Irregular or Smooth Grain Boundaries Evolved after ... · Keywords: abnormal grain growth,...
4
Irregular or Smooth Grain Boundaries Evolved after Secondary Recyrstallization of Fe-3%Si Steel Hyung-Ki Park 1,2 , Chang-Soo Park 1 , Tae-Wook Na 1 , Chan-Hee Han 2 and Nong-Moon Hwang 1,+ 1 Department of Materials Science and Engineering, Seoul National University, Seoul 151-742, Korea 2 POSCO Technical Research Laboratories, POSCO, Pohang, Korea Irregular or smooth grain boundaries after secondary recrystallization are formed depending on whether interpass aging is adopted or not during cold rolling of Fe-3%Si steel. Interpass aging induces the primary texture of high intensity of {111}©112ª and {411}©148ª orientations, which have a ∑9 relation with respect to the Goss orientation and therefore are favorable for Goss grains to grow by solid-state wetting. Under this condition Goss grains come in contact with each other by wetting without leaving any small grains in between, resulting in irregular boundaries. [doi:10.2320/matertrans.M2011268] (Received August 30, 2011; Accepted December 27, 2011; Published February 15, 2012) Keywords: abnormal grain growth, secondary recrystallization, grain boundary wetting, interpass aging 1. Introduction Abnormal grain growth (AGG), which is also referred to as secondary recrystallization, often occurs in polycrystalline metals. During AGG only a few grains grow abnormally consuming the neighboring matrix grains. This phenomenon is particularly famous in Fe-3%Si alloy because the core loss of Fe-3%Si can be minimized by properly controlling AGG. Since the first report of the selective AGG of Goss grains in Fe-3%Si alloy by Goss in 1935, 1) many efforts have been made to understand the mechanism because of its techno- logical importance and scientific interest. In spite of intensive efforts, however, the underlying principle has not yet been clarified. The selective AGG of Goss grains has been approached mainly based on the high mobility of grain boundaries shared by Goss grains with other matrix grains. 2-4) However, misorientation analyses between Goss grains and matrix grains do not support this assumption. 5,6) On the other hand, Hwang et al. 7-10) has suggested the sub-boundary enhanced solid-state wetting for selective AGG of Goss grains, where the growth advantage of abnormally-growing grains is approached in terms of grain boundary energy. According to this mechanism, if a grain has sub-boundaries of very low energy, the probability of solid-state wetting becomes so high that the grain containing sub-boundaries undergo selective AGG. In relation to this possibility, Park et al. 11) observed that sub-boundaries exist exclusively in abnormally-growing Goss grains using transmission electron microscopy. Also, Dorner et al. 12) confirmed the existence of sub-boundaries exclusively within Goss grains after primary recrystallization of a sheet sample reduced 89% by cold rolling of a single crystal of Goss orientation. The solid-state wetting mech- anism could successfully explain many puzzling structures such as the formation of island and peninsular grains at the growth front 7,13) and the pancake-shaped growth of Goss grains in highly grain-oriented electrical steel of Fe-3%Si. 14) Normally smooth grain boundaries are evolved after secondary recrystallization of Fe-3%Si steel, whereas highly irregular grain boundaries are evolved if interpass aging is adopted during cold rolling. Interpass aging, which refers to short time aging treatments at 150-300°C between passes of the single stage cold rolling has been reported to improve the magnetic properties of highly grain-oriented electrical steel. 15-17) Interpass aging has been studied extensively but its effect is not clearly understood. Quite different micro- structures are evolved after secondary recrystallization, depending on whether interpass aging is adopted or not: highly irregular grain boundaries are produced with interpass aging whereas smooth grain boundaries are produced without interpass aging. The purpose of this work is to study why irregular grain boundaries are evolved after secondary recrystallization in the Fe-3%Si sheet specimens undergoing interpass aging. Rather than studying the bene- ficial effect of the interpass aging, our study is limited to the mechanism as to the different morphology evolution of grain boundaries between the two processes with and without interpass aging. 2. Experimental Procedure An ingot of Fe-3%Si, which employed AlN as an inhibitor of grain growth to produce the highly grain-oriented electrical steel, was used as a starting material. The steel ingot was hot rolled to 2.3 mm and the thickness was further reduced to 0.3 mm by one stage cold rolling method, which is equivalent to a total engineering thickness reduction of ¾ = 87%. To investigate the effect of interpass aging during cold rolling on the microstructure evolution after secondary recrystallization, specimens with and without interpass aging were prepared during cold rolling. The specimen with interpass aging was given four separate treatments for 5 min at 250°C at intermediate thicknesses of 1.5, 0.95, 0.62, and 0.43 mm. Hereafter, these two specimens with and without interpass aging will be referred to as ‘A ’ and ‘B’, respectively. Precipitates play a critical role in inducing AGG. In highly grain-oriented electrical steel, precipitates such as AlN serve as a grain growth inhibitor at low temperature and AGG occurs when precipitates start to dissolve or coarsen at high + Corresponding author, E-mail: nmhwang@snu.ac.kr Materials Transactions, Vol. 53, No. 4 (2012) pp. 658 to 661 © 2012 The Japan Institute of Metals
Transcript of Irregular or Smooth Grain Boundaries Evolved after ... · Keywords: abnormal grain growth,...
