Quantitative Analysis of Texture Evolution of Direct Chill Cast and Continuous Cast AA 1100 Aluminum Alloys during Cold Rolling

Transcription

1 Materials Transactions, Vol. 48, No. 7 (7) pp to 189 #7 The Japan Institute of Metals Quantitative Analysis of Texture Evolution of Direct Chill Cast and Continuous Cast AA 11 Aluminum Alloys during Cold Rolling Hui Yuan, Jing Li, Dayong Cai, Qingxiang Yang and Wenchang Liu* Key Laboratory of Metastable Materials Science and Technology, College of Materials Science and Engineering (West Campus), Yanshan University, Qinhuangdao 664, P.R. China The hot bands of direct chill cast (DC) and continuous cast (CC) AA 11 aluminum alloys were annealed at 454 C for 3 hours, and then cold rolled to different reductions. The texture of the cold rolled samples was measured by X-ray diffraction in order to compare the texture evolution of the DC and CC aluminum alloys during rolling. The texture volume fractions were calculated by an improved integration method. Mathematical formulae of the texture volume fractions and rolling true strain were established to simulate rolling texture evolution. The results show that the DC AA 11 aluminum alloy exhibits a lower formation rate of the fiber than the CC AA 11 aluminum alloy. [doi:1.3/matertrans.mra73] (Received January 5, 7; Accepted April 1, 7; Published June, 7) Keywords: aluminum, cold rolling, texture, x-ray diffraction, texture volume fraction 1. Introduction The microstructure and texture of aluminum alloy sheets have been extensively investigated for the improvement of their formability. For a given aluminum alloy, the characteristics of texture and microstructure depend strongly on the processing method utilized and the processing variables. Therefore, the determination of the correlation between the parameters describing the characteristics of texture and microstructure and the processing variables is very important to produce high quality aluminum alloy sheets. Twin-belt continuous cast (CC) processing of aluminum alloy sheets has recently gained extensive attention because of its high productivity and low conversion cost. In the CC processing, the molten metal is poured between two rotating steel belts to produce a cast slab, which is immediately fed into three consecutive hot rolling mills, forming hot-band products. Aluminum sheet produced by the CC processing provides an energy saving of at least 5% and an economic saving of more than 14% over sheet made from conventional direct chill (DC) cast ingots. Due to different processing routes, there are distinct differences in microstructure and texture between CC and DC hot bands, which strongly affect the evolution of microstructure and texture during subsequent processing and hence affect the formability of aluminum sheets. 1 4) Liu et al. 1) studied the recrystallization behavior of cold rolled DC and CC AA 34 aluminum alloys. They found that it was difficult to obtain a strong cube texture in annealed CC hot bands. The recrystallization occurred more easily in cold rolled DC hot bands than in cold rolled CC hot bands. Similar results were also obtained in DC and CC AA 518 aluminum alloys.,3) In the present work the texture evolution of DC and CC AA 11 aluminum alloys during cold rolling was investigated by X-ray diffraction. The texture volume fractions of cold rolled samples were calculated by an improved integration method. The effect of the processing method (DC vs. CC) on the texture evolution during cold rolling was determined. *Corresponding author, wcliu@ysu.edu.cn. Experimental The materials used in the present investigation were CC and DC AA 11 aluminum alloys. The as-received materials were commercially produced CC and DC hot bands. The thickness of the CC and DC hot bands was 3.6 and.53 mm, respectively. The DC and CC hot bands were annealed at 454 C for 3 hours followed by air cooling. In order to investigate the effect of the processing method, the microstructure and texture of the DC and CC hot bands before and after annealing at 454 C were characterized. The annealed hot bands were then cold rolled to different reductions ranging from to 9% on a laboratory rolling mill with rolls of 13 mm in diameter. Oil was used as a lubricant. The samples were rolled by alternating the top and bottom sides between passes during multipass rolling. The rolling thickness change per pass was about.1 mm. Texture measurements were performed at one fourththickness of the cold rolled sheets. The (111), (), and () pole figures were measured up to a maximum tilt angle of 75 by the Schulz back-reflection method using CuK radiation. The orientation distribution functions (ODFs) were calculated from the incomplete pole figures using the series expansion method (l max ¼ 16). 