In situ and ex situ neutron diffraction study on deformation behavior of high-nitrogen, Ni-free duplex stainless steel

In situ and ex situ neutron diffraction study on deformation behavior of high-nitrogen, Ni-free duplex stainless steel

Available online at www.sciencedirect.com Scripta Materialia 67 (2012) 141–144 www.elsevier.com/locate/scriptamat In situ and ex situ neutron diffrac...

490KB Sizes 0 Downloads 14 Views

Available online at www.sciencedirect.com

Scripta Materialia 67 (2012) 141–144 www.elsevier.com/locate/scriptamat

In situ and ex situ neutron diffraction study on deformation behavior of high-nitrogen, Ni-free duplex stainless steel Tae-Ho Lee,a,⇑ Heon-Young Ha,a Jun-Yun Kang,a Byoungchul Hwang,a Wanchuk Woob and Eunjoo Shinb a

Ferrous Alloy Department, Advanced Metallic Materials Division, Korea Institute of Materials Science, 797 Changwondaero, Changwon 641-831, South Korea b Neutron Physics Department, Korea Atomic Energy Research Institute, PO Box 105, Yuseong, Daejeon 305-600, South Korea Received 8 March 2012; revised 30 March 2012; accepted 30 March 2012 Available online 6 April 2012

In situ and ex situ neutron diffraction were used to investigate the deformation behavior of high-nitrogen, nickel-free duplex stainless steel. During in situ deformation, both ferrite and austenite deformed plastically to the same degree. The stacking fault energy of austenite was evaluated to be 36.2 mJ m–2 from ex situ neutron profiles, which correlated well with the observed deformation microstructure of twin formation. A significant decrease in the effective particle size of austenite was due to the formation of deformation twins. Ó 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Stainless steels; Nitrogen; Neutron diffraction; Twinning; Stacking fault energy

Duplex stainless steels (DSS) composed of comparable proportions of ferrite (a) and austenite (c) offer an attractive combination of properties, including high strength and excellent resistance to stress corrosion cracking [1,2]. The Ni, which improves both c stability and formability, is contained at levels ranging from 4 to 7 wt.% in the currently popular DSS. Reducing the use of Ni in DSS not only provides substantial economical saving but also avoids the hazard of causing an allergic reaction in human skin [1–3]. Over the years, efforts have been made to develop cost-effective DSS (termed “lean DSS”). For this, N has been considered as a candidate alloying element to achieve a reduction in Ni and improvements in both the strength and corrosion resistance of DSS [2,3]. Due to its theoretical and practical importance, extensive studies on the deformation behavior of DSS have been carried out [4–7]. Since the two constituent phases have different elastic/plastic properties, stress partitioning usually occurs under plastic deformation. Notable gains in understanding the inhomogeneity of deformation in DSS have been made through the

⇑ Corresponding

author. Tel.: +82 55 280 3434; fax: +82 55 280 3599; e-mail: [email protected]

application of in situ diffraction [4–6]. Despite the many investigations, there are still inconsistencies regarding the stress partition between a and c. It has been shown that a is considered to be the stronger phase and c as the softer, more ductile phase [4,5]. However, other experiments have reported the opposite to be the case, i.e. c is the stronger phase [6]. It is conceivable that these disagreements between the previous studies may arise from the differences in chemical composition, anisotropy and texture [4–7]. The N, acting as c stabilizer, is mainly partitioned into c phase and thus promotes solid-solution hardening of c. Since recently developed DSS are alloyed with a significant amount of N to increase the strength of the material and to balance the phase fractions, the inhomogeneous deformation behavior of high-N DSS can be considered of practical importance. In an effort to understand the inhomogeneity in the deformation behavior of high-N, Ni-free DSS, a complementary study using in situ and ex situ neutron diffractions was carried out. The stacking fault energy (SFE) of c was evaluated from Rietveld whole-profile fitting, combined with double-Voigt size-strain analysis [8,9], and the deformation behavior of the nitrogenstrengthened c in DSS was discussed in terms of SFE and the partitioning of N.

