Materials Sciences and Applicatio n, 2011, 2, 911-916
doi:10.4236/msa.2011.27121 Published Online July 2011 (http://www.SciRP.org/journal/msa)
Copyright © 2011 SciRes. MSA
911
EBSD Analysis of the Submicron Width Fibber
Shaped Grain Copper Fabricated by Drawing
Masafumi Matsushita1*, Tomoya Kuji2, Hiromitsu Kuroda2, Seigi Aoyama2, Hiroaki Ohfuji3
1Department of Mechanical Engineering, Ehime University, Bunkyocho, Matsuyama, Japan; 2Metallic Materials Research Depart-
ment, Hitachi Cable. Ltd., Kawajiri, Hitachi, Japan; 3Geodynamics Research Center, Ehime University, Bunkyocho, Matsuyama,
Japan.
Email: *matsushita@eng.ehime-u.ac.jp
Received February 16th, 2011; revised March 21st, 2011; accepted May 5th, 2011.
ABSTRACT
Several drawing processes of 3N pure copper (Cu) with ultra high reduction in area have been performed, and the tex-
ture has been observed using electron back scattered diffraction. It is well known that the texture of drawn Cu is closely
related to its mechanical properties; in particular submicron scale width fibber shaped grain is interesting. Previously,
it was reported that drawing 3N Cu with around 95% reduction in area changes its mechanical properties. In the pre-
sent experiment, grains have been changed from round to fiber-shaped, and subsequently, submicron-width fiber grains
were generated with a 94.546% reduction in area. However, above 94.546% reduction in area, dynamic recrystalliza-
tion occurred and then, prevented the grains to be finer. Further recrystallized grains influence on the macroscopic
mechanical properties of Cu wire. Furthermore, the distribution of the misorientation angle at grain boundaries in-
creased with an increase in the degree of reduction in area; in particular, the distribution drastically increased between
94.546% and 99.997%. Further, the frequency of coincidence of site lattice boundaries increased. The abovementioned
variations in texture are closely related to mechanical properties.
Keywords: Drawing, Grain Boundary, EBSD, Cu, Fibber Shape Grain
1. Introduction
The mechanical properties of metals and alloys are
greatly influenced by their grain size, shape, and misori-
entation angle at grain boundaries (GBs). It is well
known that the effect of grain size on tensile strength
follows the Hall-Petch equation,
σY = σ0 + kd–1/2, (1)
where σY denotes yield strength, d denotes grain diameter,
and σ0 and k are material constants [1].
Furthermore, regarding the fiber grain shape, the ten-
sile strength of the grain in the oblong direction increases
with a decrease in the width of the fiber, obeying the
following Hall-Petch type equation,
σY = σ0 + kW–1/2, (2)
where W is the fiber grain width [2].
However, the Hall-Petch equation is not valid in nano-
meter scale grains [3,4]. The material having macro-
scopic grain, the yield strength increases obeying
Hall-Petch equation. On the other hand, grain being
submicron size, yield strength is obeying Hall-Petch
equation, inter-granular fracture and local shear deforma-
tion increase [3]. Further decrease of grain size, yield
strength is not obeying Hall-Petch equation and many
interesting phenomenon appear, for example, nanocrys-
talline Cu of the diameter 20 nm, which is made by elec-
tro deposition and rolling, shows super plasticity [5].
However the relation between fibber width of fibber
shape grain and mechanical behavior does not clear.
Further the behavior of dynamic strength, specifically
fatigue strength, is not clear.
In recent times, bulk metals having nanometer scale
grains, the so-called bulk nano-metals, are considered as
potential next-generation structured materials [6]. Ac-
cordingly, several methods for the fabrication of bulk
nano-metals have been proposed, although, most of these
methods are not suitable for fabrication on an industrial
scale [7-10]. The drawing process is one of the severe
plastic deformation processes and can be used for indus-
trial fabrication. In particular, the drawing of Cu is a
common method for fabricating many industrial products
such as trolley wire and magnet wire. Meanwhile, the
EBSD Analysis of the Submicron Width Fibber Shaped Grain Copper Fabricated by Drawing
912
demands for Cu wire with specific mechanical properties
have increased with the recent developments of the elec-
tric car, industrial robots, and so on.
