EQUAL CHANNEL ANGULAR PRESSING OF ALUMINIUM

ALLOYS AND COMPOSITES – AN OVERVIEW

M. Saravanan, R.M.Pillai*, B.C.Pai

Materials and Minerals Division, Regional Research Laboratory (CSIR), Thiruvananthapuram 695 019.
*Corresponding author: E-mail: [email protected]

Abstract

Equal channel angular pressing (ECAP) is a novel technique for producing ultra
fine grain structures in submicron level without change in the billet shape or dimensions
by introducing a large amount of shear strain into the materials. This process is well
suitable for aluminium alloys and is capable of producing ultra fine grain structures with
grain sizes falling between 200 to 500nm. This paper overviews the principle of equal
channel angular pressing technique and influences of its various processing parameters
on the structure and mechanical properties in aluminium alloys and composites.

Keywords: ECAP, Fine grain materials, Aluminium alloys, Severe plastic deformation,
Aluminium matrix composites

1. Introduction:

The automobile and aerospace industries have much interest in developing new
materials with high strength and low weight. Grain refinement is one of the techniques
which provide ultrahigh strength and ductility combination demanded for ambient and
cryogenic temperature applications. In aluminium based alloys, however, it is generally
difficult to reduce the grain size below ~10m through the conventional casting process
and adding grain refiner. On the other hand, severe plastic deformation is an effective
method for producing ultra fine grain materials. Different techniques are used to
introduce large amount of strain into metals. Among these, Equal channel angular
pressing developed by Segal, is a metal working process capable of producing ultra fine
grain materials by means of intense plastic straining without change in the shape or
dimensions of work materials [1-3].

Equal channel angular pressing has the following advantages

 Capability of producing ultra fine grain materials without change in the

shape or dimensions of work materials.
 Large sections of materials can be uniformly and intensively worked.
 Improved strength of the materials without any change in their chemical
composition.

1
  Capability of producing large bulk samples which are free from any residual
porosity.
 A high level of deformation can be achieved after one pass.
 A variety of different microstructures and textures can be developed by
modifying planes and direction of shear at each pass.

However, it has some disadvantages/difficulties towards its commercial viability.

 Shorter specimen effective for pressing.

 Careful design of Die and Punch is essential.
 Friction is a major problem warranting effective lubrication.

This paper overviews the principle of equal channel angular pressing technique and
influences of its various processing parameters on the structure and mechanical properties
in aluminium alloys and composites.

2. Principle of ECAP

Punch

Die
φ
Billet

Figure: 1 Schematic of ECAP

2
 The two channels of equal cross section intersect at an angle  and  is the angle
subtended by the arc of curvature at the point of intersection. A well lubricated billet
which has the same cross section is placed into one of the channels .The pressing is done
by using the punch to extrude the material into the second channel as shown in Figure
1.During the movement inside the channel, deformation is achieved by simple shear at
crossing plane of the channels. The process can easily be repeated a number of times in
the same tool which introduces large amount of shear strain into the material. The die
material used are SK3(Fe and 1.0 to 1.1%C ), SKD61(Fe-C 0.32 to 0.42%-Cr 4.50 to
5.50%-Si 0.80 to 1.20%-Mn 0.50%-Mo 1.00 to 1.50%-V 0.80 to 1.20% ) and H11(Fe-C
0.33 to 0.43%-Cr 5.2%-Mo 1.35%-V 0.45% ) tool steel particularly for aluminium alloys.
The lubricant used is molybdenum disulfide. ECAP introduces shear strain into the
material equivalent to a reduction ratio of 100:1 at four passes or 10000:1 at eight passes
without any reduction in the initial material cross sectional area [1].

3. Estimation of the strain induced in ECAP

Segal[1], using the sharp cornered die and neglecting friction at the die walls, has
estimated the strain induced on the material as,

2  
N= cot  ……………… (1)
3 2

Iwahashi et al [4] have analysed the strain induced in the round corner die and shown as,

N        
N= 2 cot 2  2    cos ec 2  2   ………….. (2)
3    

where,
N = Number of passes

Wu and Baker [5] have analysed how well these estimation work in practice by
pressing layered plasticine billets of different colours through plexiglass die. Their
analysis has shown that shear strain of the plasticine layers obtained from the centre of
the billets has matched reasonably well with the theoretical values. In contrast, at
locations away from the centre, the experimental values are considerably different due to
friction effects.

