Microalloying Effect on the Activation Energy of Hot Deformation

www.steel-research.de TECHNICAL NOTE Microalloying Effect on the Activation Energy of Hot Deformation Moo‐Young Seok, In‐Chul Choi, Yakai Zhao, Dong‐Hyun Lee, Jung‐A. Lee, and Jae‐il Jang
Although extensive research has been made for evaluating the activation energy Q of hot deformation in microalloyed steels, almost no attempt has been made for directly comparing the influence of each element. In this study, through a series of hot compression tests of four different microalloyed steels, we have systematically explored the influences of Nb, V, Ti, and Mo on the Q and have directly compared the effect of each element in a quantitative manner. Hot working process has been widely performed to obtain excellent mechanical performance of steels by grain refinement and microstructure homogenization.[1,2] During the process, due to the low stacking fault energy of austenite phase, dynamic recrystallization (DRX) rather than dynamic recovery (DRC) plays an important role in controlling microstructure and mechanical properties.[3] Thus, information about representative flow behavior during DRX-included hot deformation is essential for better understanding and developing of hot working process in the steel industry.[3,4] One of the most popular ways to describe the effects of temperature and strain rate on hot deformation is to link the exponent-type Zener–Hollomon parameter Z to the flow stress s as[5–7]   Q ¼ f ðsÞ ð1Þ Z ¼ e_  exp RT where e_ is the strain rate (in s1), T is the absolute temperature (in K), R is the gas constant (8.314 J mol1 K1), and Q is the activation energy of hot deformation (in J/mol). It is well known that in microalloyed steels, the alloying elements in austenite (such as Nb, V, Ti, and Mo) can induce the retardation of DRX and thus can seriously change the Q. For example, Nb and Ti atoms and their precipitates are known to retard the DRX by suppressing the grain boundary migration.[8,9] In addition, V and Mo can tackle the DRX by inducing coarse prior austenite grains and densely-distributed fine MC-type precipitates, respectively.[10,11] Despite extensive research to evaluate Q in specific microalloyed steels, somewhat interestingly, [
] M.-Y. Seok, I.-C. Choi, Y. Zhao, D.-H. Lee, J.-A. Lee, Prof. J.-I. Jang Division of Materials Science and Engineering, Hanyang University, Seoul 133-791, Republic of Korea Email: jijang@hanyang.ac.kr DOI: 10.1002/srin.201400255 ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim almost no attempt has been made to directly compare the influence of each element in a quantitative manner, which became a motive of this study. In this letter, we report our experimental comparative study on the relative effects of Nb, V, Ti, and Mo on the Q of hot deformation in microalloyed steels. Chemical compositions of the examined low-carbon micro-alloyed steels are listed in Table 1. Cylindrical samples with 12 mm in length and 10 mm in diameter are prepared for hot compression tests. Before compression, austenizing treatment was performed at 1250 8C for 300 s to dissolve the precipitates in the samples. Then, the samples were cooled down to the testing temperature at a cooling rate of 5 8C s1 and maintained for 30 s before testing for the temperature stabilization. Computer-controlled hot compression tests were carried out using a thermo-mechanical simulator Gleeble 1500 (Dynamic Systems, Inc., Poestenkill, NY) in the temperature range of 950–1100 8C at an interval of 50 8C. The maximum strain and strain rate were fixed as 0.8 and 0.1 s1, respectively. All the tests were made under Ar gas environment to avoid the oxidation, and tantalum plates were attached to the samples to minimize the friction effect. Figure 1 shows flow curves obtained at four different temperatures. While true stress decreases with increasing temperature in all examined steels, typical DRX behavior (exhibiting a peak stress, sp, followed by gradual decrease to a steady-state stress, sss) is observed only at 1050 and 1100 8C. Thus, following DRX analysis is made only for the curves at 1050 and 1100 8C. The function of flow stress, f(s), in Equation (1) can be described as hyperbolic sine law for a wide range of hot deformation stress:[3,6,7,12–14] Z ¼ f ðsÞ ¼ A  fsinhða  sÞgn ð2Þ where n and a are material constants. Note that there are different descriptions of f(s) depending on the stress level; steel research int. 86 (2015) No. 7 817 TECHNICAL NOTE www.steel-research.de Class C Mn Si P S Al Nb V Ti Mo N Nb 0.06 1.5 0.2 0.015 0.003 0.03 0.05 — — — 0.006 Nb–V 0.05 0.05 — — Nb–V–Ti 0.05 0.05 0.02 — Nb–V–Ti–Mo 0.05 0.05 0.02 0.2 Table 1. Chemical compositions (wt%) of the investigated steels. Figure 1. True stress versus true strain curves obtained from hot compression tests at a strain rate of 0.1 s1; a) Nb steel, b) Nb–V steel, c) Nb–V–Ti steel, and d) Nb–V–Ti–Mo steel. i.e., a power-law function for relatively low stresses (as < 0.8) and exponential-law function for relatively high stresses (as > 1.2).[12] The stress multiplier a and the exponent n of Equation (2) are typically determined with the stress data obtained under different strain rates.[3] Since the strain rate was fixed here, a  0.015 and n  5 is adopted following previous studies.[3,13,15] Flow stress s of Equation (2) can be either peak stress (sp) or steady-state stress (sss), both of which are often used for analyzing DRX together with other characteristic stresses such as critical stress (sc) and saturation stress (ssat).