1359-6462/$ - see front matter Ó 2012 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.scriptamat.2012.03.043

142

T.-H. Lee et al. / Scripta Materialia 67 (2012) 141–144

The investigated material was a high-N, Ni-free DSS with the following composition (in wt.%): 19.04Cr, 6.64Mn, 2.89Mo, 2.09W, 0.49N, balance Fe. The alloy was fabricated utilizing a pressurized induction melting furnace (VIM 4 III-P, ALD, Germany). After homogenization at 1300 °C for 2 h, the ingots were hot-rolled into sheets of 4 mm thickness, followed by water quenching. Specimens were solution-treated at 1200 °C for 30 min, corresponding to the temperature of a:c = 50:50. Angular-dispersive in situ neutron diffraction was performed using a residual stress analysis diffractometer equipped with a deformation device enabling tensile straining up to a maximum load of 20 kN at the Korea Atomic Energy Research Institute (KAERI). A bent perfect Si (2 2 0) monochromator and a high-resolution position-sensitive detector were used in the in situ experiment. The ð1 1 1Þc , ð2 0 0Þc , ð1 1 0Þa and ð2 0 0Þa reflections were measured with a crosshead speed of 0.01 mm s1 under a controlled load of 1 kN in the elastic region, followed by displacement control corresponding to a strain of 0.1 after yielding. The lattice plane strains (LPS), eh k l , were calculated by the relation: eh k l ¼ ðdh k l  d0h k l Þ=d0h k l ¼  cot hh k l Dhh k l where dh k l and d0h k l are the measured and stress-free LPS, respectively [4–6]. The specimen (ASTM E8M) was tensile-tested at room temperature with a nominal strain rate of 1.67  103 s–1 using a servohydraulic machine (INSTRON 5882, Canton, USA). The ex situ neutron diffraction experiment was performed on tensile-strained (true strains of 0.1–0.3) specimens using a high-resolution powder diffractometer at KAERI. The neutron beam was monochromatized to a wavelength of ˚ . The ex situ neutron profiles ((1 1 1), (2 0 0), 1.8344 A (2 2 0), (3 1 1) and (2 2 2) for c; (1 1 0), (2 0 0), (2 1 1) and (2 2 0) for a) were analyzed by Rietveld whole-profile fitting and double-Voigt size-strain analysis, following our previous approach [8]. To measure the distribution of the alloying elements in the a and c pgarer, wavelength-dispersive X-ray analysis was performed in an electron-probe microanalyzer (EPMA; Camebax SX 100, Cameca, France). Thin foils for scanning transmission electron microscopy (STEM) were prepared in a twin-jet electrolytic polishing apparatus using a solution containing 15% perchloric acid and 85% methanol. They were examined in a scanning transmission electron microscope (JEM 2100F, JEOL, Japan) at 200 kV. When the measured eh k l are plotted as a function of the applied stress, the eh k l before tensile loading are tensile (positive) for c and compressive (negative) for a [4,6]. Due to the difference in thermal expansion coefficients, the residual tensile stress in c, generated during cooling from the solution-annealing temperature, is balanced by the compressive stress in a. Thus, the variation of eh k l with the applied stress was replotted considering the thermal residual stresses following the approach by Harjo et al. [4]. To clarify the dissimilar responses of the two phases to plastic deformation, the applied nominal stresses were converted to true stress following the

recent approach by Baczmanski et al. [10], which can indirectly take the instantaneous reduction in the cross-section into account. It was found that the calculated true stress correlated well with the true stress measured from tensile testing. Figure 1(a) and (b) shows the changes in eh k l and normalized peak intensities as a function of applied true stress for the ð1 1 1Þc , ð2 0 0Þc and ð1 1 0Þa and ð2 0 0Þa reflections. At the initial stage of tensile deformation, an almost linear response of lattice strain to applied stress is observed for all the h k l planes before loading up to 650 MPa, which indicates that the elastic response is dominant during this stage in both phases. The different slopes for the different h k l reflections are associated mainly with the difference in individual diffraction elastic modulus. When the loading is above 650 MPa, deviations from the elastic linear response occur in the eh k l for a and c, implying the onset of macroscopic plastic flow in both phases. After yielding, the slope of LPS vs. applied stress increases steeply, the increase being more pronounced in the elastically soft (2 0 0) reflections of both phases. In the plastic region, the nonlinearity starts at the same macrostresses (650 MPa), indicating that both c and a phases deform plastically to almost the same degree. The critical macrostress corresponding to the onset of plastic deformation correlates well with the yield strength (688 MPa) of the investigated alloy measured from tensile testing. As shown in Figure 1(b), the peak intensities of the ð1 1 1Þc , ð2 0 0Þc and ð1 1 0Þa and reflections start to increase, whereas that

Figure 1. Changes in (a) lattice plane strain ðeh k l Þ and (b) normalized peak intensities as a function of applied stress for ð1 1 1Þc ; ð2 0 0Þc and ð1 1 0Þa ; ð2 0 0Þa reflections measured from in situ neutron diffraction.