Irregular or Smooth Grain Boundaries Evolvedafter Secondary Recyrstallization of Fe3%Si Steel
Hyung-Ki Park1,2, Chang-Soo Park1, Tae-Wook Na1, Chan-Hee Han2 and Nong-Moon Hwang1,+
1Department of Materials Science and Engineering, Seoul National University, Seoul 151-742, Korea2POSCO Technical Research Laboratories, POSCO, Pohang, Korea
Irregular or smooth grain boundaries after secondary recrystallization are formed depending on whether interpass aging is adopted or notduring cold rolling of Fe3%Si steel. Interpass aging induces the primary texture of high intensity of {111}©112ª and {411}©148ª orientations,which have a ∑9 relation with respect to the Goss orientation and therefore are favorable for Goss grains to grow by solid-state wetting. Underthis condition Goss grains come in contact with each other by wetting without leaving any small grains in between, resulting in irregularboundaries. [doi:10.2320/matertrans.M2011268]
(Received August 30, 2011; Accepted December 27, 2011; Published February 15, 2012)
Keywords: abnormal grain growth, secondary recrystallization, grain boundary wetting, interpass aging
1. Introduction
Abnormal grain growth (AGG), which is also referred to assecondary recrystallization, often occurs in polycrystallinemetals. During AGG only a few grains grow abnormallyconsuming the neighboring matrix grains. This phenomenonis particularly famous in Fe3%Si alloy because the core lossof Fe3%Si can be minimized by properly controlling AGG.Since the first report of the selective AGG of Goss grains inFe3%Si alloy by Goss in 1935,1) many efforts have beenmade to understand the mechanism because of its techno-logical importance and scientific interest. In spite of intensiveefforts, however, the underlying principle has not yet beenclarified.
The selective AGG of Goss grains has been approachedmainly based on the high mobility of grain boundariesshared by Goss grains with other matrix grains.24) However,misorientation analyses between Goss grains and matrixgrains do not support this assumption.5,6) On the other hand,Hwang et al.710) has suggested the sub-boundary enhancedsolid-state wetting for selective AGG of Goss grains, wherethe growth advantage of abnormally-growing grains isapproached in terms of grain boundary energy. Accordingto this mechanism, if a grain has sub-boundaries of very lowenergy, the probability of solid-state wetting becomes so highthat the grain containing sub-boundaries undergo selectiveAGG. In relation to this possibility, Park et al.11) observedthat sub-boundaries exist exclusively in abnormally-growingGoss grains using transmission electron microscopy. Also,Dorner et al.12) confirmed the existence of sub-boundariesexclusively within Goss grains after primary recrystallizationof a sheet sample reduced 89% by cold rolling of a singlecrystal of Goss orientation. The solid-state wetting mech-anism could successfully explain many puzzling structuressuch as the formation of island and peninsular grains at thegrowth front7,13) and the pancake-shaped growth of Gossgrains in highly grain-oriented electrical steel of Fe3%Si.14)
Normally smooth grain boundaries are evolved aftersecondary recrystallization of Fe3%Si steel, whereas highly
irregular grain boundaries are evolved if interpass aging isadopted during cold rolling. Interpass aging, which refers toshort time aging treatments at 150300°C between passesof the single stage cold rolling has been reported to improvethe magnetic properties of highly grain-oriented electricalsteel.1517) Interpass aging has been studied extensively butits effect is not clearly understood. Quite different micro-structures are evolved after secondary recrystallization,depending on whether interpass aging is adopted or not:highly irregular grain boundaries are produced with interpassaging whereas smooth grain boundaries are producedwithout interpass aging. The purpose of this work is tostudy why irregular grain boundaries are evolved aftersecondary recrystallization in the Fe3%Si sheet specimensundergoing interpass aging. Rather than studying the bene-ficial effect of the interpass aging, our study is limited to themechanism as to the different morphology evolution of grainboundaries between the two processes with and withoutinterpass aging.
