creep are found alone on martensite lath boundaries or subgrain boundaries with no M 23 C 6 carbides in close vicinity in our experiment. It can conclude that the former one is the dominant formation mechanism for Laves phase in X12. Fig. 3 is the UHR-FESEM image of the tested steel, presenting the distribution sites of Laves phase precipitates (indicated by black arrows) after short-term creep under various conditions. It is apparent that the prior austenite grain boundaries (PAGBs) and martensite lath boundaries are the favorable sites for Laves phase, perhaps because the interfacial energy of these boundaries facilitates its nucleation. Sawada et al. [23] reported some Laves phase particles located inside the grain (thought to be the location of the former grain boundaries). However, in our study, no Laves phase has been observed inside the grain. Quantification of Laves phase particle in the creep specimen is shown in Fig. 4 , including the corrected mean diameter, area fraction and number of Laves phase precipitates as a function of stress and temperature. The error bars in Fig. 4 (a) are derived from the standard deviation and the number of particles with assuming the 95% confidence internal as normal distributions. It is obvious that the corrected mean diameter of Laves phase precipitated at 650 °C is larger than that at 600 °C. However, the area fraction and the number of particles of specimen at 650 °C are lower than those at 600 °C ( Fig. 4 (b)), that is, creep specimens exposed at 650 °C display a lower density of Laves phase precipitates as compared with those at 600 °C, which is consistent with the results of steel GX12 and steel P92 reported in Ref. [15] . Furthermore, it must be noted that Laves phases in the specimens exposed at 600 °C reveal insignificant change in size with increasing the stress, as is in agreement with investigations on the steel GX12 using TEM/EFTEM by Hofer et al. [8] that strain has little influence on the size of Laves phase precipitates. In Fig. 4 (b), the area fraction of Laves phase precipitates slightly decreases with raising the stress from 130 MPa to 230 MPa at 600 °C and the number of particles of the creep specimens under higher stress is slightly lower than those under lower stress, which suggest that stress (or strain) has negative influence on the area fraction and the number of Laves phase, in accordance with the experimental results of steel GX12 during short-term creep [15] . Cui et al. [24] studied stress effect on the precipitation behavior of the Laves phase at the early stage of aging and creep in the Fe–10%Cr–6%W–(Co) alloys and found that the creep stress enhanced the Laves phase precipitation at 600 °C and 650 °C. That is, average size of precipitates in the creep specimen is slightly larger than that in the aged specimens and the number density of Laves phase precipitates increases in the creep specimen. The discrepancy of the above results is attributed to the different microstructure and microstructure evolution during creep for the steels. The Fe–10%Cr–6%W–(Co) steel studied in Ref. [24] is fully ferritic, and creep stress and strain promote the formation of dislocation substructure, which act as nucleation sites for Laves phase, i.e., stress promotes the nucleation of Laves phase in the fully ferritic steel. However, the tested steel X12 in this study and steel GX12 studied in Ref. [15] are martensitic, and have a microstructure of tempered martensite with high density of dislocations and subgrains, which are potential nucleation sites for Laves Phase. Stress and strain reduce the dislocation density and cause growth of subgrains, i.e., stress suppresses the nucleation of Laves phase, which is in accordance with the observations. The thermodynamic calculation in this study has been performed using the software JMatPro [25] , as shown in Fig. 5 . M 23 C 6 , MX, and Laves phase are the equilibrium phases for the experimental material at creep temperature (600 °C/650 °C). These results are in accordance with experimentally found phases except for the as-received state, in which the equilibrium conditions have not been achieved yet. According to Fig. 5 , precipitation of Laves phase may occur at temperatures lower than the solution temperature of Laves phase (698.6 °C), and the volume fraction of Laves phase increases with decreasing the temperature below 698.6 °C. The equilibrium volume fraction of Laves phase at 600 °C (1.04%) is higher than that at 650 °C (0.61%), while the experimental results on the volume fraction of Laves phase is between 0.25% and 0.34% at both temperatures, which is far from that at the equilibrium state. The precipitation of Laves phase is possible to occur at both tempering temperature 570 °C and 690 °C from the thermodynamic calculation result ( Fig. 5 ). However, the Laves phase which has been formed during the first-step tempering process (570 °C/21 h) probably have been dissolved during the