Severe drawing deformation will make submicron
width fibber shape grain. Previous studies reported the
relation between the mechanical properties and the re-
duction in area by drawing. Hardness and tensile strength
increase in direct proportion to the reduction in area just
below 95% (hereafter we call this range as stage 1). At
around 95% reduction, the tensile strength and hardness
show a drastic increase (hereafter we call this range as
stage 2). However furthermore increase in the area re-
duction, the tensile strength and hardness become con-
stant (hereafter we call this range as stage 3). The twist
fatigue strength test was performed for an effective
length of 65 mm, a diameter of 2.8 mm, and at the rate of
1000 times/min. Twist fatigue strength also increases in
direct proportion to the reduction in area just below 95%,
namely in stage1. And then at around 95% reduction,
namely in stage 2, twist fatigue strength shows a drastic
increase. However, as the area reduces further, namely in
stage 3, twist fatigue strength decreases while the tensile
strength and hardness is constant [11,12]. This discon-
tinuous behavior of hardness, tensile and twist fatigue
strengths depended on reduction area and mismatch of
the variation among hardness, tensile and twist fatigue
strengths are so interesting. It is considered that above-
mentioned variations of mechanical properties depending
on the reduction in area are closely relation to the texture
variations, but the textures, such as grain size, grain
boundary characters have not been clear.
With the recent developments in electron back scat-
tered diffraction (EBSD), we can determine the misori-
entation angle information of GBs and the crystal orien-
tation of the local area, as the result, grain boundary en-
gineering have been developed [13]. The textures of
drawn metal with ultra high reduction in area have not
been investigated by EBSD thus far. In this study, we
have performed EBSD measurements of the longitudinal
dimensions of drawn Cu wires corresponding to the
abovementioned three stages in order to demonstrate the
textures of drawn metal, such as the grain size, shape,
preferred orientation of the grains and GB misorientation
angle. Furthermore, the relationships among these fea-
tures and the mechanical properties of the material have
been discussed.
2. Experimental Method
2.1. Sample Preparation
The samples were fabricated by the drawing of 99.9% -
pure Cu with various reductions in area, as listed in Ta-
ble 1 (hereafter, these sample are referred to by the
Table 1. A list of reduction in area of each sample.
Sample No. Reduction in area (%)
1 73.450
2 94.546
3 99.997
assigned numbers). The fabrication process involved
performing the drawing process several times. After the
copper was drawn once, the area was reduced by 15%.
For the sake of convenience in the EBSD experiments,
the final diameters of all the samples were kept at 42 μm.
The samples were embedded in resin, and then their lon-
gitudinal cross sections were polished using diamond
powder and colloidal silica with diameters of 1 μm, 50
and 20 nm.
2.2. Electron Backscattered Diffraction
Measurement Method
The longitudinal cross sections of each sample were then
observed through EBSD. The measurement area was
positioned halfway from the central axis, as shown in
Figure 1, since the texture of the drawn Cu wire samples
was symmetric around the central axis. The measurement
step of the EBSD was 100 nm, and the accelerated volt-
age of the electron beam was 15 kV. Channel 5 (HKL
Technology, Co.) was used for analysis. The minimum
misorientation angle defined in the present experiment
was 1˚.
In this paper, we have classified the GB into the fol-
lowing three types: 1) coincidence site lattice (CSL)
boundaries, having higher geometric coordination; 2)
low-angle grain boundaries (LGBs), which have a misori-
entation angle of <15˚; and 3) high-angle grain boundaries
(HGBs), which have a misorientation angle of >15˚ but are
not CSL grain boundaries. HGBs do not have specific
rules and include some atomic voids [14,15].
Figure 1. Observed longitudinal area of Cu wire.
Copyright © 2011 SciRes. MSA
EBSD Analysis of the Submicron Width Fibber Shaped Grain Copper Fabricated by Drawing913
3. Results and Discussions
3.1. Grain shape and Grain Boundary
Characteristics
Grain boundary maps of each sample listed in Table 1
are shown in Figure 2. The black, red, and thin gray
lines are HGBs, Σ3 boundaries, and LGBs, respectively.
The Σ3 boundary has the highest geometric coordination,
namely, twin boundary.
The grain shape changes from a round to a fiber shape
when the area is reduced, as shown in Figures 2(a) and
(b). The fiber width becomes several hundred nm and the
aspect ratio is larger than 5, as shown in Figure 2(b).
Deformation twinning by room temperature plastic de-
formation has not been reported previously. However,
present EBSD measurement results clearly show the ex-
istence of deformation twin. It is considered that this
elongation of the nano-fiber grain along the drawing di-
rection is mainly responsible in increasing the tensile and
twists fatigue strengths.
(a)
(b)
(c)
Figure 2. Grain boundary maps of samples. (a), (b) and (c)
are Sample 1, 2 and 3, respectively. Thin gray lines show
LGBs. Red and black lines show Σ3 CSL boundaries and
HGBs, respectively. (a) Sample 1 (reduction in area
73.450%); (b) Sample 2 (reduction in area 94.546%); (c)
Sample 3 (reduction in area 99.997%).
With further reduction in area, round grains appear and
some of them are oblong to the central axis, as shown in
Figure 2(c). These grains are considered as recrystallized
grains. Main reason of recrystallization is the strain en-
ergy generated by deformation. Moreover, there is possi-
bility that solute atoms of Cu are concentrated in HGBs
by severe plastic deformation, and then a part of the grain
becomes the starting point for dynamic recrystallization.
These dynamic recrystallizations prohibit the grain be-
coming finer.
The frequency of CSL boundaries is shown in Figure
3. CSL boundaries increase with an increase in the re-
duction in area, particularly in sample 3. However, the
relative frequency for the misorientation angle of the
CSL boundary is much lower than Cu of similar area
recrystallized grain obtained equal channel angular ex-
trusion [10]. The relative frequencies for the misorienta-
tion angle of each sample are shown in Figure 4. The
relative frequency of HGBs drastically increases as the
percentage reduction in area grows from 94.546% to
99.9997%. GBs with misorientation angles of <5˚ de-
crease drastically. The recrystallized grains effect on the
macroscopic mechanical behavior of materials. Further-
more, an increase in the relative frequency of HGBs
causes inter-granular fracture by impact [16]. Recrystal-
lized grain and the increase of HGBs influence tensile
and twist fatigue strengths. In particular, twist fatigue
strengths is changed between stage 2 and 3, which cor-
respond sample 2 and 3. It means intergranular fracture
occurs at HGBs is one of reasons of this change. Further,
Copyright © 2011 SciRes. MSA
EBSD Analysis of the Submicron Width Fibber Shaped Grain Copper Fabricated by Drawing
914
Figure 3. Frequency of CSL boundaries at each sample.
grain boundary sliding would occur in stage 3 as same in
case of non-fibber submicron grain [4].
3.2. Preferred Orientations
Inverse pole figures (IPFs) of the outer and central area
of each sample are shown in Figure 5.
In sample 1, very weak <100> + <111> double orien-
tation textures are confirmed in the drawing direction,
namely, the X direction; these textures are stronger in the
outer area. In the Y and Z directions, a clear orientation
direction cannot be seen; that is, the direction of the ra-
dius has a random orientation.
In sample 2, the clear preferred orientation is con-
firmed in the X direction. In the central area, a <100>
axis orientation is seen. Furthermore, a <100> weak ori-
entation is confirmed in the Y and Z directions. These
results indicate that a cube texture was formed. In the
outer area, <100> + <111> double axis orientations are
seen in the X direction. The <100> orientation is weak
compared to the center area. Further rotation from <100>
to <101> occurs in the Z direction. Considering these
results, a (100) [100] cube orientation is found to col-
lapse by shear stress.
IPFs of sample 3 are shown in Figure 5. Very strong
orientations are confirmed in the X direction. In the outer
area, a <111> axis orientation is clearly confirmed. In the
central area, a <100> rich <100> + <111> double axis
orientation is confirmed. The radius direction is in both
Y and Z directions, and a weak <101> axis orientation is
clearly seen in both the outer and center areas. Consider-
ing these results, a <111> axis orientation is generated,
and the cube orientation shown in sample 2 disappears. A
strong <111> axis orientation in the X direction would
affect the mechanical properties since the slip direction is
vertical to the drawing direction.