4. Role of processing parameters

In ECAP process, understanding the deformation behaviour with in the die is very
important. The strain achieved is sensitive to the process parameters, namely, die angle,
deformation route, friction between the billet and die walls and the application of back

3
pressure, pressing temperature and pressing speed. The role of these processing
parameters on ECAP process is discussed below.

4.1 Die angle

.
Table 1 lists the values of strain induced with various tool angles [1]. Although
0
=90 tool angle results in maximum strain, difficulties associated with its choice
include: 1. Requirement of high pressure for successful pressing of the sample through
the die and breaking of the die and punch and 2. Incomplete filling of die corner is
encountered with less ductile materials. Use of a round cornered die (=200), a large die
angle and many passes for a given strain can minimise the above. However, Nakashima
et al [6], experimenting on the use of different combinations of large die angle and , ((i)
=900; =200,(ii) =112.50;=300, (iii) =1350 ;=130 and (iv) =157.50;=100), have
shown, its capability to achieve large strain with multi passes and incapability of
producing ultrafine grains. Finally, they concluded that ultra fine grain structure with
high angle grain boundaries(HAGB) may be achieved by imposing a very intense plastic
strain on the sample in each separate passage through the die with =900. Recently,
Furuno et al [7] studying ECAP of pure aluminium using a 600 die angle have observed
slightly smaller grain size than the 900 die angle. However, the tensile strength of pure
aluminium is mostly the same for both 900 and 600 die angles. Hence, the optimum die
angle is =900 to achieve effective strain. But, =900and =200 die angle is the
combination resulting in reduced friction and homogeneous microstructure.

Table 1: Values for strain induced for various die angles from equation (1)

Die angle
Strain induced
 
1500 00 0.31
1200 00 0.68
900 00 1.15
900 200 1.05*
* from equation (2)

4.2 Deformation route

Deformation route significantly affects the effectiveness of the grain refinement

than the other processing parameters. There are three deformation routes normally used
namely, 1.Route A (The material pressed without rotation between passes), 2. Route BC
(The rotation of 900 in the same direction between passes) and 3. Route C (The rotation
of 1800 between passes).

Ferrasse et al [8] have suggested that after one pass, intragranular slip within the
original grains is the main mechanism responsible for the creation of shear bands during

4
ECAP. After an increased number of passes, substantial evidence shows that subgrain
rotation is the main mechanism for the creation of submicron grain structure during
ECAP. Vedani et al [9] also have suggested that at higher strains the cell block spacing
decreases while the misorientation angle across both blocks and single cells increases.
This grain subdivision can rotate toward different orientation to create a set of new small
equiaxed grains with high angle grain boundaries, thus generating an ultrafine grain
structure. Furukawa et al [10] have concluded that microstructural evolution occurs most
rapidly in route BC. Iwahashi et al [11] have studied the deformation route A and C by
using the die having =900, =200 in pure aluminium. After 10 passes the same equiaxed
grain size is achieved in both A and C routes. Gholinia et al [12] have studied the
deformation routes by using the die having =1200 in Al-Mg and Al-Mn alloys and
concluded that Route A is an efficient one in terms of forming new high angle grain
boundaries (the boundary with misorientation angle greater than 150). Zhu and Lowe [13]
have suggested that the interactions of shear plane with texture and crystal structure play
a primary role in grain refinement while the accumulative strain plays a secondary role.
Iwahashi et al [14] have analysed A, BC and C routes by using pure Al. Their
microstructural evidence of shearing characteristics of pure aluminium are shown in
Table 2, which clearly shows that route BC is the preferable route for achieving equiaxed
grain with high angle grain boundaries.

Recently, Pei-Ling Sun et al [15] have analysed A, BC and C routes carefully by

using pure Al. The die angles used were =900, =200 and =1200,=00 . The equivalent
strain obtained has been approximately 8 with 8 and 12 passes with the 900 and 1200 die
respectively. The microstructural parameters of the samples are given in Table 3. The
findings of this study on three routes are summarized below.

1. The efficiency of high angle grain boundary generation for both dies is
maximum for route A followed by BC and C.
2. The effectiveness of grain refinement is maximum with route BC followed by A
and C in the 900 die and route BC and A followed by C in the 1200 die
3. The efficiency for generating equiaxed grains for both dies is maximum in
route BC followed by C and A.

4.3 Friction and backpressure

Friction is an important factor in ECAP process. High friction introduces

inhomogeneous deformation structure inside the billet and increases the working load.
Bowen et al [16] have analysed pure Al samples with three extrusion conditions, namely,
low friction, high friction and low friction with an applied back pressure. Low friction
has been achieved by using a highly polished die and lubricating the sample. The back
pressure has been obtained by applying a constant opposing force to the extruded end of
the billet using a secondary ram. They have shown that the best pressing condition is low
friction die with constraining back pressure. Aidang Shan et al [17] have shown that a die

5
 Table:2 Microstructures of pure aluminium in various processing routes and three directions[14]

Route After one pass After two passes After Three passes After 4 passes

BC

6
 Table 3: Grain size and HAGB obtained at different routes [15]

Average High angle

Extrusion condition
grain size grain boundaries
Die No.of (approx) (approx)
Route
angle passes m %
A 900 8 0.55 70
BC 900 8 0.45 62
C 900 8 0.64 38
A 1200 12 0.61 74
BC 1200 12 0.60 56
C 1200 12 0.84 32

set up with the moving second channel reduces the working load and friction for pure Al
and 6005 Al alloy.

4.4 Pressing temperature

In order to use ECAP process for some less ductile materials, the increasing
pressing temperature is used to reduce the working load and cracks during the process.
ECAP experiments conducted to evaluate the influence of pressing temperature on
different materials are summarised in Table 4. At elevated temperatures, the evolution of
the low angle boundaries is due to the absorption of moving dislocations into the
boundaries during deformation and annihilation of dislocations by cross slip and climb
becomes easier [18]. Hence, the increase of boundary misorientation becomes less
efficient at elevated temperatures. It is clear that increasing pressing temperature
increases the grain size and decreases high angle grain boundary.

4.5 Pressing speed

The influence of varying pressing speeds 8.5x10 -3, 8.5x10-2, 8.5x10-1 and 7.6
-1
mms in equal channel angular pressing at room temperature has been analysed by
Berbon et al [22] using pure Al and Al-1%Mg alloy samples and a die with =900 tool
angle. They have shown that the various pressing speeds have no significant influence on
the grain size after eight passes. But lower pressing speed provides a longer time for
recovery, so that a higher proportion of the extrinsic dislocations become absorbed in the
grain boundaries. Hence, the microstructure is more equilibrated and contains fewer
external dislocations at slow pressing speed.

7
Table 4: Grain size and High angle boundaries obtained at various pressing temperatures

High angle
Pressing conditions Grain
grain
size
boundaries
Alloy Die angle (approx)
Temperature (approx)
Route m
K %
 

300 0.2
373 0.25
Pure Al [18] 900 130 BC -
473 0.3
573 0.7

68
298 0.55
66
373 0.59
Pure Al [19] 1200 00 A 69
473 0.9
49
523 2.2
300 0.3
373 0.4
Al-3%Mg [18] 900 130 BC -
473 0.8
573 1.5
300 1.5
Al-3%Mg-0.2%Sc 373 1.7
900 130 BC -
[18] 473 1.9
573 4.5
323 0.25 55
373 0.3 25
423 0.4 15
Al 5052 [20] 1200 00 BC
473 0.7 12
523 0.8 8
573 2 15
523 0.6 80
573 1 -
Cast 7475 Al [21] 900 - A
623 1.2 -
673 1.7 76

5. ECAP Tool Design

In equal channel angular pressing, the tooling is the most important parameter
because of the introduction of very high shear strain on the billet. Friction between the
die walls and billet is also high which increases the working load and creates
inhomogeneous deformation structure. Segal et al [23] have developed an advanced

8
design die set up which has three movable walls (Figure 2). The movable wall eliminates
the sticking problem and wear of die wall. Lee et al [24] have developed a new die set up
for processing very long billets (Figure 3). In this set up, the roller has a slot which acts
as a first channel. The die and the support segment act as a secondary channel. Hence, the
continuous specimen pressing is possible using this die set up. Nishida et al [25] have
developed a new rotary type die (Figure 4). This comprises of a die set up having four
punches, rotary die and stationary die holder. In each pass, the die is rotated at angle 900
and the specimen is pressed by another punch. Thus, after each pass, the specimen is not
required to be removed from the die for the next pass.

Figure 2: Movable walls die set up(23)

Figure 3: Continuous ECAP die set up(24)

9
 Figure 4: Rotary ECAP die set up(25)

6. Mechanical properties

6.1 Strength and ductility

A reduction in grain size to the sub micrometer level has two significant
advantages [26, 27]. (i) There is an increase in the tensile strength through the Hall-Petch
relationship(y = o+ Kd-1/2, y = Yield strength, o= Frictional stress, K= Constant, d =
mean grain size) at low temperature with little or no corresponding reduction in the
overall ductility and (ii) If the ultra fine grains are stable at temperature where diffusion
is reasonably rapid , there is a possibility of achieving super plasticity at very high
strains.

Many researchers have analysed the mechanical properties of equal channel

pressed aluminium alloys, Horita et al [27] studying ECAP of six commercial aluminium
alloys have found that the tensile strengths of aluminium alloys increase with the
increasing number of pressings(Figure 5a). On the other hand after one pressing, the
elongation to failure decreases largely only with little change in the next pressing (Figure
5b). Table 6 lists the results of Stolyarov et al [28] on brittle cast Al-5 wt % Fe alloy
subjected to ECAP with 16 passes and varying backpressure. A back pressure of 275
MPa has resulted in the best combination of enhanced UTS of 244 MPa and elongation of
5.8 %.

6.2 Superplasticity

The development of ultrafine grained materials having reasonably stable

microstructures at elevated temperatures may provide an opportunity to increase the rate
of superplastic forming so that the technology can be expanded into the fabrication of

10
high volume parts for the automobile and consumer product industries. Many researchers
studied the superplasticity on ECA pressed aluminium alloys and their findings are
tabulated in Table 7. It can be inferred that equal channel angular pressed alloys have
very high percentage of elongation to failure. This characteristic has the potential for
achieving superplastic forming.

Figure 5: (a) 0.2% proof stress vs equivalent strain (b) Elongation to failure vs equivalent
strain [27]

Table 6: Properties of Al-5 wt % Fe alloy under different conditions [28]

Elongation
Backpressure(MPa) No.of passes UTS (MPa) YS (MPa)
(%)
Undeformed - 102 70 3.4
0 2 - - -
40 8 261 250 2.1
100 8 235 195 3.4
275 8 214 200 4.1
275 16 244 216 5.8

6.3 Fatigue properties

Ultrafine grain materials exhibit high tensile strength with fairly large ductility.
Such properties alone are not sufficient for engineering applications. If the ultrafine grain
materials are to be considered for successful engineering applications, their fatigue
properties will be major design issue. Fluctuating load impairs the strength of any
material in general aluminium and its alloys in particular. Steel’s fatigue strength is

11
 Table 7: Superplasticity by various aluminium alloys produced by ECAP

Testing conditions Elongation

Temperature to failure
Alloy Strain rate
K (approx.)
s-1 %
Al-3%Mg -0.2%Sc [29] 673 3.3 X 10-2 1030
Al-5.5%Mg-2.2%Li-0.12%Zr
603 10-3 550
[30]
Al -7034 (spray casting) [31] 673 10-2 1000
Al 1570 (Al-5.76%Mg-
0.32%Sc-0.3%Mn-0.2%Si- 723 5.6 X 10-2 2000
0.1%Fe) [32]
Al 2024 alloy [33] 673 1 X 10-2 500
Al-5.5%Mg-2.2%Li-0.12%Zr -1 1210
673 10
[34]
Al-5.5%Mg-2.2%Li-0.12%Zr
673 1 950
[34]

nearly half its tensile strength while that of high strength aluminium alloys is only about
one fifth of their corresponding tensile strength.

Patlan et al [35] studying the fatigue performance of ECA pressed 5056 Al alloy
have observed a certain enhancement of fatigue performance in the high–stress regime.
This advantage is missed during plastic strain controlled tests in the low cycle regime.
Chung et al [36] investigating the fatigue performance of ECA pressed 6061 Al alloy
have observed a large enhancement in fatigue life by a factor of about 10 compared to a
T6 treated alloy after a single pass. This result is remarkable since the improvement
obtained is in both low and high cycle regimes. However, this advantage disappears with
further pressing especially in the high cycle regime. Hence, when a balance between high
strength and fatigue performance is a primary interest, a single passed material can be a
good choice.

6.4 Creep resistance

At elevated temperatures, the mechanical characteristics of aluminium and its

alloys are lower than at room temperature due to enhanced creep deformation. Recently,
Sklenica et al [37] have analysed creep resistance of the ECA pressed high pure
aluminium by using 900 die and routes A,BC,C. Creep tests were performed at 473K in
compression under constant applied stress ranging between 10 and 25 MPa. The creep
resistance after one ECAP pass is found to considerably improve compared to unpressed
aluminium and decrease thereafter with successive passes.

12
6.5 Impact toughness:

Al-Si eutectic alloys are widely used in the automobile industries as pistons and
cylinder liners because of their good wear resistance and high tensile strength at elevated
temperatures. However, they have poor fracture toughness. Equal channel angular
pressing can transform these brittle casting aluminium alloys into a tougher material.
Recently, Aibin Ma et al [38] have studied the impact toughness of Al-11%Si
alloy. The absorbed energy of the alloy has increased with increasing number of passes.
A maximum value of 10 J/cm2 is obtained in the sample processed by ECAP for 32
passes at 623 K, which is 11 times the value of as cast alloy (0.9 J/cm2). The increased
impact toughness of this alloy is attributed to the breakage of the large aluminium
dendrites and interdendritic networks of eutectic silicon, the content of fine particles and
the homogenized microstructure resulting after ECAP.
.
7. ECAP of aluminium metal matrix composites

There are no reports on ECAP of metal matrix composites except one. Equal
channel angular pressing has been successfully carried out in an aluminium matrix
composite reinforced with 15vol% SiC whiskers with an average diameter and length of
~1 and ~40µm respectively[39]. The die angle used for pressing is 900and the pressing is
carried out at 723 and 753 K for the first and the second passes respectively with a speed
of pressing of 6mms-1. The study has shown homogeneous distribution of the
reinforcement in the matrix due to breaking of reinforcement whiskers by the severe
deformation during ECAP.

8. General Discussion

Equal channel angular pressing is a novel technique for producing ultrafine grain
structures with grain sizes falling between 200 to 500nm especially for aluminium alloys.
ECAP introduces large amount of shear strain into the material without any change in the
dimensions.

In ECAP process, several process parameters influence the effectiveness of grain

refinement. Although 900 die angle results in the maximum strain, difficulties are
associated because of the requirement of high pressure and incomplete filling of die
corner with less ductile materials. Use of large die angle with increased number of
pressings gives the same amount of strain. Despite its incapability to produce ultrafine
grains, the die with angle =900, =200 gives optimum strain and homogeneous
microstructure. Deformation route chosen significantly affects the grain refinement
compared to other processing parameters. The deformation route BC has the maximum
efficiency of high angle grain boundary and equiaxed grain generation compared to other
routes. The friction problem will reduce by providing good lubrication between the die
wall and the billet. The pressing speed has no significant influence on the equiaxed grain

13
size after eight passes. Increasing pressing temperature increases the grain size, makes
grain shape more equiaxed and also decreases high angle grain boundary.
ECA pressed materials have ultrafine grain size and exhibit high strength with
little or no reduction in ductility. ECA pressed materials have very high percentage of
elongation to failure and exhibit a reasonably stable microstructure at elevated
temperature. These advantages enable the possibility of achieving superplastic forming.
A large enhancement of fatigue life has been observed in Al 6061 alloy after a single
pass. The creep resistance of pure aluminium is considerably improved after one pass
compared to unpressed aluminium. ECAP can transform brittle cast Al-11%Si alloy into
a tougher material by breaking of large aluminium dendrites and interdendritic networks
of eutectic silicon, the formation of fine particles and homogeneous microstructure. By
doing ECAP, a homogeneous distribution of the reinforcement in the matrix has been
obtained in aluminium matrix composite reinforced with 15% volume fraction of SiC
whiskers.

Equal channel angular pressing is definitely a prospective technique for producing

ultrafine structures at low cost compared to other techniques namely torsion straining,
accumulative roll bonding, ball milling and asymmetric rolling. ECAP produces ultrafine
grain structure in relatively large bulk samples and is easily adopted into manufacturing
practice. ECA pressed materials have both superior properties and performace. Dearth of
knowledge on fatigue properties, creep resistance, wear resistance and fracture toughness
of ECA pressed aluminium alloys and ECAP of aluminium metal matrix composites
warrants further studies. The process still requires more engineering and technological
development in the tool design for industrial applications.

9. Summary

1. Equal channel angular pressing is a novel technique for producing ultrafine

grain structure in submicron level in bulk materials by introducing extremely
large plastic strain during the process.
2. During the process, the billet experiences simple shear without change in cross
sectional area, making number of pressings possible.
3. Commercial cast alloys with coarse grain size can be transformed into
ultrafine grain structure by ECAP.
4. In ECAP, the optimum conditions for achieving uniform and intensive shear and
homogeneous microstructure are die angle ~900,=200 and deformation route
BC with low friction die.
5. ECA pressed materials are porosity free in nature and exhibit high strength at low
temperature and superplastic at high temperature.
6. ECA pressed 6061 aluminium alloy after one pass has the higher fatigue life
compared to T6 treated 6061 aluminium alloy.
7. The creep resistance of pure aluminium is improved after one ECAP pass
compared to unpressed material.
8. In equal channel angular pressing, the impact toughness of the brittle cast

14
 Al-11%mass Si alloy is improved approximately 11 times greater than the as
cast alloy.
9. ECAP results in homogeneous distribution of the reinforcement in the matrix in
auminium matrix composite reinforced with 15% volume fraction of SiC
whiskers .
10. Further studies are essential to make ECAP for industrial applications.

Acknowledgement
The authors thank the Director, Regional Research Laboratory,
Thiruvananthapuram, the members of the Metals Processing Groups for their
encouragement and Council of Scientific and Industrial Research, New Delhi, for the
award of CSIR diamond jubilee Research Internship to the first author. The authors also
thank all the authors and the respective publishers, whose illustrations/figures have been
made use of.

References
[1] V.M.Segal , Mater.Sci.Engg.A197 (1995) 157.
[2] V.M.Segal , Mater.Sci.Engg.A271 (1999) 322.
[3] V.M.Segal , Mater.Sci.Engg.A386 (2004) 269.
[4] Y.Iwahashi ,J.Wang, Z.Horita and M.Nemoto , Scripta Mater.35 (1996) 143.
[5] Y.Wu and I.Baker , Scripta Mater.37 (1997) 437.
[6] K.Nakashima, Z.Horita , M.Nemoto and T.G.Langdon , Acta Mater. 46 (1998) 1589.
[7] K.Furuno, H. Akamatsu , K.Oh-ishi , M.Furukawa , Z.Horita and T.G.Langdon ,
Acta Mater. 52 (2004) 2497.
[8] S.Ferrasse , V.M.Segal , K.T.Hartwig and R.E.Goforth, Metall . Mater . Trans.28A
(1997) 1047.
[9] M.Vedani, P.Bassani, A.Tuissi and G.Angella, Metallurgical science & technology,
22 (2004),2,21-30.
[10] M.Furukawa , Y.Iwahashi , Z.Horita,M.Nemoto and T.G.Langdon , Mater.Sci.Engg
A 257 (1998) 328.
[11] Y.Iwahashi , Z.Horita,M.Nemoto and T.G.Langdon , Acta Mater. 45 (1997) 4733.
[12] A.Gholinia , P.B.Prangnell and M.V. Markushev , Acta Mater. 48 (2000) 1115.
[13] Y.T.Zhu and C.Lowe , Mater.Sci.Engg A 291 (2000) 46.
[14] Y.Iwahashi , Z.Horita,M.Nemoto and T.G.Langdon , Acta Mater. 46 (1998) 3317.
[15] Pei-Ling Sun, Po-We kao and Chih-Pu Chang , Metall . Mater. Trans.35A
(2004)1359.
[16] J.R.Bowen , A.Gholinia , S.M.Roberts and P.B.Prangnell , Mater.Sci.Engg A 287
(2000) 87.
[17] A.Shan , In-Ge Moon and Jong-Woo Park , J.Mater.Process.Technol., 22(2002) 255.
[18] A.Yamashita , D.Yamasuchi , Z. Horita and T.G Longdon, Mater.Sci.Engg A 287
(2000)100.
[19] Y.C.Chen, Y.Y.Huang , C.P.Chang and P.W.Kao, Acta Mater. 51 (2003) 2005.
[20] Y.Y.Wang , P.L.Sun , P.W.Kao and C.P.Chang , Scripta Mater.50 (2004) 613.
[21] A.Goloborodko, O.Sitdikov,R .Kaibyshov , H. Miura and T.Sakai , Mater.Sci.Engg

