Influence of smelting processes on precipitation behaviors and mechanical properties of low activation ferrite steels
Z.X. Xia,C. Zhang,H. Lan,Z.G. Yang,P.H. Wang,J.M. Chen,Z.Y. Xu,X.W. Li,S. Liu
DOI: https://doi.org/10.1016/j.msea.2010.09.088
2010-01-01
Abstract:Research highlights ▶ Creep properties could be improved dramatically by control of smelting process. ▶ VIM + ESR smelting process could improve the W macrosegregation. ▶ W could accelerate the transition of M 7 C 3 to M 23 C 6 in RAFM steel. ▶ The synergetic effect of carbides and tungsten depletion of solid solution lead to the decrease in creep resistance. Abstract In this paper, the influence of smelting processes on precipitation behaviors and mechanical properties of CLF-1 (China low activation ferrite) steel was investigated. Mechanical properties of CLF-1 steels melted by vacuum induction melting (VIM, CS25 steel) and vacuum induction melting followed by consumable electrode remelting (VIM + ESR, CS350 steel) had been studied. The results indicated that the steels showed similar values of hardness and tensile strength, however, creep properties of CLF-1 steel were obviously improved by VIM + ESR process. M 23 C 6 and M 7 C 3 Carbides were precipitated in the CS25 steel, and M 23 C 6 carbides were precipitated in CS350 steel after high-temperature tempering at 1013 K for 90 min. Creep properties of CS25 steel were shown to be dramatically deteriorated by the existence of rod M 7 C 3 precipitates, which was due to W macrosegregation in the CS25 steel melted by VIM. M 7 C 3 was replaced by M 23 C 6 after ageing at 823 K for 5000 h. Keywords CLF-1 steel Precipitation behaviors Smelting processes Creep properties Macrosegregation 1 Introduction Reduced activation ferritic/martensitic (RAFM) steel was selected as the preferred candidate for the structural material of fusion reactors [1–3] . Ageing resistance of RAFM steels is vital for application in fusion reactors. The evolution of the microstructure and its effect on mechanical properties after different schedules of heat treatment have been the object of studies, either for tempering or ageing [4,5] as well as for austenitization treatments [6,7] . Suitable dispersion of the fine particles within the matrix is required for excellent creep properties, which can act as barriers to the movement of dislocations and the migrations of lath boundaries [8] . The various creep models [9,10] suggest that the precipitation will reduce creep strength if the new phase coarsens rapidly and is formed at the expense of an existing finer particles or if it leads to withdrawal of elements from solid solution [11,12] . For high creep resistance, a trend in the alloy compositions moves towards decreasing carbon as well as increasing contents of refractory elements such as W [13,14] . Macrosegregation of tungsten usually occurs during steel smelting, it maybe influences the precipitation behaviors during ageing. In Li's study [15,16] , the thermal ageing at 823 K for 2000 h did not degrade the tensile properties and even improved the creep properties for CLAM and JLF-1 steels melted by vacuum induction melting. This phenomenon is different from the results for Eurofer97 steel with the same chemical compositions melted by vacuum induction melting and consumable electrode remelting (VIM + ESR) [17] , which showed that creep rupture tests was not influenced by ageing at 823 K for 20,000 h. The above works paid more attention to the effect of heat treatment on mechanical properties in RAFM steel. However, it is interesting to find that smelting processes also influence precipitation behaviors and mechanical properties of RAFM steel. The aim of the work was to investigate the influence of smelting processes on precipitation behaviors and mechanical properties of CLF-1 steel. 