The preferred orientation of recrystallized Cu wire by
annealing, i.e. , before drawing the sample, has a <001>
axis orientation in the X direction and a <101> axis ori-
(a)
(b)
(c)
Figure 4. Relative frequencies for the misorientation angle
of each sample. (a), (b) and (c) are Sample 1, 2 and 3, re-
spectively. (a) Sample 1 (reduction in area 73.450%); (b)
Sample 2 (reduction in area 94.546%); (c) Sample 3 (reduc-
tion in area 99.997%).
entation in the Y and Z directions.
Brown discussed the relationship between the orienta-
tion of the fiber texture rotation and the stacking fault
Copyright © 2011 SciRes. MSA
EBSD Analysis of the Submicron Width Fibber Shaped Grain Copper Fabricated by Drawing915
(a)
(b)
(c)
Figure 5. Inverse pole figures of the outer and central areas
of each sample. (a), (b) and (c) are Sample 1, 2 and 3, re-
spectively. (a) Sample 1 (reduction in area 73.450%); (b)
Sample 2 (reduction in area 94.546%); (c) Sample 3 (reduc-
tion in area 99.997%).
energy (SFE) [17]. According to his theory, a high SFE
metal causes cross slip more easily than a low SFE metal,
and subsequently, the cross slip leads to the rotation of
the <100> axis orientation to <111>. The SFE of Al is
higher than that of Cu. The present results compared to
past experimental results for Al [18], obtained through
X-ray diffraction, show that the <100> axis orientation of
Al rotates to <111> in lower than that of Cu. This result
corresponds to ref. [17], and cross slips are considered to
be the origin of this change.
During the drawing process, a two-axis stress system,
corresponding to compression and tensile stress, is added
to the central area. On the other hand, in the outer area, a
strong shear stress parallel to the drawing dice is added
to the two-axis stress of the central area. Inakazu has
analyzed the orientation change caused by the above-
mentioned stress systems and proved the stability of the
<100> and <111> anises [19]. However, in [19], orienta-
tion change from <001> to <112> and then <111> is
predicted with an increase in the reduction in area. Fur-
thermore, these changes are confirmed in Al experimen-
tally. In the present experiment for Cu, the data almost
corresponds to the results of a previous experiment for Al;
however, the X-axis orientation does not change via
<112>.
4. Conclusions
Grains of Cu were changed to submicron-width fiber
grains with ultra high reduction in area. However, when
the degree of reduction in area was increased for 94.546
and 99.997%, dynamic recrystallization occurred. Re-
crystallizations prevent the grain to be finer and influ-
ence on the macroscopic mechanical properties, which
mentioned in ref. [11,12].
The distributions of the relative frequency of HGBs
increase with an increase in the degree of reduction in
area; this increase in relative frequency distribution is
drastic between 94.546 and 99.997%. Furthermore, the
frequency of CSL boundaries increases. These changes
in textures would cause a discontinuous change in me-
chanical properties reported in ref. [11] and [12]. Con-
sidering from the present grain and grain boundaries
character changes, intergranular fracture and grain
boundary slip would cause the remarked change of twist
fatigue fracture.
Preferred orientations of drawing and radius directions
change to the <111> axis. These changes occur for
higher reductions in area for Cu rather than Al, because
the SFE of Al is smaller than that of Cu.
REFERENCES
[1] E. O. Hall, “Yield Point Phenomena in Metals and Al-
Copyright © 2011 SciRes. MSA
EBSD Analysis of the Submicron Width Fibber Shaped Grain Copper Fabricated by Drawing
Copyright © 2011 SciRes. MSA
916
loys,” Plenum Press, New York, 1970.
[2] Y. Sakai, “Strength of Heavily Cold Worked Two-Phase-
Copper Alloy,” Materia Japan, Vol. 36, 1997, pp. 692-
696.
[3] T. G. Nieh and J. Wadsworth, “Hall-Petch Relation in
Nanocrystalline Solid,” Scripta Metallurgica et Materi-
alia, Vol. 25, No. 4, 1991, pp. 955-958.
doi:10.1016/0956-716X(91)90256-Z
[4] P. G. Sanders, J. A. Eastman and J. R. Weertman, “Elas-
tic and Tensile Behavior of Nanocrystalline Copper and
Palladium,” Acta Materialia, Vol. 45, No. 10, 1997, pp.