15
 A 381 (2004) 121.
[22] P.B.Berbon , M.Furukawa , Z.Horita,M.Nemoto and T.G.Langdon , Metall . Mater .
Trans.30A (1999) 1989.
[23] V.M.Segal, R.E.Goforth, T.D.Hartwig, U.S. Patent No. 5,400,633 (1995).
[24] J.-C.Lee, H.-K.Seok, J.-W.Park, Y.-H.Chung and H.-J.Lee, U.S. Patent No.
6,370,930 (2000).
[25] Y.Nishida, H.Arima, J.C.Kim and T.Ando, Scripta Mater.45 (2001) 261.
[26] Z.Horita , T.Fujinami , M.Nemoto and T.G.Langdon , Metall . Mater . Trans.31A
(2000) 691.
[27] Z.Horita , T.Fujinami , M.Nemoto and T.G.Langdon , J.Mater.Process.Technol,117
(2001) 288.
[28] V.V.Stolyarov , R.Lapovok , I.G.Brodova and P.F.Thomson , Mater.Sci.Engg A
357 (2003)159.
[29] S.Komura , P.B.Berbon , M.Furukawa , Z.Horita , M.Nemoto and T.G.Langdon ,
Scripta Mater.38(1998) 1851.
[30] M.Furukawa , P.B.Berbon , Z.Horita,M.Nemoto , N.K.Tsenev , R.Z.Valiev and T.G.
Longdon , Metall . Mater . Trans.29A (1998)169.
[31] Cheng Xu, M.Furukawa , Z.Horita and T.G Longdon, Acta Mater. 51(2003) 6139.
[32] F.Musin , R.Kaibyshev,Y.Motohashi and G.Itoh, Metall . Mater . Trans.35A
(2004) 2383.
[33] S.Lee , M.Furukawa , Z.Horita and T.G.Longdon , Mater.Sci.Engg A 342 (2003)
294.
[34] P.Lee , M.Furukawa , P.B.Berbon , Z.Horita,M.Nemoto , N.K.Tsenev , R.Z.Valiev
and T.G .Longdon , Mater.Sci.Engg A 272 (1999) 63.

[35] V.Patlan , A.Vinogradov , K.Higashi and K.Kitakawa , Mater.Sci.Engg A 300(2001)

171.
[36] C.S Chung , J.K.Kim , H.K.Kim and W.J.Kim , Mater.Sci.Engg A 337 (2002) 39.
[37] V.Skelenica , J.Dvorak and M.svoboda , Mater.Sci.Engg A 387-389 (2004) 696.
[38] A.Ma , K.Suzuki , Y.Nishida , N.Saito, I.Shigematsu , M.Takagi , H.Iwata,
A.Watazu and Toru Imura ,Acta Mater.53 (2005) 211.
[39] D.Ma, J.Wang and K.Xu , Materials Letters 56 (2002) 999.

16