process of tempering at higher temperature 690 °C for a longer term 23 h, which promotes the diffusion of Fe, Mo and W atoms. Thus Laves phases formed at 570 °C are negligible. The number of Laves phase would not change with temperature provided that the nucleation process of Laves phase has been finished during the second-step tempering process (690 °C/23 h). Whereas, the experimental result shows that the number of Laves phase decreases with raising temperature from 600 °C to 650 °C ( Fig. 4 (b)). So the nucleation process of Laves phase has not finished after heat-treatment and some new nuclei of Laves phase should be formed during creep exposure. The nucleation of Laves phase in 9–12% Cr steels is believed to be heterogeneous [15,18,20,21] , which is also confirmed by this work. The number of nuclei is related to the potential nucleation sites for Laves phase such as the dislocation networks and subgrain boundaries. The increase in temperature accelerates the process of recovery and recrystallization, and thereby reduces dislocation density and promotes the subgrain growth, i.e., the increase in temperature suppresses the nucleation of Laves phase. Accordingly, the number of nuclei formed in the steel decreases with raising the temperature, in agreement with our experimental results. 4 Conclusions Laves phase precipitates in the creep specimens made of steel X12 during short-term exposure of 200 h at 650 °C and 400 h at 600 °C. They have irregular shape and have stacking faults in them. The SAED patterns of the phase are characteristic of streaked. Two mechanisms of the nucleation and growth for Laves phase present in steel X12 have been found. One is that Laves phases are formed alone on martensite lath boundaries and they are coherent with a grain but grow into the adjacent grain having no rational orientation relationship with them. The other is that Laves phases occur in the regions adjacent to M 23 C 6 particles, and then they grow at the expense of the Cr-rich M 23 C 6 carbides in the vicinity. It is concluded that the former one is the primary mechanism for the formation of Laves phase in steel X12. The PAGBs and martensite lath boundaries are the favorable locations for Laves phase. The size and the number of the phase are strongly dependent on temperature, but show a weaker variation with stress. Furthermore, perhaps it is because Laves phase is one of the thermodynamically stable phases present in steel X12 at 600 °C and 650 °C that Laves phase has been detected in such a short-time. In addition, it is believed that the nucleation process of Laves phase has not been finished after heat-treatment. Temperature has negative influence on the number of nuclei for Laves phase, i.e., the number of nuclei formed in the steel decreases with raising the temperature. Acknowledgement This work was supported by a grant from the Scientific Research Innovation program of Shanghai Educational Committee in PR China (No. 09ZZ16). References [1] T.U. Ken B. Scarlin R.W. Vanstone K.H. Mayer J. Lecomte-Beckers Materials for Advanced Power Engineering, Part I 1998 Jülich GmbH, Forschungszentrum Jülich 53 [2] V.K. Sikka C.T. Ward K.C. Thomas Fabrication, Properties and Applications of Ferritic Steels for High-Temperature Applications 1981 ASM Int. Conf. Production Warrendale, PA [3] K. Maruyama K. Sawada J.-I. Koike ISIJ Int. 41 2001 641 653 [4] J. Hald S. Straub J. Lecomte-Beckers F. Schubert P.J. Ennis Materials for Advanced Power Engineering, 1998 Forschungszentrum Julich, Germany 155 [5] F. Masuyama ISIJ Int. 41 2001 612 625 [6] K. Yamada M. Igarashi S. Muneki F. Abe ISIJ Int. 43 2003 1438 1443 [7] H. Cerjak P. Hofer B. Schaffernak ISIJ Int. 39 1999 874 888 [8] P. Hofer H. Cerjak B. Schaffernak P. Warbichler Steel Res. 69 1998 343 348 [9] Y. Qin G. Götzb W. Blum Z.G. Zhu J. Alloys Compd. 52 2003 260 264 [10] Y. Hosoi N. Wade S. Kunimitsu T. Urita J. Nucl. Mater. 461 1986 141 143 [11] K. Miyahara J.-H. Hwang Y. Shimoide Scr. Metall. Mater. 32 1995 1917 1921 [12] S. Kunimitsu T. Iwamoto A. Hotta Y. Sasaki Y. Hosoi Proc Int. Conf. on Stainless Steels Chiba ISIJ 1991 627 [13] J. Hald Steel Res. 9 1996 369 374 [14] J. Hald L. Korcakova ISIJ Int. 43 2003 420 427 [15] G. Dimmler P. Weinert E. Kozeschnik H. Cerjak Mater. Charact. 51 2003 341 352 [16] L. Korcakova J. Hald M.A.J. Somers Mater. Charact. 47 2001 111 117 [17] Media Cybernetics, Computer software. [18] Q. Li Metall. Mater. Trans. A 37A 2006 89 97 [19] P.J. Ennis A. Czyrska-Filemonowicz Sādhanā 28 2003 709 730 [20] Y. Tsuchida K. Okamoto Y. Tokunaga ISIJ Int. 35 1995 317 323 [21] C. Yang F. Sun L. Zhang J. Shanghai Jiao Tong Univ. (Chin. Ed.) 43 2009 1640 1643 [22] B.A. Senior Mater. Sci. Eng., A 119 1989 5 9 [23] K. Sawada M. Takeda K. Maruyama R. Ishii M. Yamada Y. Naga Mater. Sci. Eng., A 267 1999 19 25 [24] J. Cui I.S. Kim C.Y. Kang K. Miyahara ISIJ Int. 41 2001 368 371 [25] Web article: http://www.sentesoftware.co.uk/biblio.html , 2007.