T.-H. Lee et al. / Scripta Materialia 67 (2012) 141–144

of ð2 0 0Þa decreases, after a critical macrostress of 650 MPa. These dissimilar responses can be explained by the formation of a deformation texture under tensile deformation, and the trends match with other previous works [4–6]. Many attempts have been made to investigate the inhomogeneous deformation behavior of DSS using in situ neutron as well as X-ray diffraction. It is known that there are several factors affecting the flow behavior and the stress partitioning in DSS, but it is not easy to separate the individual contributions from the different factors. In situ experiments on a series of DSS with different phase fractions, carried out using angular-dispersive neutron diffraction, showed that the plastic deformation started preferentially in the c phase, implying that c is relatively softer than the a phase [4]. Jia et al. [5] confirmed the a as the harder phase, but anisotropic flow behavior was also observed, depending on the orientation. Moverare and Ode´n [6] reported that the loading direction played an important role in the stress partitioning of DSS due to the crystallographic texture of a: for loading in the transverse direction, preferential plastic deformation occurred in c, while simultaneous plastic deformation of both phases was observed at the same strain under loading parallel to the rolling direction, which is in accordance with our results. Another factor contributing to a complex stress partitioning between the two phases is the N content. It is reported that the a is harder than c if the N content is less than 0.12 wt.%, but the opposite is the case when the N content exceeded 0.12 wt.% [6]. To investigate the partitioning of alloying elements, the chemical composition of a and c were analyzed using EPMA. The measured concentrations in a and c are as follows: (i) a: 19.26Cr, 6.49Mn, 3.27Mo, 1.80W, 0.11N; (ii) c: 20.23Cr, 7.63Mn, 2.43Mo, 1.17W, 1.08N (all in wt.%). EPMA analysis shows that most of N was partitioned into c due to the higher solubility of N in face-centered cubic structures [11]. The hardness for a and c in the solution-annealed specimen was measured to be 281 and 302 Hv, respectively. Although the critical N content reported in previous studies [6,8] cannot be directly applied to other alloy systems due to the difference in chemical composition and thermal history, it is conceivable that the higher the N content, the harder the c phase. Two important sources of peak broadening, effective particle size (EPS, the size of coherently diffracting domains) and mean-squared (MS) strain, were evaluated from the double-Voigt size-strain analysis [9] on ex situ neutron profiles ((1 1 1) and (2 2 2) for c; (1 1 0) and (2 2 0) for a). Figure 2(a) and (b) presents the changes in EPS and MS strains of a and c plotted as a function of tensile strain. In a general trend, the EPS of both phases decreases with increasing strain, implying that the decrease is due to the deformation faults (stacking faults (SFs) and twins). The EPS of c decreases rapidly at small strains (up to e = 0.1), then at higher strain. The EPS of a, on the other hand, gradually decreases up to e = 0.2, then approaches a constant value. As shown in Figure 2(b), the MS strains of both phases increase with the degree of deformation, the increment being more pronounced in c.

143

Figure 2. Variations in (a) effective particle size and (b) mean-squared strain for a and c plotted as a function of tensile strain measured from ex situ neutron diffraction.

The changes in deformation microstructure of a and c were characterized by TEM. In the initial stage of deformation of c, perfect dislocations dissociated into partial dislocations, and SFs formed between the dissociated partials. As the deformation proceeded, the SFs overlapped with each other or extended preferentially along f1 1 1gc , as in the case of austenitic steels [12]. Figure 3(a) represents a STEM bright-field (BF) image and corresponding selected area diffraction (SAD) pattern of c taken from the specimen strained to e = 0.1 under a 1 1 1 two-beam condition with a zone axis of h1 1 0i. Pronounced deformation twins were identified and have a f1 1 1gh1 1 2i crystallographic component. In the later stage of deformation, conjugate twin systems were also activated and some of them intersected with primary twins. Figure 3(b) shows a STEM micrograph of a taken from the same specimen viewed in the 0 1 1 two-beam condition with a zone axis of h1 1 1i. At the beginning of deformation, a single slip system, {1 1 2}h1 1 1i, is observed, and characteristic features of a slip band structure can be distinguished. As deformation goes on, another slip system is also activated and the intersection of two slip systems can produce the formation of dislocation jog, as indicated by dotted circles. Further deformation produces a complex multiplication of dislocations. TEM observations show that the pronounced decrease in EPS of c can be explained by the formation of deformation twins, while the small decrease in EPS of a corresponds to the formation of a slip band, which can hardly act as incoherent domain boundaries. The SFE of c was evaluated for ex situ neutron diffracted profiles following our previous approach [8]. Figure 4 shows a plot of MS strain vs. stacking fault

144

T.-H. Lee et al. / Scripta Materialia 67 (2012) 141–144

Figure 3. (a) STEM BF image and SAD pattern of c taken from the specimen strained to e = 0.1 under the 1 1 1 two-beam condition with a zone axis of h1 1 0i; (b) STEM micrograph of a taken from the same specimen viewed in the 0 1 1 two-beam condition with a zone axis of h1 1 1i.