2 Experimental The chemical compositions of CS25 and CS350 steels are listed in Table 1 . CS25 steel was melted in a vacuum induction furnace into an ingot of 25 kg, followed by hot-forging and rolling into a 15-mm-thick plate. CS350 steel was melted by vacuum induction melting and consumable electrode remelting into an ingot of 350 kg, followed by hot-forging and rolling into a 15-mm-thick plate. The as-received state heat treatment includes austenitizing at 1253 K for 45 min, followed by water quenching (WQ), then tempering at 1013 K for 90 min, followed by air cooling (AC). The detailed heat treatment conditions for both steels are given in Table 2 . Thermal ageing treatments were carried out at 823 K to simulate in-service condition and 873 K to provide heavy overageing up to 5000 h in a muffle furnace, in which the samples were vacuum sealed in quartz capsules in order to avoid oxidation. The Vickers hardness was measured at room temperature with a load of 200 g and loading time of 20 s, and the average values were calculated out of 5–10 data points for every result. The miniaturized tensile (SSJ) specimens with a gauge size of 5 × 1.2 × 0.25 mm 3 were machined parallel to the rolling direction. The tensile tests were conducted at an initial strain rate of 6.67 × 10 −4 s −1 at room temperature (RT). The 0.2% proof strength was measured as yield strength. The uniaxial creep tests up to rupture were performed at 823 K with the applied stress of 270 MPa in a vacuum of <1 × 10 −4 Pa using the specimen of the same shape as for the tensile tests. The carbides inherited from the specimens were analyzed. Carbides were extracted electrolytically using a water solution of 5% potassium chloride, 1% citric acid at the current density of 0.03 A/cm 2 and the residues were examined with X-ray diffraction using Cu Kα radiation. The sections perpendicular to rolling direction of specimens was mechanically polished, and then electrochemically etched in methanol containing 10% acetylacetone and 1% lithium chloride. The martensitic microstructure was investigated by optical microscopy (OM), SEM and TEM with EDX. Thin foil and extracted carbon replica were prepared for TEM observation. 3 Results 3.1 Mechanical properties The ultimate and yield tensile strengths of CS25 steel are similar to that of CS350 steel at room temperature, the Vickers hardness also shows a reasonable tendency as shown in Fig. 1 , as it reflects lath width and dislocations density. The steels showed similar values of hardness, ultimate tensile strength and yield tensile strengths (0.2% proof stress) regardless of the material condition. It could be concluded that tensile strength is hardly sensitive to the morphology of precipitates, mainly depend on size distribution and volume fraction of precipitates as well as the size of martensitic packets. Fig. 2 shows creep properties of CS25 steel and CS350 steel for uniaxial tensile creep tests conducted at 823 K under the level of the applied stress 270 MPa. As demonstrated by the figure, significant differences were found in the creep behaviors of both steels in the as-received state. Firstly, the CS25 steel exhibits markedly shorter times to fracture than the CS350 steel in the as-received state. Secondly, the minimum creep rate of CS350 steel is much smaller than that of CS25 steel. After ageing at 823 K for 5000 h, the minimum creep rate decreased and the rupture time increased for CS25 steel, relative to the as-received. However, ageing at 873 K for 5000 h caused a significant increase in the minimum creep rate and decrease in rupture time than the other. The thermal ageing at 823 K for 5000 h did not degrade the tensile properties and even improved the creep properties for CLF-1 steels melted by vacuum induction melting as shown in Fig. 2 . The results of CS25 steel are in good agreement with the results for CLAM and JLF-1steels [15,16] . It also indicated that creep resistance of CLF-1 steel was obviously improved by control of smelting process or heat treatment as shown in Fig. 2 . 3.2 Precipitation behaviors The microstructure of CS25 steel and CS350 steel in the as-received state was fully martensitic with equiaxed prior austenitic grain size 13 μm and 14 μm, respectively (as shown in Fig. 3 ), according to our OM images, the effect of inclusions on creep properties is almost invariable for similar distribution and morphology of inclusion in both steels. According to the SEM image, the morphology and distribution of carbides has obvious difference, the rod precipitates are along grain and packet boundaries in CS25 steel, but there are mostly spherical particles along grain and packet boundaries in CS350 steel as shown in Fig. 4 . The microstructure gave no evidence of difference about lath width and dislocation density in both steel in as-tempered condition as shown in TEM images, Fig. 5 . In addition, it could be deduced that there are similar dislocation density and lath width in both steels because of the same chemical