4019-4025. doi:10.1016/S1359-6454(97)00092-X
[5] L. Lu, M. L. Sui and K. Lu, “Superplastic Extensibility of
Nanocrystalline Copper at Room Temperature,” Science,
Vol. 287, No. 5457, 2000, pp. 1463-1466.
doi:10.1126/science.287.5457.1463
[6] L. L. Shaw, “Processing Nanostructured Materials: An
Overview,” JOM Journal of the Minerals, Metals and
Materials Society, Vol. 52, No. 12, 2000, pp. 41-45.
doi:10.1007/s11837-000-0068-2
[7] Y. Saito, H. Utsunomiya, N. Tsuji and T. Sakai, “Novel
ultra-High Straining Process for Bulk Materials - Devel-
opment of the Accumulative Roll-Bonding (ARB) Proc-
ess,” Acta Materialia, Vol. 47, No. 2, 1999, pp. 579-583.
doi:10.1016/S1359-6454(98)00365-6
[8] Z. Horita, D. J. Smith, M. Furukawa, M. Nemoto, R. Z.
Valiv and T. G. Langdon, “An Investigation of Grain
Boundaries in Submicrometer-Grained Al-Mg Solid So-
lution Alloys Using High—Resolution Electron Micro-
scope,” Journal of Materials Research, Vol. 11, No. 8,
1996, pp. 1880-1890. doi:10.1557/JMR.1996.0239
[9] A. Belyakov, W. Gao, H. Miura and T. Sakai, “Strain-
Induced Grain Evolution in Polycrystalline Copper during
Warm Deformation,” Metallurgical and Materials Tran-
sactions A, Vol. 29, No. 12, 1998, pp. 2957-2965.
doi:10.1007/s11661-998-0203-1
[10] G. Wang, S. D. Wu, L. Zuo, C. Esling, Z. G. Wang and G.
Y. Li, “Microstructure, texture, Grain Boundaries in Re-
crystallization Regions in Pure Cu ECAE Samples,” Ma-
terials Science and Engineering A, Vol. 346, No. 1-2,
2003, pp. 83-90. doi:10.1016/S0921-5093(02)00521-X
[11] H. Yamamoto and N. Inakazu, “On the Relation between
Drawing Ratio and Mechanical Properties for Copper
Wire,” Journal of the Japan Copper and Brass Research
Association, Vol. 11, 1974, pp. 133-141.
[12] N. Inakazu and H. Kawakami, “Drawing,” In: H. Tanaka,
Ed., The Japan Society of Technology of Plasticity, Co-
rona Publishing Co. Ltd., Tokyo, 1994, pp. 80-86.
[13] V. Randle, “Grain Boundary Engineering: An Overview
after 25 Years,” Materials Science and Technology, Vol.
36, No. 3, 2010, pp. 253-261.
doi:10.1179/026708309X12601952777747
[14] L. N. Brewer, M. A. Othon, L. M. Young and T. M. An-
geliu, “Misorientation mapping for Visualization of Plas-
tic Deformation via Electron Back-Scattered Diffraction,”
Microscopy and Microanalysis, Vol. 12, No. 1, 2006, pp.
85-91. doi:10.1017/S1431927606060120
[15] M. A. Arafin and J. A. Szpunar, “Modeling of Grain
Boundary Character Reconstruction and Predicting Inter-
granular Fracture Susceptibility of Texture and Random
Polycrystalline,” Computional Matterials Science, Vol.
50, No. 2, 2010, pp. 656-665.
doi:10.1016/j.commatsci.2010.09.031
[16] T. Watanabe, “The Impact of Grain Boundary Character
Distribution on Fracture in Polycrystals,” Material Sci-
ence and Engineering A, Vol. 176, No. 1-2, 1994, pp.
39-49. doi:10.1016/0921-5093(94)90957-1
[17] N. Brown, “The Dependence of Wire Texture in FCC
Metals on Stacking Fault Energy,” Transactions of Met-
allization Society AIME, Vol. 221, 1969, pp. 236-238.
[18] N. Inakazu and H. Yamamoto, “A Study on the Process
of Fiber Texture Formation in Aluminium,” The Japan
Institution of Metals, Vol. 37, 1973, pp. 1224-1229.
[19] N. Inakazu, “Metal Drawing,” Kindai Hensyu Publishing
Co. Ltd., Tokyo, 1985, p. 157.