Figure 4. Plot of MS strain vs. SFP, calculated from double-Voigt sizestrain analysis on ex situ neutron profiles.

probability (SFP) wherein different values of MS strain and SFP correspond to different amounts of tensile deformation. The individual values of both MS strain and SFP have a positive correlation with the degree of deformation, and the SFE could be determined by slopes through a linear regression, incorporating the origin, where the SFP and MS strain of a strain-free sample are assumed to be zero. In the calculation, the chemical composition of c measured from EPMA was used, the

other input parameters being G = 77 GPa, A = 3.78 and K 1 1 1 x0 ¼ 6:6 [13]. The calculated SFE of c in the investigated alloy was 36.2 ± 0.4 mJ m–2. This SFE value falls within the region where the deformation is controlled by deformation twinning, which is in good accordance with the TEM observation. Recently, Choi et al. [14] reported that the SFE measured in austenitic stainless steels could successfully be applied to predict the deformation microstructure (strain-induced martensitic transformation) of the c phase in DSS. Apart from the resolution assured and limitation imposed in the evaluation of SFE, it is conceivable that the deformation mode of c in DSS can also be predicted according to its SFE value. The deformation behavior of high-N, Ni-free DSS was investigated by means of both in situ and ex situ neutron diffraction, and by TEM. During in situ tensile deformation, both ferrite and austenite deformed plastically to the same degree. The stacking fault energy of austenite was evaluated from Rietveld whole-profile fitting combined with double-Voigt size-strain analysis on ex situ neutron profiles, and corresponded to the region of deformation twinning (36.2 mJ m–2). TEM observation showed that a significant decrease in the effective particle size of austenite was due to the formation of deformation twins. It is suggested that the combined analysis of in situ and ex situ neutron diffraction can help achieve a comprehensive understanding of the deformation behavior of DSS. This work was financially supported by the Ministry of Knowledge Economy, Korea. The authors would like to express their gratitude to Mr. Jong-In Bae for his valuable support in experimental works. [1] H.D. Solomon, T.M. Devine, in: R.A. Lula (Ed.), Duplex Stainless Steels Conference Proceedings, ASM, Metals Park, OH, 1984, pp. 693–756. [2] J.O. Nilsson, Mater. Sci. Technol. 8 (1992) 685–700. [3] T.-H. Lee, H.-Y. Ha, B. Hwang, S.-J. Kim, Metall. Mater. Trans. A 43 (2012) 822. [4] S. Harjo, Y. Tomota, P. Lukas, D. Deov, M. Vrana, P. Mikula, M. Ono, Acta Mater. 49 (2001) 2471. [5] N. Jia, R.L. Peng, D.W. Brown, B. Clausen, Y.D. Wang, Metall. Mater. Trans. A 39A (2008) 3134. [6] J.J. Moverane, M. Ode´n, Metall. Mater. Trans. A 33A (2002) 57. [7] J. Foct, N. Akdut, Scripta Metall. 29 (1993) 153. [8] T.-H. Lee, E. Shin, C.-S. Oh, H.-Y. Ha, S.-J. Kim, Acta Mater. 58 (2010) 3173. [9] D. Balzar, H. Ledbetter, J. Appl. Crystallogr. 26 (1993) 97. [10] A. Baczmanski, L. Le Joncour, B. Panicaud, M. Francois, C. Braham, A.M. Paradowska, S. Wronski, S. Amarab, R. Chirone, J. Appl. Crystallogr. 44 (2011) 966. [11] J.-O. Nilsson, P. Kangas, T. Karlsson, A. Wilson, Metall. Mater. Trans. A 31A (2000) 35. [12] T.-H. Lee, C.-S. Oh, S.-J. Kim, S. Takaki, Acta Mater. 55 (2007) 3649. [13] S. Lin, H.M. Ledbetter, Mater. Sci. Eng. A 167 (1993) 81. [14] J.Y. Choi, J.H. Ji, S.W. Hwang, K.-T. Park, Mater. Sci. Eng. A 528 (2011) 6012.