compositions and heat treatment of CS25 steel with that of CS350 steel. The microstructure in both steels consists of thin plates of laths of 300 ± 100 nm width, where rod M 7 C 3 (M = Cr, W, Fe) carbides shown as black arrows are distributed mainly along lath boundaries in Fig. 5 a. However, spherical M 23 C 6 (M = Cr, W, Fe) carbides are distributed mainly along lath and grain boundaries in Fig. 5 b. The identification of the carbides by X-ray diffraction measurements made on the extracted precipitate residues is shown in Fig. 6 . The carbides were identified as MC (M = Ta, V), M 23 C 6 and M 7 C 3 for the CS25 steel as-received, only MC, M 23 C 6 for the steels after ageing at 823 K and 873 K, and MN (M = Ta, V) and M 23 C 6 for the CS350 steel. Fig. 7 shows the distribution of precipitates after different heat treatments for an extracted carbon replica. Most of precipitates in CS25 steel after tempering are rod M 7 C 3 , spherical M 23 C 6 and MC carbides. M 7 C 3 and M 23 C 6 carbides are distributed along packet and prior austenite grain boundaries, while MC carbides are distributed in matrix. The tendency for carbides distribution in other steels is similar. The size distribution and volume fraction of MX precipitates is similar between CS25 steel and CS350 steel, the effect of MX on the difference of mechanical properties could be almost neglected. The chemical composition of chromium carbides along grain or packet boundaries in extraction replica is shown in Fig. 8 . It should be noticed that the larger M 23 C 6 precipitates with W content of 15 wt.% appears during ageing at 823 K for 5000 h, while the agglomeration of coarse M 23 C 6 phase with W content of 57 wt.% takes place during ageing at 873 K for 5000 h. The creep resistance of CS25 steel aged at 873 K is shown to be dramatically deteriorated by the agglomeration of coarse M 23 C 6 phase according to Figs. 2 and 7 . 4 Discussion The creep resistance of RAFM steels is considered to depend on the microstructure. Dispersion of carbides can act as barriers to the movement of dislocations and the migrations of lath boundaries. The effects of the second refining ESR process on Ta were investigated, there is little difference on Ductile-Brittle Transition Temperature (DBTT) of as-VIMed and ESRed F82H because of similar inclusions [16] . The effect of inclusions on creep properties can be neglected due to the similar distribution and morphology. There are similar dislocation density and lath width in both steel because of chemical compositions and heat treatment between CS25 and CS350 steel. Effect of microstructure stability on creep properties mainly depends on precipitations. There are different precipitates namely M 7 C 3 , M 23 C 6 and MX (Ta and V rich), which have been detected in the as-received and aged CLF-1 steels. The results indicate that the main difference of precipitation behaviors between CS25 steel and CS350 steel is the type and distribution of chromium carbides. Besides MX (M = Ta, V; X = C, N) particles within matrix, there are the rod M 7 C 3 carbides distributed mainly along lath boundaries and spherical M 23 C 6 carbides distributed along lath and grain boundaries in CS25 steel, however, it is only spherical M 23 C 6 carbides distributed along lath and grain boundaries in CS350 steel. The formation of rod M 7 C 3 carbides may be due to the lack of W in CS25 steel. Abe's work [17] suggested that M 7 C 3 was precipitated in the 9Cr steel after tempering at 1023 K, with increasing W concentration, M 7 C 3 was replaced by M 23 C 6 , and M 6 C formed in addition to M 23 C 6 . We could assume that distribution of W is not homogeneous as un-melted W is found in tensile fracture in CS25 steel as shown in Fig. 9 , however, this phenomenon do not occur in CS350 steel as the chemical compositions is fully homogeneous. We suggest M 7 C 3 is precipitated in zone with the low W concentration in CS25 steel. The schematic graphic of effect of W concentration on chromium carbides is shown in Fig. 10 . Subsequently ageing at 823 K and 873 K, M 7 C 3 is replaced by M 23 C 6 . M 23 C 6 rapidly coarsens after ageing at 873 K, the pinning effect of carbides on boundaries become much lower than that of the fine M 23 C 6 after tempering and ageing at 823 K. It suggests that aging could increase the stability of chromium carbides, and W could accelerate the transition of M 7 C 3 to M 23 C 6 in RAFM steel as well. Creep cavities were nucleated at precipitates. Creep cavities are more easily nucleated at the coarse and rod precipitates along grain boundaries, the voids of CS25 steel make it easier that the main crack proceeds by rupture, so CS25 steel has much worse creep resistance than CS350 steel. We also can find that the larger M 23 C 6 precipitates with the lower W content appears in CS25 steel during ageing at 823 K for 5000 h, while the agglomeration of coarse M 23 C 6 phase with the higher W content takes place during ageing at 873 K for 5000 h, coarse particles were the crack initiation sites, and their interfaces with matrix were the main propagation path of cracks. The synergistic effect of coarse M 23 C 6 phase and tungsten depletion of solid solution could be the reason for a decrease in creep resistance. It also well interprets the effect of ageing on creep properties of CS25 steel and Li's results [18] . The thermal ageing at 823 K for 5000 h did not degrade the tensile properties and even improved the creep properties for CS25 steel. This is different from the results for Eurofer97 steel [19] , which showed that creep properties was not influenced by ageing at 823 K for 20,000 h, because there are not rod M 7 C 3 carbides both in the as-received steel and the aging steel. It is the main reason of the difference of creep properties between Li's results and Eurofer97 steel. The creep resistance is considered to be affected by microstructural changes and precipitation behaviors. 5 Conclusions The effect of smelting processes on precipitation behaviors and mechanical properties of CLF-1 steel was investigated, the main results are summarized as follows: The high creep strength should be achieved through the suitable size distribution and volume fraction of precipitates, also depends on the type and morphology of precipitates. Precipitation behaviors and creep properties could be improved dramatically by control of smelting process. The formation of rod M 7 C 3 carbides is due to macrosegregation of W in as-received CS25 steel, M 7 C 3 was precipitated in zone with the low W concentration after tempering in CS25 steel, however, it do not occur in CS350 steel. W could accelerate the transition of M 7 C 3 to M 23 C 6 in RAFM steel. The creep strength is high at 823 K up to 5000 h due to the stabile microstructure and the presence of tiny precipitates of MX and relatively small M 23 C 6 with the lower W concentration. The agglomeration of coarse M 23 C 6 carbides with the higher W concentration occur in CS25 steel aged at 873 K. The synergistic effect of M 23 C 6 and tungsten depletion of solid solution is the reason for a decrease in creep resistance. Acknowledgement This work was carried out with great financial supports from National Basic Research Program of China (No. 2010CB731600 ). References [1] R.L. Klueh J. Nucl. Mater. 378 2008 159 166 [2] A. Kimura R. Kasada A. Kohyama H. Tanigawa T. Hirose K. Shiba S. Jitsukawa S. Ohtsuka S. Ukai M.A. Sokolov R.L. Klueh T. Yamamoto G.R. Odette J. Nucl. Mater. 367 2007 60 67 [3] J.N. Yu Q.Y. Huang F.R. Wan J. Nucl. Mater. 367 2007 97 101 [4] R. Jarayam R.L. Klueh Metall. Mater. Trans. 29A 1998 1551 1558 [5] Y. de Carlan A. Alamo M.H. Mathon G. Geoffroy A. Castaing J. Nucl. Mater. 283 2000 672 676 [6] F. Abe T. Noda M. Okada J. Nucl. Mater. 195 1992 51 67 [7] A. Alamo J.C. Brachet A. Castaing C. Lepoittevin F. Barcelo J. Nucl. Mater. 258 1998 1228 1235 [8] P. Fernandez A.M. Lancha J. Lapena R. Lindau M. Rieth M. Schirra Fusion Eng. Des. 75–79 2005 1003 1008 [9] F.R. Larson J. Miller Trans. ASME 74 1952 756 761 [10] R.L. Orr O.D. Sherby J.E. Dorn Trans. ASM 46 1954 113 128 [11] Z. Peng L. Cai F. Peng Y. Hu F. Chen Acta Metall. Sin 46 2010 429 434 [12] Z. Peng Y. Dang F. Peng Acta Metall. Sin. 46 2010 435 443 [13] F.-S. Yin W.-S. Jung W.-S. Chuang Scripta Mater. 57 2007 469 472 [14] T. Masaka F. Abe K. Sawada Nature 424 2003 294 296 [15] Y. Li Q. Huang Y. Wu T. Nagasaka T. Muroga J. Nucl. Mater. 367 2007 117 121 [16] A. Sawahata H. Tanigawa M. Enomoto J. Jpn. Inst. Metal 72 2008 176 180 [17] F. Abe H. Araki T. Noda Metall. Mater. Trans. 22A 1991 2225 2235 [18] Y. Li T. Nagasaka T. Muroga Q. Huang Y. Wu J. Nucl. Mater. 386 2009 495 498 [19] M. Rieth, M. Schirra, A. Falkenstein, P. Graf, S. Heger, H. Kempe, et al., Eurofer97-Tensile, Charpy, Creep and Structural Tests, FZKA 6911, 2003.