2. Experimental Procedure
An ingot of Fe3%Si, which employed AlN as an inhibitorof grain growth to produce the highly grain-orientedelectrical steel, was used as a starting material. The steelingot was hot rolled to 2.3mm and the thickness was furtherreduced to 0.3mm by one stage cold rolling method, whichis equivalent to a total engineering thickness reduction of¾ = 87%. To investigate the effect of interpass aging duringcold rolling on the microstructure evolution after secondaryrecrystallization, specimens with and without interpass agingwere prepared during cold rolling. The specimen withinterpass aging was given four separate treatments for5min at 250°C at intermediate thicknesses of 1.5, 0.95,0.62, and 0.43mm. Hereafter, these two specimens with andwithout interpass aging will be referred to as ‘A’ and ‘B’,respectively.
Precipitates play a critical role in inducing AGG. In highlygrain-oriented electrical steel, precipitates such as AlN serveas a grain growth inhibitor at low temperature and AGGoccurs when precipitates start to dissolve or coarsen at high+Corresponding author, E-mail: [email protected]
Materials Transactions, Vol. 53, No. 4 (2012) pp. 658 to 661©2012 The Japan Institute of Metals
temperatures. The effect of AlN on secondary recrystalliza-tion would not be changed by such a low temperature heattreatment by interpass aging, and therefore would be thesame in both specimens. The two specimens were annealed at850°C for 150 s for decarburization and primary recrystalli-zation. These specimens underwent a stepwise heat treatmentat 15°C/h up to 1200°C and held for 10 h for secondaryrecrystallization. The sheet specimens were etched with50% water50% HCl at 80°C to identify the grain bounda-ries of abnormally-grown grains.
To investigate the texture after primary recrystallization atthe surface layer, where Goss nuclei for AGG weredistributed, the two specimens after primary recrystallizationwith a thickness of 0.3mm were polished down to a thicknessof 0.26mm. The macrotextures were examined by aconventional X-ray texture goniometer (D8 Advance,Bruker) using Co K¡ radiation. From three incomplete polefigures, the three-dimensional orientation distribution func-tion (ODF) was calculated by the series expansion methodand presented in the orientation space defined by the Eulerangles (¤1, ¯, ¤2).18) To determine the orientation relation-ships of individual grains relative to the Goss orientation,{110}©001ª, the specimens after primary recrystallizationwere analyzed by electron backscattered diffractometer(EBSD) attached to a field-emission scanning electronmicroscope (JSM-6500F, JEOL). EDAX/TSL software wasused to analyze the orientation.
3. Results and Discussion
Figures 1(a) and 1(b) are the microstructures of ‘A’ and‘B’, respectively, after secondary recrystallization and itscorresponding (100) and (110) pole figures are shown on theright inset of each figure. The pole figures indicate thatabnormally-grown grains in both specimens had a Goss
orientation. Figure 1(a) for ‘A’ with interpass aging showshighly irregular grain boundaries whereas Fig. 1(b) for ‘B’without interpass aging shows smooth grain boundaries. Thisdifferent grain boundary morphology would come from thedifferent surroundings of Goss grains in the texture afterprimary recrystallization, which affects the growth rate of theGoss grains during secondary recrystallization. Indeed,Homma and Hutchinson19) reported the effect of the textureafter primary recrystallization on the growth rate of the Gossgrains. Also, in the viewpoint of solid-state wettingmechanism, the primary texture affects the growth rate ofthe Goss grains because it determines the distribution ofthe orientation relation between the Goss grains and matrixgrains.14) Therefore, to clarify the different grain boundarymorphology between the ‘A’ and ‘B’, the textures afterprimary recrystallization were compared by X-ray diffraction.
Figures 2(a) and 2(b) show ODFs of ‘A’ and ‘B’ afterprimary recrystallization, respectively. Mainly evolved ori-entations were the same between the two specimens.However, the texture of ‘B’ was more broadly distributedthan that of ‘A’. In both specimens, {111}©112ª and{411}©148ª orientations developed during primary recrystal-lization. Heavily cold rolled body-centered cubic steel,consisting of ¡-fiber (RD // ©110ª) and £-fiber (ND //©111ª) textures, is known to have a tendency to generate{111}©112ª and {411}©148ª orientations after primaryrecrystallization.20) The stored energy of the grains developedafter cold rolling is maximal in the {111}©112ª orientation,resulting in preferred nucleation of this orientation.21) The{411}©148ª orientation was experimentally shown to nucle-ate in the deformed ¡-fiber texture.20,22) Although thedevelopment of the {411}©148ª texture is well established,its formation mechanism has not been understood clearly.23)
To clarify the difference of the texture in more detail, theODF sections at ¤2 = 45° of the specimen with and without
Fig. 1 The microstructures after secondary recrystallization (a) with and (b) without interpass aging during cold rolling.
Irregular or Smooth Grain Boundaries Evolved after Secondary Recyrstallization of Fe3%Si Steel 659
interpass aging, which include orientations mainly evolved inbody-centered cubic iron, were magnified as shown inFigs. 2(c) and 2(d), respectively. The drastic difference inthe texture intensity between the two specimens is revealed.Compared with the main texture after primary recrystalliza-tion of ‘B’, that of ‘A’ has a significantly high intensity oforientations near {111}©112ª and {411}©148ª, which have a∑9 relationship with respect to the Goss orientation. Indeed,the changes of texture after primary recrystallization byinterpass aging were reported.1517) In relation to this changeof texture, Matsuo et al.15) reported that interpass agingduring cold rolling causes free carbon to pin dislocations,which inhibits a complex slip such as cross-slip, andpromotes the deformation by a single slip. Therefore,interpass aging would affect the deformation behavior duringcold rolling, which might again affect the change of thetexture after primary recyrstallization.
The difference of the texture determines a percentage oflow energy boundaries relative to the Goss orientation, whichaffects the growth rate of Goss grains during secondaryrecrystallization. In the mechanism of secondary recrystalli-zation by solid-state wetting, the high growth rate corre-sponds to the high probability of wetting, which woulddepend on the texture after primary recrystallization.14) For
example, if the texture produces a higher percentage of lowenergy grain boundaries with respect to the Goss orientation,the wetting probability of Goss grains increases.7,10) For this,the percentage of coincidence site lattice (CSL) boundaries,which would be considered as low energy boundaries,24,25)
relative to the Goss orientation was calculated using EBSD.In fact, the crystallographic relationship as well as theboundary plane should be considered to determine theboundary energy.26) However, the data for the boundaryplane are not available; we had to assume that the boundarieswith CSL relationship have low energy.
Figure 3 shows the percentage of CSL boundaries relativeto the Goss orientation in the specimens after primaryrecrystallization. ∑ values were calculated up to ∑29b usingBrandon’s criteria.27) In order to get statistically reliable data,a sufficiently large number of grains were examined: 9105and 8926 grains in the ‘A’ and ‘B’, respectively. The blackand gray colored bars indicate the percentage of the matrixgrains with the CSL relationship in the ‘A’ and ‘B’,respectively. The specimen with interpass aging had a higherpercentage of CSL boundaries, especially low ∑ index (∑3∑9), than that without pass aging. The higher percentage of∑9 boundary can be understood by the primary texture ofthe specimen with interpass aging which had a high intensity
Fig. 2 ODFs of the specimens (a) with and (b) without interpass aging after primary recyrstallization. To clarify the difference of thetexture in more detail, ODF sections at ¤2 = 45° of the specimen (c) with and (d) without interpass aging were magnified.
H.-K. Park, C.-S. Park, T.-W. Na, C.-H. Han and N.-M. Hwang660
of the {111}©112ª and {411}©148ª orientations having ∑9relation with the Goss orientation.
If ∑1∑9 boundaries in Fig. 3 are assumed to have lowenergy enough for the Goss grains to undergo solid-statewetting, the difference in their percentage between ‘A’ and‘B’ is worth noticing. The percentage of ∑1∑9 boundariesin Fig. 3 is 8.62% in ‘A’ and 6.9% in ‘B’, which suggests thatthe wetting probability in ‘A’ should be higher than that in‘B’. This means that the growth rate of Goss grains by solid-state wetting in ‘A’ would be higher than that in ‘B’ due to itshigher percentage of low energy boundaries. Therefore, therate of secondary recrystallization in ‘A’ would be higherthan that in ‘B’.
In the case of ‘A’, an abnormally-growing Goss graincomes in contact with another abnormally-growing Gossgrain by solid-state wetting without leaving any small grainsin between, before the appreciable normal grain growth(NGG) occurs among matrix grains. It should be noted thatthe growth front of Goss grains growing by solid-statewetting is highly irregular as featured by very high frequencyof peninsular grains.7,13) As a result, the grain boundaries of‘A’, whose growth front proceeds by solid-state wetting,would be highly irregular as shown in Fig. 1(a).
In the case of ‘B’, however, the rate of secondaryrecrystallization is relatively low and some NGG of matrixgrains would occur before abnormally-growing Goss grainsmeet each other during secondary recrystallization. WhenNGG occurs, high energy boundaries between the matrixgrains are replaced by low energy boundaries. If the matrixgrain boundaries have predominantly low energy, solid-statewetting is not favored.8) In this condition, the abnormally-growing Goss grains undergo further growth not by solid-state wetting but by the non-wetting mode, where theboundary migrates by the difference in the grain boundarycurvature. Therefore, the microstructure after secondaryrecrystallization is similar to that of NGG with smooth grainboundaries as shown in Fig. 1(b).
4. Conclusion
In conclusion, the Fe3%Si sheet steel, which underwentinterpass aging during cold rolling, shows highly irregulargrain boundaries after secondary recrystallization. This aspectis attributed to the high growth rate of Goss grains by solid-state wetting, which is again attributed to a higher percentageof low energy boundaries relative to the Goss orientationresulted from interpass aging.
Acknowledgements
This work was financially supported by POSCO TechnicalResearch Laboratories.
REFERENCES
1) N. P. Goss: Trans. Am. Soc. Met. 23 (1935) 511544.2) J. Harase, R. Shimizu and D. J. Dingley: Acta Metall. Mater. 39 (1991)
763770.3) N. Rajmohan, J. A. Szpunar and Y. Hayakawa: Acta Mater. 47 (1999)
29993008.4) Y. Ushigami, T. Kumano, T. Haratani, S. Nakamura, S. Takebayashi
and T. Kubota: Mater. Sci. Forum 467470 (2004) 853862.5) A. L. Etter, T. Baudin and R. Pnelle: Scr. Mater. 47 (2002) 725730.6) A. Morawiec: Scr. Mater. 43 (2000) 275278.7) K. J. Ko, P. R. Cha, D. Srolovitz and N. M. Hwang: Acta Mater. 57
(2009) 838845.8) K. J. Ko, A. D. Rollett and N. M. Hwang: Acta Mater. 58 (2010) 4414
4423.9) D. K. Lee, K. J. Ko, B. J. Lee and N. M. Hwang: Scr. Mater. 58 (2008)
683686.10) D. K. Lee, B. J. Lee, K. J. Ko and N. M. Hwang: Mater. Trans. 50
(2009) 25212525.11) H. K. Park, S. D. Kim, S. C. Park, J. T. Park and N. M. Hwang: Scr.
Mater. 62 (2010) 376378.12) D. Dorner, L. Lahn and S. Zaefferer: Mater. Sci. Forum 467470
(2004) 129134.13) S. B. Lee, N. M. Hwang, C. H. Hahn and D. Y. Yoon: Scr. Mater. 39
(1998) 825829.14) H. K. Park, J. H. Kang, C. S. Park, C. H. Han and N. M. Hwang: Mater.
Sci. Eng. A 528 (2011) 32283231.15) M. Matsuo, T. Sakai, M. Tanino, T. Shindo and S. Hayami: 6th Int.
Conf. on Text. of Mater., (1981) pp. 918927.16) T. Haratani, W. N. Hutchinson, I. L. Dillamore and P. Bate: Met. Sci. 18
(1984) 5766.17) J. W. Flowers and W. S. Wright: J. Appl. Phys. 57 (1985) 42174219.18) V. Randle and O. Engler: Introduction to Texture Analysis: Macro-
texture, Microtexture and Orientation Mapping, (Gordon & Breach,Amsterdam, 2000).
19) H. Homma and B. Hutchinson: Acta Mater. 51 (2003) 37953805.20) H. Homma, S. Nakamura and N. Yoshinaga: Mater. Sci. Forum 467
470 (2004) 269274.21) I. Samajdar, B. Verlinden, P. V. Houtte and D. Vanderschueren: Mater.
Sci. Eng. A A238 (1997) 343350.22) T. Toge, M. Muraki, M. Komatsubara and T. Obara: ISIJ Int. 38 (1998)
524530.23) S. Nakamura and H. Homma: Mater. Sci. Forum 467470 (2004) 159
164.24) G. C. Hasson and C. Goux: Scr. Metall. 5 (1971) 889894.25) D. Raabe and K. Lücke: Scr. Metal. Mater. 26 (1992) 12211226.26) M. Z. Quadir, C. S. T. Chang and B. J. Duggan: ISIJ Int. 50 (2010)
264271.27) D. G. Brandon: Acta Mater. 14 (1966) 14791484.
Fig. 3 The percentage of the matrix grains with the CSL relation with theGoss grain calculated using Brandon’s criteria. The black color indicates aCSL percentage of the specimen with interpass aging and the gray oneindicates that without interpass aging.
Irregular or Smooth Grain Boundaries Evolved after Secondary Recyrstallization of Fe3%Si Steel 661