5) The ODFs were presented as plots of constant sections with isointensity contours in Euler space defined by the Euler angles 1,, and. The volume fractions of texture components were calculated by an improved integration method. 3,6) In this method, Euler space representing all possible crystallographic orientations in rolling is subdivided into 6 regions called the cube, Goss, rotated cube (), rotated Goss (r-goss), brass and copper orientation regions. 7) The volume fractions of the cube, Goss,, and r-goss components are calculated by integration over the cube, Goss, and r- Goss orientation regions, respectively. The fiber runs from the B orientation f11gh11i through the S f13gh634i orientation to the C orientation f11gh111i. 8 1) It lies in the brass and copper orientation regions, but does not occupy the whole region. The orientation intensities f (g) within 15.5 of the center position of the fiber in the subset of Euler space

2 Quantitative Analysis of Texture Evolution of Direct Chill Cast and Continuous Cast AA 11 Aluminum Alloys during Cold Rolling 1887 Fig. 1 Microstructure of the DC and CC AA 11 aluminum alloy before and after annealing at 454 C for 3 h. DC hot band, CC hot band, (c) annealed DC hot band, and (d) annealed CC hot band. are integrated to represent the volume fraction of the fiber component. The orientations in the remaining brass and copper regions are referred to as orientations. 3. Results 3.1 Microstructure and texture of DC and CC AA 11 aluminum alloys Figure 1 shows the microstructure of the DC and CC AA 11 aluminum alloys before and after annealing at 454 C for 3 hours. All micrographs were taken from longitudinal sections as defined by the rolling direction and the normal direction. It is seen that the DC and CC hot bands possessed a typical deformation structure. After annealing at 454 C, the DC and CC hot bands were fully recrystallized, and the size of recrystallized grains in the DC aluminum alloy was smaller than that in the CC aluminum alloy. Figure shows the texture of the DC and CC AA 11 aluminum alloys before and after annealing at 454 C for 3 hours. As shown in Figs. and, the DC and CC hot bands possessed a typical fiber rolling texture. The fiber ran from the B orientation through the S orientation to the C orientation. 8 1) After the hot bands were annealed at 454 C for 3 hours, the texture of DC and CC hot bands was changed from the fiber rolling texture to the cube recrystallization texture. The DC aluminum alloy exhibited a stronger cube recrystallization texture than the CC aluminum alloy, which is in agreement with the results obtained in DC and CC AA 34 1) and 518 3) aluminum alloys. However, the difference in strength of cube recrystallization texture between the DC and CC AA 11 aluminum alloys was less than that between the DC and CC AA 518 aluminum alloys. (c) Fig. Texture of the DC and CC AA 11 aluminum alloy before and after annealing at 454 C for 3 h. DC hot band, CC hot band, (c) annealed DC hot band, and (d) annealed CC hot band. 3. Texture evolution of DC and CC AA 11 aluminum alloys during rolling Figures 3 and 4 show the texture evolution of the DC and CC AA 11 aluminum alloys during rolling, respectively. During cold rolling all initial orientations were rotated to the stable end orientations, i.e. the fiber component. As the cold rolling reduction increased, the intensity of the cube orientation decreased, whereas the intensities of orientations along the fiber increased. After 9% cold rolling, the (d)

3 1888 H. Yuan, J. Li, D. Cai, Q. Yang and W. Liu f (g) 1 C S B f (g) C S B % 38.4% ϕ, o ϕ, o Fig. 5 Intensities of the ODF f (g) at the center position of the fiber as a function of a particular angle for the DC and CC AA 11 aluminum alloys cold rolled to different reductions. (c) 7.% (d) 91.5% Fig. 3 ODFs of the DC AA 11 aluminum alloy cold rolled to reductions of 13., 38.4, (c) 7., and (d) 91.5%. 15.5% 4.% (c) 73.4% (d) 91.8% Fig. 4 ODFs of the CC AA 11 aluminum alloy cold rolled to reductions of 15.5, 4., (c) 73.4, and (d) 91.8%. texture was mainly composed of the fiber component. Figure 5 shows the maximum intensities of the ODF f (g) for a particular angle along the fiber for the cold rolled DC and CC AA 11 aluminum alloys. It is seen that the increase in the intensities of the C and S orientations with reduction was higher than that of the B orientation, especially in the DC AA 11 aluminum alloy. This was different from the results found in the CC AA 55 aluminum alloy, in which the intensities of the B and S orientations were higher than the intensity of the C orientation. 1) The lattice rotation of the cube orientation to the fiber during rolling has been investigated. 1) The cube orientation was rotated to the fiber through different lattice rotation paths. The large increase in the intensities of the C and S orientations with reduction indicates that more cube orientations are rotated to the C and S orientations. The concept of texture volume fraction has proved useful in the assessment of the orientation distribution of polycrystalline samples. 8,9,11,1) Table 1 lists the volume fractions of the cube,, Goss, r-goss, fiber and components. The texture volume fractions of the hot bands are also listed in Table 1. For the DC and CC AA 11 aluminum alloys, the volume fractions of the cube,, Goss, r-goss, fiber and components presented similar characteristics with reduction. As the cold rolled reduction increased, the volume fractions of the cube, and components decreased, whereas the volume fraction of the fiber component increased. The volume fraction of the Goss component first increased with increasing reduction. When the cube component was exhausted, the volume fraction of the Goss component decreased with increasing further reduction. 3.3 Quantitative analysis of rolling texture evolution The texture evolution of f.c.c. metals during rolling can be described by the variation in the texture volume fractions with rolling true strain. During rolling the orientations near the cube orientation in the orientation region first rotated to the cube component, and then rotated to the fiber. Therefore, the volume fraction of the component and the sum of the volume fractions of the cube and components decreased with increasing rolling true strain. The volume fraction of the component also decreased with increasing rolling true strain. The relative change of these texture volume fractions can be expressed by f i ¼ðM i M i Þ=M i ð1þ where M i and M i are the texture volume fractions before and after cold rolling, respectively. The volume fraction of the fiber component increases with increasing rolling true strain. The relative change of the volume fraction of the fiber component can be expressed by f ¼ðM M Þ=ð1 M Þ ðþ where M and M are the volume fractions of the fiber component before and after cold rolling, respectively. Liu et al. 3,6) have studied the texture evolution of cold rolled CC AA 5xxx series aluminum alloys. They found that the relation-

4 Quantitative Analysis of Texture Evolution of Direct Chill Cast and Continuous Cast AA 11 Aluminum Alloys during Cold Rolling 1889 Table 1 Texture volume fractions for the cold rolled DC and CC AA 11 aluminum alloys. Alloy Reduction (%) Texture volume fraction (%) cube Goss r-goss fiber Remainder Hot band DC Hot band CC Table Values of k i and n i in Equation (3) for the DC and CC AA 11 aluminum alloys. Alloy Texture component M i (%) k i n i r cube (:61 :69) :85 :6.988 DC 11 fiber (:15 :18) 1:11 : (:68 :88) 1:9 : (:14 :) 1:9 :.9 cube (:59 :65) 1: :5.994 CC 11 fiber (: :4) :98 : (:97 1:13) 1:13 : (:3 :34) :91 :6.989 ln[-ln(1-f i )] cube + β fibre remanider ln[-ln(1-f i )] cube + β fibre Volume fraction, % cube β fiber cube + Volume fraction, % cube β cube ln() ln() Fig. 6 ln½ lnð1 f ÞŠ vs. ln " for the DC and CC AA 11 aluminum alloys. ship between f i and rolling true strain (") takes the form of the Johnson-Mehl-Avrami-Kolmogorov (JMAK) equation: f i ¼ 1 expð k i " n i Þ ð3þ where k i is a constant, and n i is the strain exponent for different texture components. The rolling true strain (") was calculated from the rolling reduction (R) by" ¼ lnð1 RÞ. For the cube +, fiber, and components, the f i values and rolling true strain can be presented in a ln½ lnð1 f i ÞŠ vs. ln " format, as shown in Fig. 6. The values of k i and n i were determined by fitting the Fig. 7 Plots of texture volume fractions as a function of rolling true strain for the DC and CC AA 11 aluminum alloys. Points with different symbols are measured values, and solid lines are calculated by Equation (3). experimental data into eq. (3). Table shows the values of k i and n i as well as the correlation coefficients (r) of these linear fits. Figure 7 shows the volume fractions calculated based on the values of M i, k i and n i. The experimental data are also depicted in Fig. 7. It is found that the above JMAK equation could reflect the variation in the volume fractions of the cube +, fiber, and components with rolling true strain. Thus, the texture evolution of AA 11 aluminum alloy during rolling can be simulated based on the values of M i, k i and n i.

5 189 H. Yuan, J. Li, D. Cai, Q. Yang and W. Liu f i Discussion DC CC cube + β fiber Fig. 8 Effect of the processing method (DC vs. CC) on the texture evolution of AA 11 aluminum alloy during cold rolling. The experimental results reveal that the processing method affects the microstructure and texture of AA 11 aluminum alloy. Since the DC hot band is expected to experience larger plastic deformations and higher degrees of recovery than the CC hot band, the DC hot band exhibits finer deformed microstructure and stronger fiber rolling texture than the CC hot band. After annealing at 454 C for 3 hours, the microstructure and texture of the DC hot band is different from those of the CC hot band. The DC hot band exhibits finer recrystallized grains and stronger cube recrystallization texture than the CC hot band. The processing method affects not only the microstructure and texture of the hot bands, but also the texture evolution during cold rolling. The effect of the processing method on texture evolution can be revealed by the k i and n i values in these JMAK-type equations. The k i value reflects the rate of formation and disappearance of each texture component. Figure 8 shows the relationship between f i and rolling true strain for different components. It is seen that the DC alloy has a lower formation rate of the fiber, a higher rate of lattice rotation in the cube + orientation region, and a lower rate of disappearance of the component than the CC alloy. The difference in the texture evolution between DC and CC alloys is mainly attributed to their initial textures. The DC alloy possesses a stronger cube recrystallization texture than its CC counterpart. The stronger cube recrystallization texture prior to cold rolling decreases the formation rate of the fiber rolling texture. 13) 5. Summary The texture evolution of the DC and CC AA 11 aluminum alloys during rolling was investigated by X-ray diffraction. The results are summarized as follows: (1) The DC hot band possesses finer deformed microstructure and stronger fiber rolling texture than the CC hot band. After annealing at 454 C for 3 hours, the DC hot band exhibits finer recrystallized grains and stronger cube recrystallization texture than the CC hot band. () The relationship between the texture volume fractions and rolling true strain can be quantified by using equations of the JMAK type. The k i value in these equations reflects the rate of formation or disappearance of each texture component. (3) The processing method affects the texture evolution during rolling. The DC alloy has a lower formation rate of the fiber, a higher rate of lattice rotation in the cube + orientation region, and a lower rate of disappearance of the component than the CC alloy. REFERENCES 1) Y. S. Liu, X. M. Chen, J. T. Liu and J. G. Morris: Mater. Sci. Forum () ) Y. L. Liu, S. Ding and J. G. Morris: Aluminum Alloys for Packaging III, ed. S. K. Das (Warrendale, Pennsylvania: TMS, 1998) pp ) W. C. Liu and J. G. Morris: Mater. Sci. Eng. A 339 (3) ) Y. M. Zhao, W. C. Liu and J. G. Morris: Metall. Trans. A 35A (4) ) H. J. Bunge: Texture Analysis in Materials Science, (Butterworths, London, 198). 6) W. C. Liu and J. G. Morris: Metall. Trans. A 35A (4) ) W. C. Liu, D. Juul Jensen and J. G. Morris: Acta metall. 49 (1) ) J. Hirsch and K. Lücke: Acta metall. mater. 36 (1988) ) O. Daaland and E. Nes: Acta metall. 44 (1996) ) W. C. Liu, C.-S. Man and J. G. Morris: Scripta mater. 45 (1) ) K. Lücke, J. Pospiech, J. Jura and J. Hirsch: Z. Metallk. 77 (1986) ) M. B. Cortie: Textures and Microstructures 9 (1997) ) W. C. Liu, T. Zhai, C.-S. Man, B. Radhakrishnan and J. G. Morris: Phil. Mag. 84 (4)