[16] For determining sp and sss, first the flow curves in Figure 1 were smoothened by polynominal fitting after removing elastic portion. Then, each smoothened curve was re-plotted as work hardening rate u (¼ds/de) versus s. Figure 2(a) shows representative example of the u–s curve 818 steel research int. 86 (2015) No. 7 (for Nb steel at 1050 and 1100 8C). In the curve, sp and sss can be determined as the stress at the first and second point of u ¼ 0, respectively.[16] Obtained values of sp and sss are summarized in Figure 2(b). Note that the critical stress for the initiation of DRX (sc) is the stress at the inflection point of u–s curve (i.e., the minimum point of the d2u/ds2 vs. s plot), and sc/sp values of the examined steels are close to each other in the range of 0.865–0.921, which corresponds to ec/ep (that can be determined by flow curves in Figure 1) of 0.382–0.518. Both sc/sp and ec/ep values are in agreement with the literature data.[3,17] In the present study, sp rather than sss was adopted as s of Equation (2) for two reasons; first, the strain corresponding to sp is lower than that to sss, and second, sp is known to be more important for industrial process.[17] As shown in Figure 2(b), sp obtained here is in the range of ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.steel-research.de TECHNICAL NOTE Figure 3. Influence of microalloying on a) experimentally obtained Q and b) amount-corrected Q. Figure 2. Evaluation of characteristic stresses: a) representative example showing how to determine the peak stress and steadystate stress from u to s plot (for Nb steel); b) summary of the obtained peak stresses and steady-state stresses. 70–97 MPa and thus as is 1.05–1.46, implying that application of hyperbolic sine function (Equation (2)) is more appropriate than that of power-law or exponentiallaw function. Combining Equation (1) and (2) and taking natural logarithm lead to    Q 1 ln e_ þ ¼ ln A þ nln sinhðas p Þ R T ð3Þ from which, at a given e_ , Q could be calculated from the slope of the relationship between (1/T) and ln{sinh(asp)}. Obtained Q values are summarized in Figure 3(a) where literature Q value of C-Mn non-microalloyed steel (280 kJ mol1)[14] is also provided. It is obvious that microalloying significantly affects Q. Since the added amount of each alloying element is different from each other, it is need to be normalized for directly comparing the effect of each element. Thus, a simple regression was performed, resulting in the following equation of the Q. Q ½kJ mol1  ¼ 280 þ 2:29  103 ½Nb þ 1:62  103 ½V þ 2:73  103 ½Ti þ 3:10  102 ½Mo ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim where [element] indicates the dissolved content in steels (in at%). This amount-corrected Q values are shown in Figure 3(b). It is seen that the Ti shows the largest influence, the effects of Nb and V are the second and third, respectively, and the influence of Mo is the smallest among the four elements; i.e., the microalloying effect on Q is in the order of “Mo < V < Nb < Ti.” This also suggests that, since Q is proportional to the decreasing amount of sp during increasing T,[13] sp can be reduced more largely by addition of Ti, Nb, and V rather than that of Mo. From all above results, the hot deformation behavior of the examined steels can be summarized as follows; Nb steel: Nb-V steel:   3:49  106 e_ exp RT ¼ 4:62  1011 fsinhð0:015  s p Þg5   4:38  106 e_ exp RT ¼ 1:48  1015 fsinhð0:015  s p Þg5   5:01  106 Nb-V-Ti steel: e_ exp RT Nb-V-Ti-Mo steel: ð4Þ ¼ 1:94  1017 fsinhð0:015  s p Þg5   5:37  106 _eexp RT ¼ 4:66  1018 fsinhð0:015  s p Þg5 : steel research int. 86 (2015) No. 7 819 TECHNICAL NOTE www.steel-research.de Acknowledgments This research was supported in part by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2013R1A1A2A10058551), and in part by the Human Resources Development program of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government (MOTIE) (No. 20134030200360). Received: July 25, 2014; Published online: October 24, 2014 Keywords: microalloyed steel; dynamic recrystallization; hot deformation References [1] J. J. Jonas, Mater. Sci. Eng. A 1994, 184, 155. [2] H. Mirzadeh, J. M. Cabrera, A. Najafizadeh, Metall. Mater. Trans. A 2012, 43, 108. [3] H. Mirzadeh, J. M. Cabrera, J. M. Prado, A. Najafizadeh, Mater. Sci. Eng. A 2011, 528, 3876. [4] M. Frommert, G. Gottstein, Mater. Sci. Eng. A 2009, 506, 101. 820 steel research int. 86 (2015) No. 7 [5] C. Zener, H. Hollomon, Appl. Phys. 1944, 15, 22. [6] G. E. Dieter, Mechanical Metallurgy, 2nd Ed., McGraw-Hill Book Co., New York 1976. [7] S. F. Medina, C. A. Hernandez, Acta Mater. 1996, 44, 137. [8] S. Yamamoto, C. Ouchi, T. Osuka, Proceedings of the International Conference on Thermomechanical Processing of Microalloyed Austenite, Pittsburgh, PA, USA, 1982. [9] M. Elwazri, P. Wanjara, S. Yue, Mater. Sci. Eng. A 2003, 339, 209. [10] W. G. Leslie, The Physical Metallurgy of Steels, McGraw-Hill, London 1981. [11] W. B. Lee, S. G. Hong, C. G. Park, K. H. Kim, S. H. Park, Scripta Mater. 2000, 43, 319. [12] M. Meysami, S. A. A. A. Mousavi, Mater. Sci. Eng. A 2011, 528, 3049. [13] H. J. McQueen, N. D. Ryan, Mater. Sci. Eng. A 2002, 322, 43. [14] S. F. Medina, M. I. Vega, M. Gomez, P. P. Gomez, ISIJ Int. 2005, 45, 1307. [15] J. M. Cabrera, A. Al, J. J. Omar, J. M. Jonas, Metall. Mater. Trans. A 1997, 28A, 2233. [16] J. J. Jonas, X. Quelennec, L. Jiang, E. Martin, Acta Mater. 2009, 57, 2748. [17] E. I. Poliak, J. J. Jonas, ISIJ Int. 2003, 43, 684. ß 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim