Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Investigation of the influence of severe plastic deformation on the microstructure and mechanical properties of Al-7136 alloy
(USC Thesis Other)
Investigation of the influence of severe plastic deformation on the microstructure and mechanical properties of Al-7136 alloy
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INVESTIGATION OF THE INFLUENCE OF SEVERE PLASTIC DEFORMATION
ON THE MICROSTRUCTURE AND MECHANICAL PROPERTIES OF
Al-7136 ALLOY
by
Zhichao Duan
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MECHANICAL ENGINEERING)
August 2009
Copyright 2009 Zhichao Duan
ii
ACKNOWLEDGEMENTS
First and foremost, I would like to express my great gratitude to my advisor ,
Professor Terence G. Langdon for his great support and invaluable guidance throughout
my Ph.D. study at the University of Southern California. I have been fortunate to get to
know Prof. Langdon and am always admiring his profound knowledge and great
enthusiasms in materials research. My achievement would not even be possible without
his encouragements and inputs.
I would like to thank my dissertation committee members, Professor Andrea M.
Hodge and Professor Edward Goo, for their evaluation of my Ph.D. study. I would also
like to show my sincere appreciations to my collaborators of this research: Professor
Nguyen Q. Chinh, from the Department of General Physics, Eotvos University, Hungary
for his valuable discussions on the ECAP processings.; Professor Xiaozhou Liao, from
the School of Aerospace, Mechanical and Mechatronic Engineering, University of
Sydney ,Australia for his great work in TEM studies; and Dr. Gang Sha, from the
Electron Microscopy Unit (EMU) at the University of Sydney, Australia, for his excellent
contributions in 3-D Atom Probe Tomography (3-D APT) work.
My thanks also extend to my group members at USC: Dr. Cheng Xu, Dr. Megumi
Kawasaki , Dr. Roberto Figueiredo and others, for their instructive discussions and
helpful advices in the last three years.
Finally, I would like to present this thesis as a gift to my grandpa, my parents
iii
and the whole family thousands of miles away in China. It is their everlasting love,
supports and encouragements that intensified my determination in pursuing this degree.
iv
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ii
LIST OF TABLES vi
LIST OF FIGURES vii
ABSTRACT xvi
Chapter 1: Introduction 1
Chapter 2: Literature Review 5
2.1. Severe Plastic Deformation 5
2.2. Equal Channel Angular Pressing (ECAP) 6
2.2.1 Introduction of Equal Channel Angular Pressing (ECAP) 6
2.2.2 Parameters during ECAP processing 10
2.2.3 Microstructural Characteristics after ECAP 22
2.2.4 Mechanical Properties achieved after ECAP 25
2.2.5 ECAP on 7000 Series Aluminum Alloys 28
2.3. High Pressure Torsion 37
2.3.1 The Fundamental parameters in processing by HPT 40
2.3.2 Microstructural development in more complex metallic alloys 53
2.3.3 Combing processing by HPT with other processing methods 56
2.3.4 Materials Softening during HPT 62
Chapter 3: Experimental Materials and Procedures 66
3.1. Experimental materials 66
3.2. Experimental facilities ,analytical methods and sample preparation 66
3.2.1 ECAP 66
3.2.2 High Pressure Torsion 69
3.2.3 Mechanical Testing 71
3.2.4 Microstructural analysis and sample preparations 75
3.3. Overall Summary of Experimental Procedures 80
3.3.1 Summary of the ECAP part 80
3.3.2 Summary of the HPT part 83
Chapter 4: Experimental Results 85
4.1. ECAP of Al-7136 alloy 85
4.1.1 ECAP of Al-7136 alloy at 473K 85
4.1.2 Develop a room temperature ECAP method 91
4.1.3 Pressing Al-7136 at 110 ℃ after solid solution 99
4.1.4 Microstructures study of Al-7136 102
v
4.2. High Pressure Torsion on Al-7136 124
4.2.1 Microhardness study of Al-7136 after HPT 124
4.2.2 Microhardness study of Al-7136 after 1 pass room temperature
ECAP plus HPT processing 133
4.2.3 Microhardness study of Al-7136 after 4 passes ECAP at
473 K plus HPT processing 138
4.2.4 Microstructures of Al-7136 after HPT processing 143
Chapter 5: Discussion 160
5.1. The effect of Severe Plastic Deformation on precipitates in Al-7136 160
5.1.1 Precipitation during ECAP of the Al-7136 162
5.1.2 The effect of ECAP on the precipitation kinetics 163
5.1.3 The effect of ECAP on h precipitate orientation and morphology 165
5.2. Segregation and depletion of alloying element s at grain boundaries
during SPD 169
5.3. Discussion of the effect of ECAP on Al-7136 177
5.4. Discussion of the effect of HPT on Al-7136 178
Chapter 6: Summary and Conclusions 190
REFERENCES 196
ALPHABETIZED BIBLIOGRAPHY 208
vi
LIST OF TABLES
Table 4.1 Hardness of Al-7136 after solid solution at various temperatures,
*denotes the test was done 5 days after solid solution. 94
Table 4.2 The nominal compositions of Al-7136. 108
Table 4.3 The compositions of large precipitates formed under different passes
ECAP in the Al-Zn-Mg-Cu alloy (7136) 129
vii
LIST OF FIGURES
Figure 2.1 iIllustration of the ECAP die with indication of the characteristic angles. 8
Figure 2.2 Die geometries (a) Segal’s die assumption [47] (b) Iwahashi’s die
assumption [40]. 13
Figure 2.3 (a)(b) Die showing equal fillet radii geometry which can be used in
ECAP, where f is the angle of intersection of the two channels ,
R is the fillet radius and r is the inner fillet radius.[14] 15
Figure 2.4 Equivalent plastic strain for different die geometry and angle for
L
0
=10 mm and R= 10 mm. [14] 16
Figure 2.5 Total equivalent plastic strain with friction modeled by Luis in the
proposed optimized die geometry [14] 16
Figure 2.6 Maximum principal stress distribution with different outer corner
angle ECAP dies at the middle of the deformation step [15]. 17
Figure 2.7 Contour maps showing the Vickers microhardness distribution in
the cross-section of pure aluminum billets subjected to ECAP for
4 passes via route BC using a solid die with y= 20 ゜ [48]. 20
Figure 2.8 Shearing characteristics for various routes [50] 21
Figure 2.9 Grain size versus the pressing temperature for pure Al and Al-3% Mg
and Al-3% Mg-0.2% Sc alloys [55]. 23
Figure 2.10 Variation of the fraction of high-angle boundaries with the number
of passes through the die for purealuminum using route BC: datum
points are shown for the X and Y planes [62]. 27
Figure 2.11 A comparison of yield strength and ductility for an Al-3004 alloy
processed by cold-rolling or ECAP [66]. 27
Figure 2.12 Microstructure in the as-received Al-7034 alloy showing the presence
of rod-shaped MgZn2 precipitates [78]. 33
viii
Figure 2.13 Microstructureof Al-7034 after 1 pass(a) and 2 passes (b) of ECAP at
473 K.[78] 33
Figure 2.14 Engineering Strain versus Engineering Stress at test (a) at ambient tem-
perature and (b) 673 K [78]. 35
Figure 2.15 Tensile engineering stress–strain curves for the UFG (solid circle)
and CG (empty circle) samples, both of which were naturally aged
for one month.[80] 38
Figure.2.16 The values of the Vickers microhardness as a function of pre-aging
time for the samples processed by only 1 pass in ECAP.[81] 38
Figure.2.17 The values of the Vickers microhardness as a function of the number of
passes in ECAP for samples pre-aged for 10 minutes [81]. 39
Figure.2.18 Scehmatic Illustration of High Pressure Torsion(HPT) [28] 42
Figure.2.19 Schematic illustration of HPT for (a) unconstrained and (b, c)
constrained conditions [29]. 42
Figure.2.20 Parameters used in estimating the total strain in HPT [28]. 43
Figure.2.21 Microstructures of HPT nickel for (a) the central part of the disk
after 5 turns at 1 GPa, (b) the edge of the disk after 5 turns at 1 GPa
the SAED patterns were taken with an aperture size of 1.8 mm [97]. 47
Figure.2.22 Microhardness profiles of nickel processed by HPT at two different
applied pressures[97] 47
Figure.2.23 Microstructures of HPT nickel at the edge of the disk after 5 turns
at 6 GPa: the SAED patterns were taken with an aperture
size of 1.8 m[97]. 50
Figure.2.24 Accumulated equivalent strain versus distance from the specimen
center in the first-order strain gradient model [98]. 52
Figure.2.25 Bright-field TEM micrographs of copper processed by HPT (a) at the
edge of the disk and (b) showing deformation twins [102]. 55
Figure.2.26 Plot of normalized minimum grain size versus normalized stacking fault
energy for Cu and two Cu–Zn alloys showing different slopes from
experimental data obtained using TEM and XRD [102]. 55
ix
Figure.2.27 Microstructures of aluminum: (a) optical microscopy showing the initial
annealed structure and (b) TEM micrograph showing the structure after
applying a pressure of P = 1 GPa without torsional straining [29]. 57
Figure.2.28 Representative TEM micrographs of commercial purity aluminum
after HPT with a pressure of P = 1 GPa (a) in the central region and
(b) near the periphery after two whole revolutions [29]. 57
Figure.2.29 Optical micrographs of the central region of a disk of an austenitic steel
processed by HPT (a) after N = 2.27 revolutions and (b) after N = 16
revolutions [97]. 59
Figure.2.30 Microhardness profiles for six samples of an austenitic steel processed
by HPT through different numbers of revolutions [97]. 60
Figure.2.31 Microstructure of nickel processed using different procedures:
(a) ECAP, (b) ECAP + CR, (c) HPT, (d) ECAP + HPT and
(e) ECAP + CR + HPT [114]. 61
Figure.2.32 The grain boundary misorientation distributions in nickel after (a) ECAP
, (b) HPT and (c) ECAP + HPT [114]. 64
Figure 2.33 Microhardness of ED NC Ni and ED NC Ni + HPT. The microhardness
of ED NC Ni + HPT is lower than that of as-deposited ED NC Ni, and it
decreases with increasing shear strains. [117] 65
Figure 2.34 Scanning electron micrographs of ED NC Ni + HPT illustrating the grain
coarsening after SPD. (a) Definition of the direction of observations; (b)
micrograph near the center in the axial direction; (c) micrograph at a
radius of about 3 mm in the axial direction; (d) micrograph at a radius of
about 3 mm in the radial direction [117]. 65
Figure 3.1 The design of (a) the ECAP die and (b) the ECAP plunger. 68
Figure 3.2 Schematic drawing of the pressing facility. 70
Figure 3.3 Schematic drawing of the High Pressure Torsion (HPT) facility. 74
Figure 3.4 Schematic drawing of the tensile specimen. 77
Figure 3.5 Grid pattern of the Vickers Hardness measurement on the HPT discs. 77
x
Figure 3.6 Illustration of the 3-D Atom Probe technology.[119] 81
Figure.4.1 Room temperature tensile tests of Al-7136 after ECAP at 473 K
via route Bc. 87
Figure 4.2 Engineering stress-strain rate curve of Al-7136 alloy after 4 passes of
ECAP, tested at 673 K and various strain rates. 92
Figure 4.3 Engineering stress-strain rate curve of Al-7136 alloy after 4 passes of
ECAP, tested at 703 K with various strain rates. 92
Figure 4.4 Engineering stress-strain rate curve of Al-7136 alloys after different
passes of ECAP, tested at 703 K. 93
Figure 4.5 Engineering stress-strain rate curve of as-received Al-7136 alloys,
tested at 703 K with various strain rates. 93
Figure 4.6 Surface topography of billets after 1 ECAP pass at room temperature
in the case of Al-7136 a) solution treated at 743 K for a) 30 mins;
b)60 mins. 97
Figure.4.7 Room Temperature Tensile properties of Al-7136 Processed by ECAP
Pressed under different conditions. 100
Figure 4.8 The high temperature stress –strain curve of the room temperature
pressed Al-7136. 101
Figure.4.9 Room Temperature Tensile properties of Al-7136 Processed by ECAP
Pressed under different conditions. 104
Figure.4.10 Room Temperature Tensile properties of Al-7136 Processed by ECAP
Pressed under different conditions. 105
Figure.4.11 The mass spectrum of the as-received Al-7136. 108
Figure.4.12 The XRD profile of Al-7136 in different conditions. 110
Figure.4.13 Microstructure of the as-received Al-7136. 112
Figure.4.14 Microstructure of the 473 K ECAP 1-Pass Al-7136. 112
Figure.4.15 Microstructure of the 473 K ECAP 4-Pass Al-7136. 114
xi
Figure.4.16 Microstructure of the 473 K ECAP 8-Pass Al-7136. 114
Figure.4.17 Microstructure of the Solid solution plus room temperature ECAP
1-Pass Al-7136. 115
Figure.4.18 TEM micrograph (a) and selected area diffraction pattern (SADP)
(b) of as-received Al-7136. 120
Figure 4.19 Element maps of an as-received Al-7136 sample, (a) Mg map, (b),
(c) and (d) are Mg, Cu and Zn map after removing matrix atoms. 120
Figure 4.20 TEM bright field (BF) images and corresponding SADPs of Al-7136
samples thermally aged at 200°C up to 40 mins. (a), (b) and (c)
are BF images and (d), (e), (f) are SADPs aged for 5, 20 and 40 mins,
respectively. 122
Figure 4.21 TEM BF micrographs of (a) 1-pass, (b) 4-pass and (c) 8-pass ECAP
samples of a Al-Zn-Mg-Cu (7136 ) alloy. 123
Figure 4.22 SADPs of Al-7136 samples experienced (a) 1-pass and (b) 4-pass
ECAP at 200 °C, (c) the SADP of 4-pass sample superimposed with
white artificial diffraction rings of the h precipitate and (d) the 4-pass
SADP superimposed with red artificial rings from S-phase. 125
Figure 4.23 Mg maps of Al-7136 alloy samples experienced with (a)1-pass,
(b) 4-pass and (c) 8-pass ECAP. 126
Figure 4.24 Matrix solute concentration evolution (a) and precipitate number
density evolution (b) in the Al-Zn-Mg Cu alloy as a result of different
number of ECAP processing. 129
Figure.4.25 Vickers Hardness (HV) distribution on Al-7136 discs after room
temperature HPT at 6 GPa. 131
Figure.4.26 The histogram of the Vickers Hardness (HV) distribution on Al-7136
discs after room temperature HPT at 6 GPa. 134
Figure.4.27 The contour maps and histograms of the Hv distribution on Al-7136
discs after room temperature HPT at 6 GPa. 139
Figure.4.28 The contour maps and histograms of the Hv distribution on Al-7136
discs after 4 passes ECAP at 473 K plus HPT processing. 144
Figure.4.29 Partition of the HPT disk for TEM investigation purpose. 146
xii
Figure.4.30 TEM images of the center part of the 1 turn HPT Al-7136 sample. 146
Figure.4.31 TEM images of the peripheral part of the 1 turn HPT Al-7136 sample. 149
Figure.4.32 TEM images of the center part of the 2 turn HPT Al-7136 sample. 149
Figure.4.33 TEM images of the peripheral part of the 2 turn HPT Al-7136 sample. 150
Figure.4.34 TEM images of the peripheral part of the 4 turn HPT Al-7136 sample. 152
Figure.4.35 TEM images of the center part of R.T.ECAP plus 1turn HPT
Al-7136 sample 156
Figure.4.36 TEM images of the peripheral part of R.T.ECAP plus 1 turn HPT
Al-7136 sample 156
Figure.4.37 TEM images of the center part of R.T.ECAP plus 2 turn HPT
Al-7136 sample 157
Figure.4.38 TEM images of the peripheral part of R.T.ECAP plus 2 turn HPT
Al-7136 sample. 158
Figure 5.1 Multiple view examinations of 7 pairs of close neighbour precipitates
adjacent to each other inside the 1-pass ECAP Al-7136 sample shown
in Fig. 6a. (a), (b), (c) shows 3 cases of close neighbour small
precipitates, (d), (e), (f) and (g) shows a small precipitate touching
a larger one. 168
Figure 5.2 Elemental maps of Al-7136 alloy sample after 4 passes of ECAP
processing: (a) Mg containing a selection box, (b) Cu, (c) Zn, (d) Si,
(e) Cr and (f) Mn maps. 172
Figure 5.3 Composition profiles from a selection box with the z axis parallel
to the grain boundary normal from Al-7136 alloy sample after 8
passes of ECAP as shown in Fig. 5. 2 (a). 174
Figure 5.4 Elemental maps of a polycrystalline region in Al-7136 sample after
8 passes of ECAP. (a) Mg, (b) Cu and (c) Zn maps. 174
Figure 5.5 Composition profiles obtained using a selection box of 20 x20 x 40 nm
with the z axis parallel to the normal of each grain boundary in the
sample after 8 passes of ECAP as shown in Fig.5. 4: the composition
profiles from (a) GB1, (b) GB2 and (c) GB3. 179
xiii
Figure.5.6 Concentration of Zn in (Al) solid solutions in various Al-Zn alloys
before and after HPT. [159] 182
Figure.5.7 Dependence of: (a) hardness and (b) lattice spacing on deformation
degree (number of torsions) for Al–30 wt.% Zn alloy.[189] 187
xiv
ABSTRACT
The super-saturated Al-Zn-Mg-Cu alloys or 7000 series Al alloys are increasingly
replacing the 2000 series Al alloys in aerospace applications due to its outstanding
strength. The potential of further strengthening the 7000 series Al alloys using SPD
methods makes it tempting to conduct ECAP as well as HPT on the selected Al-7136
alloy.
ECAP processing at 473 K was proven to be feasible for Al-7136 alloys as there
was no cracking or segmentations of the rods after 8 passes of ECAP. TEM study
revealed that this could effectively refine the grain size and achieve an equiaxed
microstructure in Al-7136 after 8 passes of ECAP. However, the mechanical properties of
the as-pressed Al-7136 were mostly deteriorated compared to the as-received material.
To avoid the negative effect of processing at 473 K, an innovative ECAP
processing routine including solid solution, quenching and immediate room temperature
ECAP pressing was successfully carried out on Al-7136 rods. The room temperature
mechanical properties were consequently strengthened but still it failed to improve the
high temperature performance.
3-D Atom Probe Tomography (3-D APT) as well as X-ray diffraction (XRD)
revealed a promotion of the precipitates growth and the segregation of the alloying
elements at the grain boundaries in the as-pressed samples, they are considered as the
mains reasons for the softening of the Al-7136 alloy despite of the grain refinement.
xv
HPT was also conducted on Al-7136, besides direct HPT processing, different
pre-processing of ECAP including 1 pass room temperature pressing and 4 passes
pressing at 473 K were also employed before HPT processing, respectively. A
complicated microhardness evolution dependent on the pre-processing as well as strains
imposed was found.
The effect of HPT on the grain refinements and diffusion coefficients of alloying
elements were discussed based on the TEM images of grain size and precipitates. It was
concluded that HPT processing could significantly refine the grain size, meanwhile, the
diffusion coefficients were much higher in materials after HPT processing that would
possible lead to similar precipitation and grain boundary segregation as that in ECAP.
This could successfully explain the hardness evolution mentioned earlier.
1
1. INTRODUCTION
The well-known Hall-Petch equation states that the strength of all polycrystalline
materials increases with a reduction in the grain size. Along with the enhanced strength
due to the Hall-Petch relation, grain refinement also improves the low and warm
temperature ductility. This has led to an ever-increasing interest in fabricating materials
with extremely small grain sizes.
Traditionally, thermo-mechanical processing is a common way to refine the grain
structure of commercially polycrystalline alloys. Reasonable balance of strength and
ductility could be achieved through hot rolling or extrusion [1-5], however, the minimum
grain size that could be achieved by these methods is in the order of a few micrometers.
In order to achieve even better combinations of strength and ductility, smaller grain size
is one prerequisite.
The definition of UFG (ultrafine-grained) materials is thus introduced. With
reference to the characteristics of polycrystalline materials, UFG materials are defined as
bulk polycrystalline materials having fully homogeneous and equiaxed microstructure
with average grain sizes less than ~1 m and with a majority of the grain boundaries
having high angle of misorientations[6]. In fact, characterization methods like Electron
Back Scattered Diffraction (EBSD) or X-ray Diffraction (XRD) often reveals the
existence of even smaller subgrains or substructures, sometimes with the size of less than
100 nm. This thus led to a more general term of defining this category of materials as
Bulk Nanostructured Materials.
2
Severe Plastic Deformation processes [7] such as Equal Channel Angular Pressing
(ECAP) [6] and High Pressure Torsion (HPT) [8]have proven to be effective to refine the
structure of several metallic materials to grain sizes lower than 1 m in most cases. Such
structures present not only remarkable strength but also provide potential to superplastic
deformations in several alloys [9-12].
Many experimental studies and new developments in ECAP took place in the last
decade. While channel angle and pressing temperature are generally considered as the
most important parameters for ECAP, the effect of die geometry on the imposed strain
and also the homogeneity of the samples has drawn the attention from researchers [13-
16], A few improvements on the die geometry have been proposed, with generally
theoretical analysis or finite element modeling methods (FEM) [14-16]. One of the new
setup from Luis et al. [14] is believed to be effective in improving the processing
according to the FEM result, thus an experimental study is tempting in evaluating the
new die geometry.
With the comparatively easy to deform f.c.c structure, Al and Al alloys are among
the most popular materials in Bulk Nanostructured Materials studies [7]. Significant grain
refinements have been reported in pure aluminum and various aluminum alloys, the grain
sizes are within the UFG range. Superplasticity was also found in Al alloys by various
researchers through ECAP [17-19]. A trend of ECAP application on Al and Al-alloys is
the use of ECAP on commercial alloys rather than lab-designed eutectic or eutectoid
alloys. The Al 7000-series alloys, which are alloyed with zinc and magnesium, can be
precipitation hardened to the highest strengths of any aluminium alloy. It is increasingly
3
used to replace the Al 2000-series alloys in the rapid-growing aero industry due to its
outstanding resistance to stress corrosion cracking.
There are a few studies on the Al 7000-series alloys, reporting superplasticity and
microstructure changes resulted from ECAP [20-22]. Due to the complicated chemical
composition, the mechanical property changes are different from other materials; also,
the effect of ECAP on the precipitates is an interaction of temperatures ,die pressures and
die configurations, thus different ECAP processing procedures could lead to different
microstructure changes and mechanical behavior even in the same alloy. More research is
needed to better understand the mechanism involved in ECAP on Al 7000-series alloys.
Meanwhile, no report has been seen on the effect of ECAP on the newly developed Al
7136 alloy.
Besides ECAP, another popular “ top down” processing method to achieve UFG
materials is High Pressure Torsion (HPT). The sample, usually in the form of a disk, is
located between two anvils where it is subjected to a compressive applied pressure, in the
order of GPa at room temperature or elevated temperature and simultaneously subjected
to a torsional strain which is imposed through rotation of the lower anvil [8]. There are
numerous scientific researches reporting the exceptional grain refinement and
strengthening effect of HPT in the last five years or so [23-27]. Due to the fundamental
factors that the imposed strain is correlated to the distance from the center of the disc,
homogeneity across an HPT disk is one of the most studied topics. Despite theoretical
calculations indicate a decreasing imposed strain when the distance to the center of the
disc decreases [28], experimental studies suggest there is potential in many materials for
4
achieving a gradual evolution into a reasonably homogeneous microstructure [23,24,29].
Nevertheless, the variation is noticeable especially in the early stages of deformation, the
microstructures appear to become reasonably homogeneous across the disks when total
strain is sufficiently high under a high imposed strain. As of date, most of these HPT
studies focused on high purity metals [30,31] or less-complicated alloys, it is thus a
challenge yet of great interest to investigate the effect of HPT on more complicated
commercial alloys for its further development in industrial applications.
5
2. LITERATURE REVIEW
2.1. Severe Plastic Deformation (SPD)
As previously introduced, Ultrafine-Grained (UFG) materials attracted much
attention in the last a couple of decades for its remarkable physical and mechanical
properties. The high room temperature strength as well as the ability to exhibit
superplastic flow if the fine grain size could be retained marked the advantage of UFG
microstructure. It is well established that bulk materials with UFG microstructure could
be fabricated through two different approaches, the “bottom up” and “top down”
procedures.
While bulk solid could be fabricated through the assembly of individual atoms or
nanoparticulate solids and the grains are exceptionally small and homogeneously
distributed in the “ bottom up” methods, the disadvantages of introducing contamination
and residual porosity as well as the incapability of scaling up set a hard-to-overcome
limitation to the further application of them into more practical areas.
All these disadvantages could be avoided in the “ top down” procedures, they
apparently showed a brighter future in industrial applications. There are several different
procedures of processing bulk solids to achieve UFG microstructures, with all of them
sharing the same idea of imposing heavy straining and thus upon the introduction of a
very high dislocation density. A term called Severe Plastic Deformation (SPD) is thus
used as the general term of describing these processing methods. The definition of SPD
processing is “any method of metal forming under an extensive hydrostatic pressure that
6
may be used to impart a very high strain to a bulk solid without the introduction of any
significant change in the overall dimensions of the sample and having the ability of
produce exceptional grain refinement.” [32] One of the biggest advantages of SPD is the
retaining of the geometric shape, which makes it possible to impose an extremely high
strain and consequently a very high density of lattice dislocation that eventually refines
the grains significantly.
Some of the most popular SPD processing techniques include Equal Channel
Angular Pressing (ECAP), High Pressure Torsion (HPT) and Accumulative Roll-Bonding
(ARB) [33]. They have been well studied and reported as effective methods of refining
microstructure and producing UFG bulk materials through high plastic straining. The
grain sizes achieved are in submicrometer or even nanometer range, thus it often leads to
superior mechanical properties [7,34-36].
Among all of the SPD techniques, ECAP has been studied the most in the last two
decades. It could provide an excellent capability for producing favorably homogeneous
ultrafine grains in large bulk metallic materials. ECAP is able to introduce intense strains
by simple shear through an L-shaped die without any reduction of the cross-sectional area
of the bulk sample.
2.2. Equal Channel Angular Pressing (ECAP)
2.2.1.Introduction of Equal Channel Angular Pressing (ECAP)
Numerous investigations have been performed concerning various aspects of SPD
processing, with the mechanical properties examined from the macro-scale to nano-
7
indentation and with the investigations covering a wide range from conventional
microstructure examinations to atomistic simulations of the deformation mechanisms.
ECAP is widely considered to be the most attractive and potentially most useful one,
where a rod is pressed through a die constrained within a channel which is bent through
an abrupt angle, , and with a corner curvature angle, , as illustrated in Figure 2.1. An
intense shear strain thus would be introduced into the sample .
Besides the advantage same as other “top down” methods in no porosity or
contaminants involved during the processing, it is also a simple processing technique that
has the capability of being developed into a continuous processing procedure [16].
Processing by ECAP leads to significant strengthening of the material at ambient
temperatures [19, 37-38], and, provided the ultra-fine grains have a reasonable thermal
stability, it often leads to the occurrence of superplastic ductility at high strain rates at
elevated temperatures [9, 10, 39].
Valiev and Langdon [6] reviewed this technique and summarized the recent
developments, effects of diverse processing parameters and resulting structure and
properties. A basic characteristic of this technique is that the sample cross section
remains unchanged enabling repetitive processing, by which very high total strain could
be imposed on the samples. The magnitude of the total strain imposed on the sample
through a series of continuous ECAP pressings could be estimated through first principles
using the two angles and ,strain imposed by the process is given by the equation[40]:
2 2
cosec
2 2
cot 2
3
N
N
(2.1)
8
Figure 2.1 Illustration of the ECAP die with indication of the characteristic angles.
9
where
N
is the strain imposed after multiple pressings and N is the number of ECAP
passes. Further, reasonable agreements to the equation are obtained from alternative
views including approaches based on friction between the die and the material generated
during pressing and a variety of sets of angles and .
While equation 2.1 is based on the ideal procedure of frictionless pressings, a
finite element modeling (FEM) analysis shows evidence for the reduction of friction
with increasing die angle [41] and a higher uniformity of deformation by pressing
using a larger work-piece [42,43]. On the other side, experimental studies further
support the equation: a model experiment was conducted using layers of colored
plasticcine passing through a plexiglass die [11]. The strain measured in the experiment
was very consistent with the value predicted by Eq.2.1,especially at points close to the
center of the sample rods; Another modeling analysis was conducted using a combination
of two half-rods of pure aluminum with grid marks on their common interfaces and using
a die having an angle of =90
o
and =0
o
[44].
A direct measurement of strain in the experiment showed reasonable agreement
with the value predicted by Eq.2.1 whereas only a limited non-uniform area was found in
the billet. Consequently, The strain imposed to the billet, ~1 per pass in the most common
dies settings, ~90
o
,promotes structure refinement by rearrangement of dislocations
resulting in final grain sizes of less than ~1 m in most metallic materials.
10
2.2.2. Parameters during ECAP processing
2.2.2.1. Channel angle
As the most important parameters of the ECAP setup, is believed to be
influencing the imposed-strain per pressing, as well as the workability of the material. An
earlier investigation analyzed the influence using four different angle of of 90
o
, 112.5
o
,135
o
, and 157.5
o
[45]. In order to highlight the effect of , the total strains imposed
on through different passes of ECAP processing with various channel angles were
identical. Surprisingly, the ideal equiaxed ultrafine grained material with homogeneous
microstructure separated by high angle grain boundaries was only observed with the
channel angle of 90
o
. It is therefore suggested an intense strain is necessary at each
ECAP pass to achieve a UFG structure rather than the total accumulated strain.
Despite it is concluded the 90
o
channel angel is the most effective condition in
ECAP processing to produce UFG materials, the intense strain per pass resulted from this
setup may introduce in noticeable cracks or catastrophic shears bands in brittle materials.
In that case, a larger channel angle such as 110
o
or 135
o
would be employed to
successfully press the materials. Earlier investigations found successful ECAP processing
of 8 passes of tungsten using a die with =110
o
at ~1000 ℃ rather than failure at =90
o
[46].
2.2.2.2. Outer corner curvature and related die geometry
The die configuration of ECAP evolved during the development of this technique.
Segal first proposed the die configuration as shown in Fig.2.2 (a) [47], in which an
11
incompressible material and a complete filling of the channel were assumed. The analysis
from Segal concluded that the shear strain as in Eq.2.2:
)
2
cot( 2
(2.2)
This analysis does not take the fillet radii into account, thus, a higher fillet radii
could lead to a discrepancy to the actual results. Iwahashi et al. proposed a die
configuration [40], which includes a fillet radius in the outer part of the ECAP die as
shown in Fig 2.2(b), a similar but more complicated strain equation was attained by
Iwahashi et al. in Equation 2.1.
While equation 1 described the imposed strain comparatively close to the real facts
in regular estimations, however, this analysis does not take both fillet radii into account
and hence does not accurately predict the plastic strain in the processed billet when the
fillet radii has significant values. Moreover, the geometry proposed by Iwahashi et al in
Fig.2.2 considers an external radius and the sharp edge in the inner radius of the die, it is
likely that stress concentration will appear.
As an effort to improve the ECAP processing procedure, Luis proposed a new die
geometry as Fig 2.3 (a) shows [14] , with both the internal radius and external radius, the
analysis from Luis concluded the strain imposed with this configuration could be:
)
2 2
sin( )
2
cos(
)
2
sin(
) ( )
2 2
cot( 2
x x
x
x
(2.3)
Calculation using Equation 2.3 with certain constants provided the imposed strain of
this new die setup with different angles and inner fillet radius, as can be seen from
12
Fig.2.4. While the equivalent strain decreases as the angle increases with identical r and
R; given the same angle , the equivalent strain reaches the maximum value when the
inner radius reaches the value of the outer radius. Based on the result above, the
optimized configuration of die geometry is 90 ° with the inner radius r equals to the
outer radius. When r=R then tan(x/2) =0, Eq.(2.3) reduces to Eq.(2.2) in this case.
Luis then proposed this die geometry as described in Fig.2.3 (b), with the optimized
configuration. As can be observed in Fig.2.5, the total equivalent strain predicted by FEM
is generally homogeneous across the cross section with the highest difference observed in
the nodes closed to the inner and outer zones of the die. This indicates that the die
geometry plays important role in affecting the strain.
Similar behavior is observed for traditional ECAP; however, the strain imposed on
different part of the materials across the cross section perpendicular to the pressing
direction is more non-uniform with a significant variation from the inner part of the die to
the outer. Figure 2.6 shows the FEM simulation result on the traditional ECAP die with
an identical angle and different angles, as can be seen in the graphs, the material is
much less homogeneous compared with the one using the new die configuration. Another
advantage of the new die setup is that as long as r=R, the equivalent strain is independent
of the fillet radius, this is true up to a maximum fillet radius after which the main
deformation effect becomes bending instead of shear.
Experimental study on the homogeneity using the conventional die configuration
has been conducted [48]. As shown in Figure 2.7, the hardness distributions become
inhomogeneous after one pass of ECAP and with lower hardness regions in the vicinity
13
(a)
(b)
Figure 2.2.Die geometries (a) Segal’s die assumption [47] (b) Iwahashi’s die assumption
[40].
14
of the bottom surfaces. The distribution becomes more homogeneous with additional
passes through the ECAP die.
One thing has to be pointed out is that the finite element model employed in the
simulation above considers only the non-friction condition or low friction conditions, and
for non-work hardening materials. It is not that difficult to obtain the first one if a proper
lubricant such as MoS
2
is used, but as to the second, as previously introduced, a non-
work-hardening material is generally not true.
However, this simulation work inspired the possibility of achieving a more
homogeneous and efficient ECAP processing procedure using the new die geometry. An
experimental study is tempting using the proposed die geometry considering the
advantages brought by the modification.
2.2.2.3. Pressing routes
While repetitive pressing is possible due to the unique die setup in ECAP
procedures, it was suggested that there is a possibility of introducing different slip
systems in s sample by simply rotating the billets around the pressing axis between each
separate pass. There are reports confirming the final microstructure in the billet can be
different when the material is pressed using several different routes [19.49].
A schematic configuration of four major pressing routes was shown in Fig.2.8. As
described, the billet is continuously pressed without rotation in route A, while in route B
A
the billet is rotated by 90 °in alternative directions at each separate pass, in route B
c
the
15
(a)
(b)
Figure 2.3.(a)(b) Die showing equal fillet radii geometry which can be used in
ECAP,where is the angle of intersection of the two channels ,R is the fillet radius and r
is the inner fillet radius.[14]
16
Figure 2.4. Equivalent plastic strain for different die geometry and angle for L
0=
10 mm
and R= 10 mm. [14]
Figure 2.5. Total equivalent plastic strain with friction modeled by Luis in the proposed
optimized die geometry [14]
17
Figure 2.6. Maximum principal stress distribution with different outer corner angle ECAP
dies at the middle of the deformation step [15].
18
billet is rotated by 90 °in the same direction between each pass and the billet is rotated
by 180 °between each pass in route C.
The effect of the pressing routes to the microstructural evolution was investigated
using Transmission Electron Microscopy (TEM) and route B showed the most rapid
evolution of subgrains into an array of high-angle grain boundaries[51]. Moreover,
additional experiments confirmed that when the die angle is 90 °, a reasonably
homogeneous microstructure and equiaxed grains separated by high-angle grain
boundaries were produced more quickly using route B
C
compared with route B
A
[52]
.
As
a consequence, route B
C
is favorable to produce UFG materials with homogeneous
microstructure in the 90 °die.
2.2.2.4. Number of ECAP passes
The unique advantage of ECAP makes it possible to press same billet for multiple
times. While the strain imposed per pressing is closely connected to the die angle and die
geometry, increasing number of passes surely affects the total imposed strain which
would lead to microctructural evolutions based on several early investigations [46, 51].
While with low passes of ECAP, elongated subgrains or grains as well as a low angle
grain boundary misorientation were found in various materials [46,51], it is confirmed
with TEM and Selected Area Electron Diffraction( SAED) observation that the grains
could be further refined and driven to a high angle misorientation with higher passes of
ECAP. The specific passes of ECAP to achieve a homogeneous equiaxed grain structure
varied with different materials: in pure Aluminum, a minimum of 4 passes was reported
19
to be able to produce such mircostructure. This number becomes ~6 in Al-1%Mg alloy
and ~ 8 in Al-3% Mg alloys [19].
It is also worth of mentioning although an immediate increase of the 0.2% proof
stress could be found after the first pass, more passes of ECAP processing would only
retain the strength instead of further strengthening [49].
2.2.2.5. Pressing temperature
Temperature plays an important role during the processing of ECAP that directly
affect the flow stress of the materials. Generally, a certain material is more difficult to
press at lower temperature, for example, pure aluminum could be pressed up to an
imposed strain of ~20.9 without fracture by continuous pressing using a die having
at ambient temperature while this value decreases to ~13.6 at a lower temperature of
123K [53]. Accordingly, increasing pressing temperature could improve the workability
of the materials, a successful example is CP titanium could only be pressed at 275 ℃ or
above, pressing at lower temperature would result in serious cracks [54].
Nevertheless, elevated pressing temperature has a negative effect on the grain size of
the as-pressed materials, as well as the fraction of low angle grain boundaries. There was
an increase in the equilibrium grain size with increasing temperature as shown in Fig.2.9.
thus further influencing the mechanical properties. A high fraction of low angle grain
boundaries was also found if pressed at higher temperatures.
It is thus concluded that although it is generally experimentally easier to press
specimens at high temperatures, optimum ultrafine-grained microstructures will be
20
Figure 2.7 Contour maps showing the Vickers microhardness distribution in the cross-
section of pure aluminum billets subjected to ECAP for 4 passes via route B
C
using a
solid die with 20 ゜ [48].
21
Figure.2.8 Shearing characteristics for various routes [50]
22
attained when the pressing is performed at the lowest possible temperature where the
pressing operation can be reasonably conducted without the introduction of any
significant cracking in the billets. By maintaining a low pressing temperature, this
ensures the potential of achieving both the smallest possible equilibrium grain size and
the highest fraction of high-angle boundaries.
2.2.3 Microstructural characteristics after ECAP
As an effective processing method to refine grain size and improve mechanical
properties of the materials, ECAP is heavily studied for its effect on the microstructure
and mechanical properties of the materials. Previous paragraph introduced the influence
of various parameters like die angle, pressing temperature, pressing numbers etc on the
performance of ECAP. Although these parameters make ECAP effect a very complex
subject to study, there are several characteristics commonly found in evolved
microstructures independence of materials.
2.2.3.1. Non-equilibrium state of grain boundaries
There have been numerous investigations on the characteristics of the
microstructures introduced by ECAP ever since the invention of this technique. Almost
all of these investigations employ transmission electron microscopy for determination of
the grain sizes produced by ECAP and the nature of dislocation interactions occurring
within the grains.
The large plastic strain imposed on the sample during repetitive ECAP accordingly
influenced the microstructure characterization of the material. Early work [56] on ECAP
23
Figure 2.9. Grain size versus the pressing temperature for pure Al and Al-3% Mg and Al-
3% Mg-0.2% Sc alloys [55].
24
demonstrated the equiaxed granular-shaped structures with sizes in the submicrometer
range, which was later defined as ultra-fine grains with non-equilibrium grain boundaries
[57].In fact, non-equilibrium state of grain boundaries is a common characteristic of
ultrafine-grained materials produced by SPD procedures: High-resolution Electron
Microscopy (HREM) revealed that the grain boundaries of Al-Mg alloys produced with a
wavy or curved nature of non-uniformity and the presence of random arrangements of
facets and steps as well as regions of elastic distortions or missing lattice fringes along
the grain boundaries [58,59]. These kind of irregular grain boundaries in high-energy
configurations were even retained after one hour of annealing of as-pressed Al-Mg-Sc
alloy [60]. While the ultrafine grains of pure metals and solid solution alloys processed
by ECAP are not stable during the annealing treatment, high thermal stability was
observed even at high temperatures in materials like Al-5.5%Mg-2.2%Li-0.12%Zr alloy
[61], or Al-3%Mg-0.2%Sc alloy [41] that contain precipitates which pin the grain
boundaries and retain the grain size in the heat treatment.
2.2.3.2. High angle grain boundary misorientation
The distribution of grain boundary misorientations after ECAP reveals a reasonable
fraction of high-angle grain boundaries but also an excess of low-angle grain boundaries.
It is reasonable to anticipate that the presence of a high fraction of low-angle boundaries,
which is a recurrent feature of the misorientation distributions after ECAP. The large
numbers of excess dislocations introduced on each separate passage through the ECAP
die directly result in this feature, however, the tendency for the average boundary
25
misorientation to increase with the increasing numbers of passes, as shown in Fig.2.10,
provides a unique opportunity to make use of ECAP processing to control the boundary
misorientation, which is specifically important for superplasticity, for which grain
boundary sliding is an important mechanism.
ECAP thus provides an important approach for grain boundary engineering [63]
where it has been proposed that the properties of materials may be effectively changes by
deliberate and careful tailoring of the distributions of boundary angles. The evidence
available to date shows that ECAP processing provides an alternative procedure for
achieving different misorientation distributions because there is high fraction low angle
boundaries after a small number of passes but a lower fraction of low-angle boundaries
when the number of passes is increased [64]. It was successfully used with an Al-3%Mg-
0.2%Sc alloy by processing samples through either 2 passes using route C or 8 passes
using route B
C
, measuring the misorientation distributions, and then using these samples
to measure the interdiffusion characteristics and especially the role of any enhanced grain
boundary diffusion [65].
2.2.4. Mechanical Properties achieved after ECAP
The high dislocation density imposed during ECAP not only influences the
microstructure of the samples, but also consequently affect the mechanical properties.
The small grain sizes and high defect densities in UFG materials processed by ECAP lead
to much higher strengths than in the coarse-grained counterparts.
26
A much higher tensile strength of about 2 times as the one before ECAP was usually
found in the materials subjected to the ECAP processing. At the same time, a much lower
elongation-to-failure was attained and significantly less strain hardening was observed.
Despite the ductility loss, it is important to notice that ECAP processing leads to a
reduction in the ductility which is generally less than in conventional deformation
processing techniques such as rolling, drawing and extrusion.
Fig.2.11 compares the strength and ductility of the Al-3004 alloy subjected to rolling
and ECAP respectively, it is obvious that ECAP retains the ductility better while
strengthen sample by a larger margin.
It was also found that some materials in the as-pressed condition have larger yield
point elongations than in the as-annealed condition [67]. This might be related to the
almost dislocation-free ultrafine grains attained by ECAP that there were insufficient
dislocation sources for a rapid yielding.
As to the mechanical performances at high temperature, while it is well
established that Rachinger grain boundary sliding is a characteristic of high temperature
deformation, it should be noted that the role and nature of Rachinger sliding is different at
large grain size from that at small grain size [68, 69], superplasticity is more likely to
exhibit with small grain size, e.g, less than 1 m.
ECAP is capable of producing ultrafine grains, providing a potential to achieve
superplasticity. Indeed, many reports have been published so far claiming the superplastic
behaviors in various materials [18,19]. Although very large elongations, for example over
1000%, were reported on only part of the materials subjected to ECAP, it is generally
27
Figure. 2.10. Variation of the fraction of high-angle boundaries with the number of passes
through the die for purealuminum using route BC: datum points are shown for the X and
Y planes [62].
Figure 2.11 A comparison of yield strength and ductility for an Al-3004 alloy processed
by cold-rolling or ECAP [66].
28
true that materials exhibit an improved ductility at elevated temperatures due to the
refinement of grain sizes.
As previously discussed, alloying is beneficial to the grain refinement during
ECAP as the recovery rate would be decreased given the solid solution or precipitation
conditions; the magnesium addition by Iwahhashi et al [19] in aluminum alloys caused a
significant decrease in the ultimate grain size. Other chemical elements are needed to be
added into the alloy to retain the non-equilibrium grained structure by ECAP in high
temperature condition. Horita et al [18] demonstrated that an addition of 0.2% scandium
into the Al-3%Mg solid solution alloy achieves an elongation of over 1500% at a high
strain rate through ECAP. This superplasticity performance was credited to the
precipitates, Al
3
Zr, introduced by the addition of scandium. Similar effect of Zr addition
was reported by Furukawa et al. in Al-1420 alloy [70].
Copper and lithium was similar to magnesium in affecting the Al alloy performance,
but nevertheless the roles of these two elements are much more complicated. It is very
likely that the addition of copper and lithium could lead to new precipitates phases, for
example, CuAl
2
[71] . Superplasticity was achieved through ECAP in a commercial Al-
2004 alloy with a chemical composition of Al-6% Cu-0.4%Zr[72].
2.2.5. ECAP on 7000 Series Aluminum Alloys
Despite a great variety of materials subjected to ECAP, aluminum alloys drew
special attention due to its comparatively easy to deform f.c.c structure. However, most
of the ECAP studies focused on Al-Mg alloys [73-75] or Al alloys with less complicated
29
chemical compositions, the studies on Al-Zn-Mg-(Cu) alloys, or 7000 series Al alloys
which are widely used for high strength structural applications such as high performance
sporting goods or aircrafts are limited. Despite the potential of combing the effect of
ECAP and precipitation hardening in these alloys is tempting, it is almost impossible to
press them at room temperatures directly due to their limited deformability.
In other hard-to-press metals or alloys like Ti [76] or Mg alloys [77] , processing by
ECAP at an elevated temperature could effectively improve the workability. And it is
comparatively easy to employ this strategy, there are many reports describing the effect
of ECAP at elevated temperatures to the microstructure as well as mechanical properties
of the Al-7000 series alloys[78,79] However, pressing at elevated temperature could
incur possible grain growth and as to these age-hardenable alloys, new phases as well as
the precipitates formed during the ECAP processing would have additional influence on
the mechanical behavior of the as-pressed Al alloys. Consequently, it is important to
develop a processing routine that could be employed to successfully process Al 7000
alloys by ECAP at room temperature. Successful pressing of these alloys by ECAP at
room temperature was reported by a few researchers employing different processing
routines [80,81] Details of both of the strategies would be described in the following
sections.
2.2.5.1 Pressing Al-7000 alloys by ECAP at elevated temepratures
There have been a few studies on the effect of ECAP at elevated temperatures on
Al-Zn-Mg-Cu alloy, as known as the 7000-series Al alloys [78,79]. Loss in tensile
30
strength at room temperatures were commonly reported , despite an apparent grain
refinment after ECAP processing. The uncommon strength loss was contributed to the
fragmentation of strengthening precipitates, as well as the phase transformation of these
precipitates accelerated by ECAP processing. Details of these studies are discussed below.
The as-received 7034 alloys in Xu’s study [78] was observed with reasonably
equiaxed grains and low dislocation density, the average grain sizes were measured to be
around 2.1 m. Microstructure study was also conducted on the same material after
ECAP processing by ECAP in different passes at 473 K. It is apparent that ECAP
produces an array of ultrafine grains generally in equiaxed shape with 6 or more passes
at this temperature. The average grain size measured in the samples subjected to 6 passes
of ECAP at 473 K was around 300 nm and the SAED pattern suggests that a high fraction
of these boundaries have high angles of misorientations.
The fragmentation of precipitations was reported by various researchers on different
alloys [59]. It was first reported on ’-precipitates in an Al-Cu alloy [82] and later on ’-
precipitates in Al-Mg-Si alloys [83]. As to the Al 7000-serires alloys, similar reports
could be found on MgZn
2
phase in Al-7050 [84] and Al-7034 alloys [78].Xu et al studied
the microstructure change of spray cast Al-7034 alloy by ECAP systematically. The
chemical composition in wt% was Al-11.5%Zn,2.5%Mg,0.9%Cu and 0.2%Zr [78].
Fig.2.12 shows the microstructure of this alloy in the as-received condition. The
presence of fairly large rod-like -phase (MgZn
2
) and an array of very fine particles
which was identified as primarily ’-phase but also Al
3
Zr particles. A careful
measurement showed the large rod-shaped MgZn
2
particles had lengths of ~0.48 m and
31
widths of ~0.07 m, but the very fine particles had dimensions of around 10 nm. This is
considered as typical microstructures of 7000-series aluminum alloys containing
additions of Zn, Mg and Zr [85,86].
Fig.2.13 (a) and (b) show microstructures of Al-7034 alloys after one and two passes
of ECAP at 473 K, respectively. It is important to notice that the large rod-shaped
precipitates have become fragmented by the high pressure imposed in ECAP while the
very fine particles of Al
3
Zr appeared to be unaffected by the ECAP processing. There
were also fine particles having size in the range of ~30-100 nm which was identified as
the phase. Although many of the more irregular particles were formed by
fragmentation of the larger precipitates, it was reasonable to assume that a large portion
of the smaller MgZn
2
particles were formed through a direct transformation of the ’-
phase into the phase with subsequent coarsening. The differential scanning
calorimetry (DSC) study later confirmed the assumption.
Xu et al. conducted tensile tests at ambient temperature and high temperature
respectively [78]. As to ambient temperature tests, strain rate of 3.3 x 10
-4
s
-1
was chosen.
As shown in the plot in Fig.2.14 (a), the as-pressed samples exhibit very short regions of
strain hardening, low values of ultimate tensile strengths, longer regions of strain
softening and reductions in the total elongation to failure. This behavior is different from
conventional alloys processed by ECAP where there is generally a significant increase in
strength after ECAP processing. As discussed in the microstructure part earlier, there is
fragmentation of the rod-shaped -phase precipitates; however, this could not be the
reason for the strength loss. Xu et al contributed the strength loss to the dissolution or
32
transformation of the metastable hardening ’-phase during ECAP at 473 K, which is
known to occur in 7000-series aluminum alloys [87].
High temperature tests were conducted at 673 K; the results are plotted in Fig.2.14
(b). Exceptionally high superplastic elongations, up to ~1100% are attained in the 6-pass
and 8-pass samples whereas the maximum elongations are less than 650% in other
samples. As shown in the microstructure study, all of the samples processed by ECAP
achieved a reasonably fine grain size, which cannot be attributed to the elongation
changes. Thus, grain boundary misorientations should be responsible for the plastic
behaviors. Several reports have demonstrated that the fraction of high-angle boundaries
increases with increasing numbers of passes in ECAP [62,88].
2.2.5.2.Pressing Al-7000 alloys by ECAP at room temepratures
Althoug it is feasible to press Al-7000 alloys at elevated temepratures, However,
this processing strategy could incur possible grain growth and as to these age-hardneable
alloys, new phases as well as the precipitates formed during the ECAP processing would
have additional influence on the mechanical behaviors of the as-pressed Al alloys. To
avoid these side effects , it is of great imprtance to develop some new processing routines
to press Al-7000 alloys by ECAP at room temperature.
Zhao et al.[80] and Chinh et al[81] conducated researches on processing Al-7000
alloys by ECAP at room temperature respectively. Acoording to a series of researches
from their reoprts, it is possible to press these alloys following a certain processing
routine: solid solution tretment , water quenching immediately followed by pressing by
33
Figure 2.12 Microstructure in the as-received Al-7034 alloy showing the presence of rod-
shaped MgZn
2
precipitates [78].
Figure 2.13.Microstructureof Al-7034 after 1 pass(a) and 2 passes (b) of ECAP at 473
K.[78]
34
ECAP. The specified processing parameters like the solid solution tretment temperature,
duration are dependent on the specific materials.
The ECAP routine employed by Zhao et al.[80] for commercial 7075 Al alloy
was to treat the rods at 480 ℃ for 5 h and then quenched to room temperature. The ECAP
processing was carried out immediately after quenching for 2 passes following route B
C
.
The grain size of this alloy decreased drasctically from 40 m to elongated grains with an
average width of about 150 ±20 nm and a length of about 430±30 nm. The grain
boundries of the UFG samples are wavy and diffuse suggesting high strain and non-
equilibrium boundary configurations. Systmetric Differential Scanning Calorimetry (DSC)
and X-ray Difffraction (XRD) investigations were also conducted, the results suggested
ECAP could accelerate the precipitation as well as prcipitates coarsening,however, it
would not change the sequence of the precipitates phase transformation, i.e. from
supersatureated solid solution to Guimier-Preston (G-P) Zones, then metastable
’(MgZn2) , finally stable (MgZn2) .
Similar approaches were applied to other Al 7000 series alloys by Chinh et al [81],
with the solid solution temperature of around 470 ℃ and duration of around 30 minutes.
Billets subjected up to 7 passes of ECAP processing at room temperature was achieved.
Grain size measurement of the sample underwent 4 passes of ECAP at room temperature
concluded an average grain size of around 300 nm was achieved. This is much smaller
than the grain size of ~500 nm in the same materials subjected to 8 passes of ECAP at
473 K. The dislocation density measured from XRD profiles is also substantially higher
35
(a)
(b)
Figure.2.14 Engineering Strain versus Engineering Stress at test (a) at ambient
temperature and (b) 673 K [78].
36
with a value of (6.5±0.7)×10
14
m
2
compared with (3.2±0.4)×10
14
m
2
in the same materials
subjected to 8 passes of ECAP at 473 K.
Mechanical property tests were conducted on the samples processed by ECAP at
room temperature. The engineering stress-strain curves of the UFG and CG sampples
from Zhao et al was plotted in Figure.2.15 The tensile yield strength and ultimate
strength of the UFG sample are 650 and 720 Mpa, respectively,which are about 103%
and 35% higher, respectively, than those of the CG sample. The elongation to failure of
the UFG sample is smaller than that of the CG sample.
The Vickers hardness value (HV) as a function of aging time for two different Al
7000 samples ( AA2 and AA3) processed by 1 passes in ECAP at RT from Chinh’s
results was shown Figure 2.16, the hardness values obtained on the quenched and
naturally-aged samples without ECAP are also plotted for comparison. It could be
concluded that even 1 pass of ECAP at room temperature could strengthen the Al 7000
alloys substantially. An increase of the hardness was found both in AA2 and AA3
without ECAP with the increase of the natural ageing time up to 1 year, when a saturated
HV was achieved. The corresponding saturation HV values of the lower concentration
AA2 alloy and the higher concentration AA3 alloy are ~85 and ~185, respectively. After
processed by ECAP of 1 pass at RT, the hardness of AA2 and AA3 underwent different
pre-aging duration prior to ECAP was reasonably same.
Figure 2.17 shows the Vickers hardness value, as a function of the number of passes
in ECAP for AA2 and AA3 samples pre-aged for the shortest time of 10 minutes, there
was no continuous strengthening with more passes of ECAP at room temperature.
37
2.3. High pressure torsion
While the origin of processing by HPT could be traced back to the idea of combining
torsion with compression in the paper by Prof.Bridgeman from Harvard University in
1943 [89], researches on HPT were mostly conducted in the last two decades, the
principles of modern HPT are depicted schematically in Fig.2.18. The sample disk is
located between two anvils where it is subjected to a compressive applied pressure,P, of
several GPa at room temperature or an elevated temperature and simultaneously it is
subjected to a torsional strain which is imposed through rotation of the lower anvil.
Surface frictional forces therefore deform the disk by shear so that deformation proceeds
under a quasi-hydrostatic pressure.
There are practically two distinct types of processing: unconstrained and
constrained HPT [29]. In unconstrained HPT, the specimen is placed on the lower anvil
and then it is subjected to an applied pressure and torsional straining. Since the sample is
not constrained, it will flow outwards partially which may incur only a partial back-
pressure during the processing. This disadvantage could be avoided by introducing the
constrained HPT, which is shown in Fig.2.19 (b), the sample is machined to fit into a
cavity in the lower anvil and the load is applied such that there is no outward flow of
material during the torsional straining. It is supposed that true constrained HPT is
conducted in the presence of an effective back-pressure, however, in practice, it is much
easier to set up a quasi-constrained HPT as shown in Fig.2.19 (c) rather than the
idealized constrained one, there will still be limited flow outwards between the upper and
lower anvils.
38
Figure. 2.15. Tensile engineering stress–strain curves for the UFG (solid circle) and CG
(empty circle) samples, both of which were naturally aged for one month.[80]
Figure.2.16 .The values of the Vickers microhardness as a function of pre-aging time for
the samples processed by only 1 pass in ECAP.[81]
39
Figure.2.17 .The values of the Vickers microhardness as a function of the number of
passes in ECAP for samples pre-aged for 10 minutes [81].
40
2.3.1.The Fundamental parameters in processing by HPT
2.3.1.1.Definition of the strain imposed in HPT
As depicted earlier, during HPT processing, the sample, in the form of a disk, is
located between two anvils where it is subjected to a compressive applied pressure, P, of
several GPa at room temperature or at an elevated temperature and simultaneously it is
subjected to a torsional strain which is imposed through rotation of the lower anvil.
Surface frictional forces therefore deform the disk by shear so that deformation proceeds
under a quasi-hydrostatic pressure.
For an infinitely small rotation, d , and a displacement, dl, it follows from Fig.2.20
that dl = r*d where r is the radius of the disk, and the incremental shear strain, d , is
then given by [28]
(2.4)
where h is the disk thickness. By further assuming that the thickness of the disk is
independent of the rotation angle, h, it follows from formal integration that, since h =
2 N, the shear strain, , is given by
(2.5)
where N is the number of revolutions. Finally, in many investigations the equivalent von
Mises strain is then calculated using the relationship [92-94]
(2.6)
41
The use of Eq. (2.6) is correct for small imposed shear strains but for large strains, where
≥0.8, the equivalent strain is given by [93]
(2.7)
In practice, the preceding equations provide alternative relationships which may
be used to estimate the total strains imposed on disks subjected to HPT. Interestingly,
calculation results show that although there is a variation of the imposed strain correlated
to the position on the disk, i.e., the imposed strain is larger at the peripheral part than the
center part , the variation of the accumulated imposed strain is not that significant after a
few turns of HPT. In fact, even for r =0.1 mm, which corresponds to a point in the
immediate vicinity of the disk center, the accumulated strain after 5 whole revolutions is
equal to approximately 4 and this is only lower by a factor of about 2.5 than the strain
estimated at the periphery.
2.3.1.2. Homogeneity across the HPT disk
The nature of torsion decides that the imposed strain varies across the sample and
that value decreases to zero in the center theoretically during HPT processing. It is thus
reasonable to anticipate that the microstructure produced by HPT will be extremely
inhomogeneous. Despite a variation of the homogeneity was found in most of the
materials after low turns of HPT, it was also suggested that this variation could
eventually vanished with increasing the number of turns in many materials according to
the experimental data up to date. Much of the experimental information on the presence
or absence of homogeneity is obtained most conveniently by taking measurements of the
42
Figure.2.18. Scehmatic Illustration of High Pressure Torsion(HPT) [28]
Figure. 2.19. Schematic illustration of HPT for (a) unconstrained and (b, c) constrained
conditions [29].
43
Figure. 2.20. Parameters used in estimating the total strain in HPT [28].
44
local microhardness, and then correlating a selected set of these hardness values with
microstructural observations undertaken using transmission electron microscopy (TEM).
The experimental results indicate a possible dichotomy. In the investigations on
austenitic steel[94],Cu[95] and high-purity Ni [96], significant variations in the values of
the microhardness across the diameters of disks were found, with lower hardness values
in the center of the disks and higher values in the peripheral regions; while the results on
commercially purity Al [29], an Al-Mg-Sr alloy[23], Cu[25] and high purity Ni [97]
showing the microstructures appear to become reasonably homogeneous across the disks
when torsional straining is continued to a sufficiently high total strain under a high
imposed pressure.
TEM investigations were employed in most of the studies on the microstructure
evolution with HPT processing. According to the results available to date, it is generally
expected to see a transition from relatively larger subgrains in the central region to
smaller grains having high-angle boundaries in the peripheral region after HPT
processing with low pressure and/or low turns. The bright field TEM images and
corresponding selected area electron diffraction ( SAED) patterns shown in Fig.2.21
revealed the microstructure of commercially purity Nickel after HPT. A higher
microstructural refinement is found near the edge of the disk and the mean grain size in
this region was measured ~0.17 m [97].
Homogeneity evolution is also closely correlated to the applied pressure. The
hardness results of high purity Ni after HPT processing under different values of the
applied pressure ,as shown in Fig.2.22, while all of the microhardness values are higher
45
than for the unprocessed nickel, with lower applied pressure of 1 GPa there is a non-
uniform distribution of microhardness values across the sample diameter with
significantly lower values in the center of the disk whereas at the higher pressure of 9
GPa, there is a minor increase in Hv at the edge of the disk but a significant increase in
the center so that the microhardness values become more uniform. In short, a better
homogeneity could be achieved with increasing the applied pressure. Thus, it could be
concluded that the processing parameters of strain (in the form of numbers of turns) and
applied pressure are both very critical to the homogeneity of the materials.
2.3.1.3. Microstructure Evolution
Other than the inhomogeneous grain refinement revealed in the bright field TEM
images shown in Fig.2.23., the Selected Area Electron Diffraction (SAED) pattern
contains many spots situated around circles thereby indicating the presence of boundaries
having high misorientation angles in the edge region. While the average grain size in the
center of the disk in is about 2 times larger than the grain size at the edge, the SAED
patterns consists of separate spots showing the presence of a large fraction of
boundaries having low angles of misorientation. When the same sample is subjected
HPT at 6 GPa, the microstructures for this condition were reasonably similar at both the
center and the edge as anticipated from the hardness data in Fig.2.22 and the SAED
pattern consists of rings with many diffracted beams indicating the presence of many
small grains with multiple orientations within the selected field of view.
46
2.3.1.4. Model for the development of homogeneity in HPT
Ideally, the shear starts at a point with the largest frication force, which is correlated
to the local frictional coefficients and the level of the applied pressure. Assume the
applied pressure on the disk is homogeneous regardless of the specific position of the
point, then it is only the friction coefficient , which is related to the local surface
roughness of the disk and/or that of the upper anvil in the HPT processing facility , that
decides the starting point of shearing. Once it starts, the induced local hardening would
consequently reduce the frictional force and transfer the shearing to an adjacent point.
Practically, the shearing extends inwards to the interior with the repetitive
position alterations. A generally acknowledged fact is that reasonably homogeneous
microstructures may be attained throughout the samples when the applied pressure P and
the number of revolutions N are both sufficiently high. There are two different models of
strain gradient plasticity explaining the microstructural evolution and the development of
a homogeneous structure in HPT [98]. The materials is considered to be in the form of a
two-phase composite, i.e., the dislocation cell walls where dynamic recovery occurs by
climb processes and the cell interiors where dynamic recovery is controlled by cross-slip
of the dislocations.
The change of the dislocation density within the cell interiors,
c
as a function of
time t is given as [99,100] :
(2.8)
47
Figure. 2.21. Microstructures of HPT nickel for (a) the central part of the disk after 5
turns at 1 GPa, (b) the edge of the disk after 5 turns at 1 GPa the SAED patterns were
taken with an aperture size of 1.8 m [97].
Figure.2.22. Microhardness profiles of nickel processed by HPT at two different applied
pressures[97]
48
Where A.B and C are constants which are capable of measurement within the HPT
processing, ἐ
p
is the equivalent von Mises plastic strain rate,
w
is the dislocation density
in the cell walls, d
⊥
is the average dislocation cell size, f is the volume fraction of cell
walls, is a constant having the same dimensions as strain rate and n
cs
is the recovery
exponent for cross-slip which is generally taken as a liner function of the stacking fault
energy.
The first term in equ.2.8 denoted the increase in the dislocation density in the cell
walls and cell interiors due to the activation of Frank-Read sources in the walls, the
second term denotes the decrease in the dislocation density in the cell interiors and the
last term corresponds to the effect of dynamic recovery and annihilation of dislocations
by cross-slip at large strains. While the first two terms are more correlated to the
processing procedure, the last one depends more on the nature of the materials, i.e., a
material with a high stacking fault energy such as high purity Al will incorporate an
outstanding influence on the rate and magnitude of the recovery process.
Fig.2.24 shows the distribution of accumulated plastic strain along the specimen
radius for different numbers of turns in HPT [98].These results clearly demonstrate that
reasonably homogeneous microstructures may be attained throughout the samples,
including in the central region, when the applied pressure P and the number of
revolutions N are both sufficiently high.
49
2.3.1.5. The significance of the minimum grain size attained using HPT
While HPT has been proved to be an effective tool in reducing the grain size of
the materials, there is usually a minimum grain size that HPT processing could achieve,
this specific minimum grain size is dependent on the material. For pure Cu, Fig.2.25
shows representative TEM and HREM micrographs taken from the edge of disks by HPT
[101]. It is apparent that HPT produces equiaxed grains with reasonably random
orientations. These measured average grain size correspond to the minimum possible
grain size, d
min.
The experimental data in this study indicate that the stacking fault energy
significantly affects the measured value of d
min
such that a decrease in energy leads to a
corresponding decrease in the grain size in the order from copper to bronze to brass after
processing by HPT under the same experimental condition. This trend is reasonable if the
minimum grain size is determined through the development of a dynamic balance
between dislocation generation and dislocation recovery. Thus, in materials with lower
stacking fault energy it becomes increasingly difficult for the dissociated dislocations to
recombine and cross-slip, and this impedes the recovery process and leads to a smaller
value for d
min
.
A theoretical model has also been developed to anticipate the minimum grain size
[102], a relationship between d
min
and the value of the stacking fault energy,
(2.9)
50
Figure.2.23.the edge of the disk after 5 turns at 6 GPa: the SAED patterns were taken
with an aperture size of 1.8 m[97].
51
Where A
1
is a constant. b is equal to the length of the unit dislocation along the <110>
direction for f.c.c structures such as copper-based alloys and thus it is given by / 2
where is the lattice parameter. Using appropriate values of b,t and G,Fig.26 shows a
plot of d
min
/b versus /Gb on a logarithmic scale using data obtained from TEM and XRD
[105]. If it is reasonably assumed that the apparent nonlinearity is due to a scatter in the
experimental data, it is possible to fit the data with straight lines as shown in Fig.2.26 to
yield slopes of 1.37 and 1.07 from the TEM and XRD measurements of grain size,
respectively. It is readily apparent that each of these slopes is significantly higher than the
slope of 0.5 predicted in the model and documented in Eq.2.9 . This conflicts should be
contributed to the model assumption that only a single deformation mechanism occurs,
yet other than the lattice dislocation slip existed in the copper deformation, twinning is
proven to be another important factor in the deformation of Cu-30wt% Zn alloy; even in
the same material, the deformation mechanism may change that the grain size becomes
smaller than a critical value.
It is possible to achieve very fine grains in Al and Al alloys. It is interesting to
note that the HPT processing of pure aluminum was first documented only very recently
when microstructural evolution was investigated in the HPT processing of commercial
purity aluminum as a function of the accumulated strain [24,29]. Fig.2.27 shows
examples of the microstructures in (a) the initial annealed condition and (b) after
application of a load of1 GPa but without torsional straining so that N = 0 [29]. Optical
microscopy revealed an initial grain size of 0.2–0.5 mm in Fig.2.27(a) but the grain size
was reduced to 3 m immediately upon application of the load and without any torsional
52
Figure.2.24.Accumulated equivalent strain versus distance from the specimen center in
the first-order strain gradient model [98].
53
straining as shown in Fig. 2.27(b). The micrographs in Fig.2.28 were taken on the same
material and they show the microstructures (a) in the central region and (b) near the
periphery after 2 whole revolutions. It has also been proved that HPT is an effectively
tool to substantially refine the grains to the size of ~100-300 nm or even smaller in most
of the Al alloys. The measured grain size is less than 100 nm in the commercial V96Z1
alloy Al (7.5% Zn–2.7% Mg–2.3% Cu –0.15% Zr) after HPT processing [103], similarly
a ~90 nm grain size was found in Al-3%Mg alloy [58].
2.3.2. Microstructural development in more complex metallic alloys
While the general microstructural development also occurred in the more complex
metallic alloys, there are something unique to be discovered. In a cast Al–11 wt.% Fe
alloy after HPT straining, other than the microstructural refinement, there is evidence that
a homogeneous, distribution of small (<1 m) second-phase particles formed on the
aluminum-based matrix [104]. A supersaturated solid solution of iron in the aluminum
a maximum solubility of 2.2 wt.% was attained and this provided matrix with the
opportunity to further strengthening the alloys by consequent heat treatments which
usually is invalid to this conventionally non-hardenable alloys.
The microhardness of the alloy was dramatically increased to 1750 MPa from 750
MPa in the as-cast alloy by HPT processing. Subsequent artificial aging at 373 K led to
an additional increase in the microhardness up to 3020 MPa due to a general
decomposition of the supersaturated solid solution and the occurrence of precipitation
strengthening. Detailed TEM studies reveals that well-defined yet curved and wavy grain
54
boundaries as well as poorly delineated grain boundaries both existed in the Al-Fe alloys
after HPT processing; a high level of internal stresses within the grains was also found.
Since this contrast is present both within grains containing dislocations and within grains
where no visible dislocations are present, it is concluded that the source of these internal
stresses is associated with the presence of defects within the grain boundaries.
Another important feature of the as-processed alloys is that the dendritic structure
of the castings was eliminated by HPT, which implies the enhanced solubility due to SPD
processing. An enhanced solubility of the second-phase during HPT straining was
observed also in a commercial 01959 aluminum alloy (Al–6% Zn–2.8% Mg–1% Cu–
0.37% Zr) [105]
In the more complex material systems, despite a lot in common, the homogeneity
development is slightly different from that in metals or simple alloys. For example,
Fig.2.29 shows optical microstructures in HPT disks of an austenitic steel strained
through either (a) 2.27 or (b) 16 whole revolutions [92]. It is readily apparent that, at
least after 16 revolutions, a significant refinement of the microstructure has taken place in
the immediate vicinity of the center of the disk. This same conclusion is reached also
from Fig.2.30 which depicts the microhardness profiles as a function of a normalized
distance from the center of the HPT disk. These data suggest the occurrence of some
grain refinement in the central region although it is also apparent that this refinement
develops more slowly than in pure nickel [97]. One reason for the a slower refinement in
austenitic steel may lie in the additional strain energy that is needed for the phase
transformations taking place in multi-phase materials [106-109].
55
Figure.2.25. Bright-field TEM micrographs of copper processed by HPT (a) at the edge
of the disk and (b) showing deformation twins [102].
Figure.2.26. Plot of normalized minimum grain size versus normalized stacking fault
energy for Cu and two Cu–Zn alloys showing different slopes from experimental data
obtained using TEM and XRD [102].
56
Although all of these proposals are reasonable and may occur in HPT processing,
nevertheless a careful inspection of the experimental results reported in this earlier
investigation suggests a possible alternative interpretation. For example, it was shown
that HPT straining induces the occurrence of a martensitic transformation and about 80%
of austenite was transformed into martensite [107]. Similarly, a stress-induced phase
transformation was experimentally detected in nanocrystalline ferrite as a consequence of
HPT processing at room temperature [106].
2.3.3. Combing processing by HPT with other processing methods.
As HPT and ECAP are both effective Severe plastic deformation method for grain
refinement, the possibility of combining these two processing procedures thus becomes
tempting in further refining the materials. In fact, it is also possible to combine HPT with
other conventional deformation methods, like rolling or drawing to optimize the physical
and mechanical properties of the material. There are several reports describing the
application of HPT and other processing techniques to different materials including Cu
[110-112] and Ni [113-115].
An investigation using high purity Ni made direct comparisons between the
microstructures produced with combinations of HPT, ECAP and cold-rolling(CR) with
the materials examined using XRD and TEM [114]. For ECAP the material was
processed at room temperature to produce an as-pressed average grain size of 350 nm
[116]. For HPT the material was strained at room temperature under an applied pressure
of 6 GPa for 5 revolutions to produce an average grain size of 170 nm [97].
57
Figure.2.27. Microstructures of aluminum: (a) optical microscopy showing the initial
annealed structure and (b) TEM micrograph showing the structure after applying a
pressure of P = 1 GPa without torsional straining [29].
Figure.2.28 Representative TEM micrographs of commercial purity aluminum after HPT
with a pressure of P = 1 GPa (a) in the central region and (b) near the periphery after two
whole revolutions [29].
58
Additional samples were prepared by various combinations: (i) ECAP + CR, (ii)
ECAP + HPT and (iii) ECAP + CR + HPT. The CR of the ECAP specimens was
performed at room temperature with a reduction in thickness from 1.7 to 0.25 mm,
equivalent to an overall reduction of 85%. Fig.2.31 shows TEM micrographs for the Ni
specimens produced by (a) ECAP, (b) ECAP + CR, (c) HPT, (d) ECAP + HPT and (e)
ECAP + CR + HPT [114]. Besides the typical non-equilibrium grain boundaries with
grain interiors having complex contrasts under SPD processing, it is also apparent that the
combinations of these various techniques could produce reasonably homogeneous
microstructures as shown in the micrograph. Additional processing was proven to be
useful in further reducing the grain size: the measured grain sizes were 140 nm after
ECAP + HPT in Fig.2.31d and 100 nm after ECAP + CR + HPT in Fig.2.31e.
Another report on the nickel samples processed using ECAP, HPT and a
combination of ECAP and HPT [113] showed that a combination of ECAP and HPT also
plotted in the Fig.2.32, the grain boundary misorientation distributions for all three
processing conditions exhibit a bimodal character with clearly defined peaks within both
the low and high-angle ranges. The magnitude of the peaks at the low angles led to a
smaller fraction of the grain boundaries having low angles of misorientation. As
decreases in the order from ECAP to HPT to ECAP+HPT , which is the same order that
the measured mean grain sizes decreased. Although the presence of an excess of low-
angle boundaries after ECAP + HPT in Fig.2.32c is not consistent with the anticipated
theoretical random distribution,nevertheless it is concluded that this combined processing
59
Figure. 2.29. Optical micrographs of the central region of a disk of an austenitic steel
processed by HPT (a) after N = 2.27 revolutions and (b) after N = 16 revolutions [97].
60
Figure.2.30. Microhardness profiles for six samples of an austenitic steel processed by
HPT through different numbers of revolutions [97].
61
Figure.2.31. Microstructure of nickel processed using different procedures: (a) ECAP, (b)
ECAP + CR, (c) HPT, (d) ECAP + HPT and (e) ECAP + CR + HPT [114].
62
route produces a distribution of misorientations which most closely approximates to the
random distribution.
2.3.4. Materials softening during HPT
While generally HPT is believed to be an effective processing method to
strengthen the materials due to the significant grain refinement capability, there are some
exceptions. Reports are available that excessive HPT processing, which would lead to an
extremely high dislocation density, could lead to dynamic recovery or dynamic
recrystallization that would not lead to further grain refinement but larger grain sizes in
different materials. Additionally, when the grain sizes of the materials subjected to HPT
processing are already very small, usually to the nanocrystalline range, HPT processing
would also be invalid to strengthen the materials, instead, softening behavior was
observed in electrodeposited nanocrystalline nickel as shown in Figure.2.33 [117] .
The reason for the softening in ED NC nickels was also due to the grain
coarsening, TEM study on the ED NC Nickel disk after HPT processing was revealed in
Figure 2.34. The high purity Nickel( ≥99.7%) was with an average grain size of ~22 nm
and a narrow grain size distribution before HPT processing. After processing by HPT
under a hydrostatic pressure of 6.2 GPa for five turns, there is an obvious grain
coarsening observed in the disc as shown in Fig.2.34. This effect is pronounced at the
edge, where the shear strain is larger (for a radius r = 3 mm) compared with that at the
center (c < 20 taking r = 0.5 mm). All these materials softening behaviors during HPT
processing mentioned above could be explained using the Hall-Petch equation, where a
63
larger grain size corresponds to a lower strength. However, this is not the only category
of softening occurred during HPT processing. A series studies by Mazilkin et al. [118] on
the effect of HPT on Al-Zn, Al-Mg and Al-Zn-Mg alloys revealed that despite an
apparent grain refinement from the HPT processing , the materials were rather softened
than strengthened. HPT processing was found to be effective in strongly decreasing the
grain size, which also led the materials to be further away from the equilibrium state than
the initial CG material.
At the same time, during HPT, the Zn- and Mg-rich supersaturated solid solutions
decompose and approach the phase equilibrium state corresponding to room temperature.
This decomposition accelerated by HPT processing is mostly likely because HPT would
introduce fluxes of vacancies that would assist the GB diffusion, and will bring the
material closer to the equilibrium state than that of the initial CG structure. Consequently,
this decomposition would incur materials softening after HPT processing, which is
claimed to be more pronounced than the Hall-Petch hardening due to the decreasing grain
size as well as work hardening. As a net result, the Al-Zn, Al-Mg and Al-Zn-Mg alloys
were softened after room temperature HPT processing.
64
Fig.2.32. The grain boundary misorientation distributions in nickel after (a) ECAP, (b)
HPT and (c) ECAP + HPT [114].
65
Figure 2.33. Microhardness of ED NC Ni and ED NC Ni + HPT. The microhardness of
ED NC Ni + HPT is lower than that of as-deposited ED NC Ni, and it decreases with
increasing shear strains. [117]
Figure 2.34. Scanning electron micrographs of ED NC Ni + HPT illustrating the grain
coarsening after SPD. (a) Definition of the direction of observations; (b) micrograph near
the center in the axial direction; (c) micrograph at a radius of about 3 mm in the axial
direction; (d) micrograph at a radius of about 3 mm in the radial direction [117].
66
3. EXPERIMENTAL MATERIALS AND PROCEDURES
3.1 Experimental materials
The commercial Al-7136 alloy was used for the inspection of the effect of ECAP
on the microstructure and mechanical properties of supersaturated aluminum alloy in this
study. The chemical composition of the as-received alloy is : Al 85 wt%, Zn 9.4 wt%,
Mg 2.5wt%, Cu 2.5wt%, Zr 0.2 wt% and Fe 0.15 wt% . The material also included
impurities such as: Ti <0.1 wt%, Cr <0.05 wt%, Mn <0.05%. The alloy was provided by
QED Inc. San Diego,CA,USA in the form of extruded rods and were cut into billets with
length of 64 mm to be processed by ECAP.
3.2. Experimental facilities, analytical methods and sample preparation
3.2.1. ECAP
3.2.1.1. ECAP die and plunger
An ECAP facility is composed of a die having two channels, a plunger and a
pressing machine including some fixture parts. The design of an ECAP die and a plunger
is one of the most important subjects in order to prevent any problems where the
continuous pressing is conducted at room temperature and at elevated temperatures.The
ECAP die was used to press commercial Al-7136 alloy following various processing
routines to investigate the effect of ECAP on the microstructure and mechanical
properties of the supersaturated Al alloys.
67
The drawings of the first set of the ECAP die and the plunger used in this study are
shown in Fig.3.1.(a) (b), respectively. For use at elevated temperatures, the die is made of
a tool-steel having high hardenability, excellent wear resistance and hot toughness and it
is subsequently followed by an appropriate heat treatment . The solid die has an angle of
with an outer arc of curvature of Thus, from the theoretical calculation
given in equ.1, this design of the die can impose a strain of ~1 into a billet in each
separate pressing process. Practically, the diameter of the channel subsequent to the
channel-direction is engineered to be slightly smaller than that of the other channel so
that there is a reduction of elastic expansion when a billet is pressed through the shear
plane at the channel-intersection. In addition, at the channel exit where a sample is
removed from the die there is a slightly larger diameter than the other parts of two
channels to make the billet pass smoothly and with less friction.
In order to monitor the pressing temperature in the die during the ECAP procedure
at elevated temperatures, a thermocouple was inserted to the position close to the
channel-intersection of the two channels. The plunger was also made of the same tool-
steel and given the same appropriate heat-treatment. The dimension of the plunger is
designed correspond to the channel entrance of the die. The plunger inserted into the die
has a length extending from the channel entrance to the intersection of the two channels.
3.2.1.2. Pressing Facility
A schematic illustration of the ECAP facility used in this study is shown in Fig.3.2.
The die and plunger were installed in the facility and supported using the die
68
(a)
(b)
Figure 3.1.The design of (a) the ECAP die and (b) the ECAP plunger.
69
supporter and a plunger cap, respectively, with the accurate alignment of each other. The
facility was designed to move the lower platen with the die upwards and downwards in
pressing a billet and insert the billet for continuous pressing. The lower platen was
covered by a refractory material and it was heated for pressing at elevated temperature
when necessary monitored by a thermometer. Two different lubricants were used for the
pressing: Moly Assembly Oil 150 produced by Sumico Lubricant Co.Ltd. for the pressing
at room temperature and Crown 9105 Anti-Seize Compound produced by Crown Oil UK
for pressing at elevated temperatures. For the elevated temperature pressing, a sample
was kept in the die for 4-5 minutes before pressing so that the temperature of the sample
is consistent with the desired pressing condition.
3.2.2. High Pressure Torsion
The High Pressure Torsion was conducted on Al-7136. The as-received Al-7136
rods with the diameter of ~10 mm were cut to a length of ~60 mm. After that, the rod was
sliced into disks having thickness of ~1.5-2.0 mm. Both sides of these disks were
polished using abrasive papers without cloth polishing to give a series of HPT samples
having a total thickness of ~0.8 mm.
Processing by HPT was conducted at room temperature using an HPT facility
similar to that shown in Fig.3.3. with upper and lower anvils made from high-strength
YXR3 tool steel and having nitride surfaces.Each anvil was machined with a spherical
depression at the center of the adjacent surfaces with a depth of 0.25 mm and a diameter
of 10 mm. A lubricant containing MoS
2
was placed around the periphery of each depre-
70
Figure 3.2 Schematic drawing of the pressing facility.
71
ssion on both the upper and lower anvils, the sample disk was placed in the depression on
the lower anvil and the lower anvil was brought into position so that the disk was
contained within the depressions on the two anvil surfaces.
A controlled applied load could be exerted by the hydraulic pressing machine with
the range from 1.6 MPa to 14.0 MPa. Since the diameter of the hydraulic cylinder is
250 mm. The accordingly force could be calculated from:
F= ×125
2
×P
h
(3.1)
Where P
h
is the hydraulic pressure, F is the applied force. This force then could be
transferred to the pressure applied on the 10-mm-diameter disk, P
d
, using :
P
d
= F / ( × 5
2
) (3.2)
An imposed pressure of 6 GPa on the disk is desired in this research. Using these
equations, one could then calculate that this corresponds to a hydraulic pressure of 49 ton
force. Torsional straining was applied by rotating the lower anvil with respect to the
upper anvil at a constant rotation speed of 1 rpm, All operations were conducted at room
temperature.
3.2.3. Mechanical testing
3.2.3.1. Tensile testing
An Instron machine was used for tensile testing in the present work. The tensile
machine included a single-zone resistance furnance and a load-cell connected to a digital
indicator so that the tensile tests could be conducted using accurate initial strain rates
72
over a wide range of temperatures. A wide range of strain rates was available for use for
the tensile testing since the ratio of the motion of a gear representing the strain rates could
be changed to various values. Constant cross-head speeds were used during the testing
regardless of the initial strain rate. Each tensile specimen was tightened by grips to a jig
and subsequently a thermocouple was set close to the gauge length of the specimen
without any contact in order to check and maintain the consistent desired testing
temperature within ±2 K throughout the tensile testing. In practice, the testing was
conducted 30 or more minutes after the desired temperature was achieved to stabilize the
temperature in the furnace.
A chart-recorder connected to the Instron machine recorded instantaneous loads
against sample elongations by the movement of chart-paper with selected speed during
the tensile testing. The chart-recorded expressed a plot of load in pounds versus
displacement in inches with desired ratios of ×10 or ×20 to the original displacement and
consequently the plot was converted manually to a plot using terms of stress and strain.
The engineering tensile stress and the elongation to failure of the engineering strain (EF)
in % were calculated as follows:
2
0
() 1
()
()145
Plb
MPa
Ain
(3.3)
0
00
(%) 100 100
LL L
EF
LL
(3.4)
where P is the instantaneous applied load, A
0
is the original cross-sectional area of the
specimen, L is the total length of the tensile specimen after testing, L
0
is the total length
73
of the tensile specimen having the original gauge length and △L is the displacement of
the tensile specimen. However, due to considerations of the elongation of only the gauge
length, a modification of the elongation of grip parts was undertaken if necessary.
All tensile specimens used in this study had the dimensions shown in Fig 3.4. It
was designed to have a gauge length of 4 mm and a cross-sectional area of 3 × 2 mm
2
.
The tensile specimens produced from the ECAP billets were machined to have a tensile
direction which is same as the pressing direction.
3.2.3.2. Vickers hardness test
The value of the Vickers micohardness were recorded on each disk processed by
HPT as well as those sliced from the ECAP processed pure Aluminum samples. Firstly,
the etched layers were removed gently by polishing with soft colths to give smooth and
clean surfaces. The microhardness measurements were then taken using an FM-1e
microhardness tester equipped with a Vicker indenter. Different loads were used
dependent on the materials to ensure a decent size of the indentation, for Al-7136 alloys ,
a load of 200 gf was chosen while for pure aluminum, a load of 50 gf was selected. A
dwell time of 15 s was held for each separate measurement. The values of the Vickers
microhardness, Hv, were measured using two different procedures to obtain detailed
information on the variation of the hardness values across the diameters of each disk and
the total distribution of hardness values over the total surface of each individual disk.
74
Figure 3.3 Schematic drawing of the High Pressure Torsion (HPT) facility.
75
Fig.3.5. shows a schematic illustration of one-quarter of an HPT disk with the
locations shown for each separate measurement of the microhardness. For hardness
measurement across the diameter of the disk, the average value of Hv were determined at
selected positions from the center of the disk to both edges where these measurements
relate to individual positions separated by incremental distance of 0.3 mm; these points
are denoted by the small open circles along the bottom edge of the one-quarter disk
depicted in Fig.2. For each of these positions , the average value of Hv was determined
from four separate hardness measurements recorded at uniformly separated points
displaced from the selected position by a distance of 0.15 mm. For the hardness profile
over the total disk surface, the individual values of Hv were recorded following a
rectilinear grid pattern with separations of 0.3 mm between each consecutive point; the
grid pattern for one-quarter of the surface is shown in Fig.2. All of these individual
values of Hv were then used to construct color-coded contour maps that provided clear
visual presentations of the distributions in hardness across the surface of each disk.
3.2.4. Microstructural analysis and sample preparations
3.2.4.1. Transmission Electron Microscopy ( TEM )
In order to investigate the effect of severe plastic deformation (SPD) on the
microstructure change of the supersaturated Aluminum alloys, observations by TEM
were conducted to describe the morphologies of the grains and precipitates. Bright field
76
images at different magnifications as well as selected area electron diffraction (SAED)
images were taken using a Philips CM12 operating at 120 kV as well as a JEOL 3000F
at 300 kV.
For ECAP samples, the disks with thickness of ~1 mm were sliced from the billets
perpendicular to the pressing direction. The disks were then polished down from both
sides to have a thickness of ~100 m using the conventional polishing method without
micro-cloths. After cutting off the edge of the disk using a punch to set the diameter of
the disks to 3 mm, the center of these disks were then ultimately thinned by a twin-jet
electropolishing methods using an electrolyte of 30% nitric acid and 70%
methanol at 20 V and at a temperature of -20 °C.For HPT samples, the original disks
were divided into two parts: the center part and the peripheral part as shown in Fig.3.
With an original diameter of the disk of 10 mm, the center part was defined as a
concentric disk with the diameter of 3mm, the outer ring-shape part of the disk was
defined as the peripheral part as illustrated. TEM samples were retracted from both the
center part and the peripheral part of each disk after initially thinning using the
conventional mechanical polishing methods (including micro-cloths) to the thickness of
~100 m. The samples were then ultimately thinned by a twin-jet electropolishing
methods using an electrolyte of 30% nitric acid and 70% methanol at 20 V and at a
temperature of -20 °C.
77
Figure 3.4 Schematic drawing of the tensile specimen.
Figure 3.5 Grid pattern of the Vickers Hardness measurement on the HPT discs.
78
3.2.4.2. 3-D Atom Probe
The atom probe is an atomic-resolution microscope used in materials science that
was invented in 1967 by Erwin Wilhelm Müller, J. A. Panitz, and S. Brooks McLane. It
is basically a combination of a field ion microscope and a mass spectrometer, as shown in
Fig.3.6[119]
As in the atom probe, single atoms on the surface of a sharply-pointed needle are
ionized by field evaporation. The ions produced are projected away from the specimen, in
this case towards a position-sensitive detector with single atom sensitivity.
The early atom probe field ion microscopes were very inefficient in the amount of
specimen analyzed due to the small acceptance angle of the mass spectrometer. Therefore,
some new variants of atom probe were developed to overcome that limitation. As these
instruments produce three-dimensional images of the internal structures of specimens
from many slices, each containing a few atoms, this technique has been termed atom
probe tomography (APT). Essentially, as atoms on the surface of the specimen field
evaporate one at a time and fly toward a two-dimensional position-sensitive detector,
their hit position in x and y is recorded. Gradually, each atom on the surface evaporates
and exposes the underlying layers. The sequence of atom hits on the detector can be used
to track both the serial evaporation of atoms in a given layer and the serial evaporation of
the layers. The three dimensional image is thus reconstructed from this combination of
two-dimensional hit positions and field evaporation sequence.
79
The unique advantages of 3-D atom probe are the very high spatial resolution and
atomic level chemical characterization capability. A modern 3-D atom probe could
achieve a resolution of ~0.3 nm. Despite there are techniques like Scanning Tunneling
Electron Microscopy (STEM) that could reach a slightly higher resolution, the thickness
of the samples are usually limited to be less than 10 nm. [1] Furthermore, 3-D atom
probe tomography (3-D APT) could provide striking, revealing exciting new perspectives
on microstructural features such as precipitates and phase interfaces by reconstruction.
This is especially useful in this study as the involvement of complicated precipitation
occurring during the SPD processing.
As a result of the previous description, 3-D Atom probe tomography (APT) was
employed in this work to reveal the precipitation evolution as well as the chemical
composition and evolution of the grain boundaries. Quantitative APT analysis was used
also to provide chemical information on the precipitates and the matrix, the spatial
distribution of precipitates during ECAP, and to yield unique information on both the
structure and chemistry of the precipitate evolution [120,121]. The new information from
this research provides additional insight into the precipitation mechanism occurring in
SPD processing.
The tip samples for atom probe characterization were prepared using a two-step
electro-polishing procedure with thin bars having cross-sections of 0.5 0.5 mm
2
. The
first step used an electrolyte of 25% perchloric acid in acetic acid at 15 V at room
temperature and the second step used an electrolyte of 4% perchloric acid in 2-
butoxethynal at 20 V. Characterization by APT was performed in a local electrode atom
80
probe (LEAP) at a specimen temperature of 20 K under a laser pulse at an energy of 1.0
nJ and with a target evaporation rate of 1%. This laser energy is sufficiently high to
consistently measure the precipitate concentrations [122].
3.2.4.3 X-ray diffraction
X-ray diffraction was also conducted on to provide more information on the
precipitation evolution as well as the chemical composition of Al-7136 subjected to
severe plastic deformation.
Quantitative X-ray diffraction measurements of the HPT and HPT +CR samples
were performed using an X-ray diffractometer equipped with a Cu target
operating at 1.8kW and a graphite curved single-crystal (0 0 0 2) monochromator. The
Cu radiation was selected at the goniometer receiving slit section, the divergence and
anti-scattering slits were set at 0.5 ◦ and 0.5 ◦, respectively, and the width of the receiving
slit was 0.3 mm. A series of θ–2 θ scans was performed to provide a record of the XRD
patterns at room temperature.
3.3 Overall summary of experimental procedures
3.3.1. Summary of ECAP part
For the first round, ECAP processing was performed at a temperature of 473 K on
the as-received Al-7136 alloy. Samples were subjected to 1,2,4,6 and 8 passes
81
Figure 3.6 Illustration of the 3-D Atom Probe technology.[119]
82
corresponding to imposed strain of ~1,~2,~4,~6 and ~8 , respectively. In repetitive
pressings, each sample was rotated by 90 °in the same sense between each pass in the
processing route termed B
C
, which was selected because it was reported to be the
optimum processing procedure. After ECAP processing, TEM analysis was carried out at
the Electron Microscopy Unit (EMU) of University of Sydney, Australia to investigate
the microstructure as well as the precipitation morphologies of the material. In order to
understand the precipitation evolution along with the ECAP processing at 473 K, 3-D
Atom Probe Tomography (3-D APT) was also conducted at EMU to depict the detailed
precipitation morphologies at different stage during the processing. To supplement the
investigation, X-ray diffraction was also conducted on samples subjected to different
passes of ECAP at 473 K.
Tensile tests at room temperature and elevated temperatures were carried out using
an Instron machine on the as-received Al-7136 specimens as well as those subjected to
different passes of ECAP. For room temperature tensile test, a strain rate of 3.3 × 10
-4
s
-1
was selected while those at elevated temperatures, various strain rates ranging from 1.0 ×
10
-4
s
-1
to 3.3 × 10
-2
s
-1
were chosen dependent on specific testing temperature.
Second part of the ECAP pressing of Al-7136 was conducted at ambient temperature and
383 K. For pressing at ambient temperature, the materials were first subjected to solution
heat-treatment at elevated temperatures, followed by water quenching to introduce
supersaturated solid solutions (S.S.S). ECAP processing at ambient temperature as well
as at 383 K was then performed on these as-quenched rods within a few minutes for 1
pass, equivalent to an imposed strain of ~1. Several pre-pressing solid solution
83
approaches were evaluated at different temperatures and/or durations of time. After
pressing, the microstructure as well as the precipitation evolution during the processing
procedure was also investigated following the similar approach conducted on those
directly pressed at 473 K using TEM, 3-D APT as well as XRD at the EMU of University
of Sydney. Meanwhile, tensile test at room temperature was carried out on the samples
subjected to 1 pass ECAP at ambient temperature.
3.3.2. Summary of HPT part
High pressure torsion was conducted on Al-7136 alloys subjected to different pre-
processing conditions. In the first group, as-received Al-7136 alloys were made into disks
with the diameter of 10 mm and thickness of ~0.8 mm, HPT processing was directly
conducted on these disks at ambient temperature with a hydraulic pressure equal to 6 GPa.
The disks were subjected to 1/8, 1/4 , 1/2, 1, 2, 3, and 4 turns of HPT processing ,
respectively. For the second group, Al-7136 rods were pre-processed by solid solution at
470 C for 60 mins ,immediately followed by water quenching and room temperature
ECAP for 1 pass, the center parts of the as-pressed rods were then cut into disks and
polished into the desired size by abrasive papers for HPT processing, an identical 6 GPa
pressure was applied to those samples with 1/2, 1 , 2 and 4 turn,respectively. The third
group of HPT specimens underwent ECAP pressing at 473 K for 4 passes first, followed
by similar HPT processing with a pressure of 6 GPa , the samples were subjected to 1
and 2 turns of HPT.
84
All these sample disks after HPT processing were carefully polished by abrasive
paper to get smooth and clean surface , and mounted for Vickers’s hardness test
following a rectilinear grid pattern with separations of 0.3 mm between each consecutive
point by a FM-1e hardness test machine.
After hardness test, the original disks were divided into two parts: the center part and
the peripheral part, TEM samples were retracted from both the center part and the
peripheral part of each disk after initially thinning using the conventional mechanical
polishing methods followed by ultimately thinning by a twin-jet electropolishing
methods. Bright field images as well as selected area electron diffraction (SAED) images
were taken using the TEM unit at the EMU of University of Sydney, Australia.
85
4. EXPERIMENTAL RESULTS
4.1 ECAP of Al-7136 alloy
4.1.1. ECAP of Al-7136 alloy at 473 K
As stated in the previous chapters, as-received Al-7136 alloy was cut into rods and
processed by ECAP at 473 K in the first round of the experiments.
4.1.1.1. Room temperature tensile results
The results for the room temperature tensile testing of the Al-7136 alloy are
plotted in Fig. 4.1 in the form of the true stress-true strain curves and the strain
hardening rates. The limits of the uniform elongation are maked on each separate curve
as
u
.
For the as-received alloy shown in Fig. 4.1(a), there was no necking and failure
occurred at a strain of ~0.2. For the samples processed by ECAP, the extent of the
uniform elongation was measured by the intersection of the plot of true stress and the
strain hardening rate. The results are given in Figs 4.1(b)-(f) and they show the uniform
elongation varies insignificantly in these samples with values between 0.09 to 0.14. Thus,
there is no simple relationship between the number of passes and the extent of the
uniform elongation.
An important result from these plots is that the measured yield stresses of the Al-
7136 alloy are significantly lower after ECAP by comparison with the as-received
86
material and, in addition, there is relatively little change in the yield stresses between 1
and 8 passes. By contrast, there is a tendency for the flow stress to decrease with
increasing numbers of ECAP passes. This trend is contrary to the usual behavior after
ECAP where there is generally a significant increase in strength but it is consistent with
the results reported earlier for a similar Al-7034 alloy [78]. Microstructure observations
on the Al-7034 alloy showed that processing by ECAP at a temperature of 473 K led to a
refining of the grain size, a fragmentation of the strengthening η-phase and also a
dissolution of the η'-phase which is the main reason for the decrease in strength due to
ECAP processing. Furthermore, with further pressing through additional passes in ECAP,
the effect of the dissolution of the strengthening phase becomes more pronounced and
further contributes to the loss in strength.
4.1.1.2. High temperature tensile test
The inherited features of severe plastic deformation methods, or more specifically
ECAP , is widely reported to be able to improve the high temperature performance of the
materials contributed to the ultrafine grain size and high angle grain boundary
misorientations. There are various reports claiming superplasticity achieved on the
materials after ECAP , including aluminum alloys. Xu et al. [21 ]also reported the
superplastic performance of Al-Zn-Mg-Cu alloys tested at 673 K and a strain rate of 1 x
10
-2
s
-1
.
87
Figure.4.1 Room temperature tensile tests of Al-7136 after ECAP at
473 K via route B
c
.
(a)
(b)
(c)
88
Figure.4.1. Continued. Room temperature tensile tests of Al-7136 after ECAP
at 473 K via route B
c
.
(d)
(e)
(f)
89
In order to investigate the high temperature mechanical behaviors of the Al-7136
alloys subjected to different processing by ECAP. High Temperature tensile
tests were conducted using the Instron machine at different temperatures and with various
strain rates. The as-received material was also tested for comparison. The engineering
strain-engineering stress curves were plotted in Figure 4.2-4.5.
For the samples pressed by ECAP at 473 K, the 4- Pass Al-7136 samples were first
tested at 673 K using strain rates: 3.3 x 10
-5
s
-1
, 1 x 10
-4
s
-1
, 1 x 10
-3
s
-1
, 1 x 10
-2
s
-1
and
3.3 x 10
-2
s
-1
.The results were plotted into figure.4.2. The flow stress of the samples
decreased as the testing strain rates decreased. At the highest strain rate of 3.3 x 10
-2
s
-1
,
a maximum flow stress of ~40 MPa was observed. All the specimens failed at the
elongation-to-failure of ~ 230% except the one tested at the strain rate of 1 x 10
-2
s
-1
,
where the elongation-to-failure was measured around 300%.
After this , the testing temperature were raised to 703 K, the 4- Pass Al-7136
samples were tested using strain rates: 1 x 10
-4
s
-
1, 1 x 10
-3
s
-1
, 1 x 10
-2
s
-1
,1 x 10
-1
s
-1
.
Again , the flow stress decreased as the testing strain rates decreased., the maximum flow
stress of 35 MPa achieved at the strain rate of 1 x 10
-1
s
-1
is slightly lower than the one
achieved at 673 K and strain rate of 3.3 x 10
-2
s
-1
, this is common considering the higher
testing temperature. The elongation-to-failure varies at different testing strain rates : at
the lowest strain rate of 1 x 10
-4
s
-1
, the elongation-to-failure of ~160% was measured,
this value increased to ~170% at the highest strain rate of 1 x 10
-1
s
-1
and ~230% at 1 x
10
-3
s
-1
. The maximum elongation-to-failure measured is ~280% at the strain rate of 1 x
10
-2
s
-1
as well.
90
Consequently, the strain rate of 1.0 x 10
-2
s
-1
is considered as the optimum strain
rate for Al-7136 subjected to ECAP to be tested at high temperature. To better
understand the effect of ECAP on the high temperature mechanical behaviors of Al-7136,
tensile specimens were cut from Al-7136 subjected to different number of passes of
ECAP at 473 K. The comparison tests were conducted at the 703 K and the strain rate
1.0 x 10
-2
s
-1
, the result in the form of engineering strain- engineering stress curves was
plotted in Figure.4.4.
As plotted, the maximum flow stress of ~ 25 MPa was achieved at the 8-Pass
sample, meanwhile, the elongation-to-failure achieved on the 8-Pass samples is the
lowest , at the value of ~230%. The maximum elongation-to-failure was measured at the
4-Pass sample with the value of ~ 280%. While for other samples the flow stress and
elongation-to-failure varies, there is not a simple relationship between the passes of
ECAP and the elongation-to-failure or flow stress.
The as-received Al-7136 was also tested at 673 K to investigate the high
temperature mechanical performance. Testing strain rate varies from 1.0 x 10
-4
s
-1
to 1.0
x 10
-2
s
-1
, the engineering stress- engineering strain curve was plotted in Figure.4.5.
Similarly, the flow stress decreased as the strain rate decreased; surprisingly, the
maximum elongation-to-failure achieved on the as-received Al-7136 at the strain rate of
1.0 x 10
-3
s -1 is close to 480%, larger than the maximum one achieved on the as-pressed
samples. One thing has to be mentioned is that the maximum elongation-to-failure was
achieved on the as-received Al-7136 at the strain rate of 1.0 x 10
-3
s
-1
, an order lower
91
than the optimum strain rate for the as-pressed samples. This is consistent to earlier
reports on the high temperature tensile tests of different alloys.[123-125]
4.1.2. Develop a RT ECAP method
It is apparently feasible to press the Al-7136 alloys at the elevated temperature of
473 K based on the results above, in fact, the surface of the rods subjected to 8 passes of
ECAP at 473 K still remained smooth without visible cracks. However, as the
mechanical test indicated, the material was softened rather than strengthened after
processing by ECAP at elevated temperature in the room temperature tensile test;
additionally, despite a retain or slightly strengthening of the flow stress in the high
temperature tensile tests, the as-pressed samples demonstrated an even inferior
elongation-to-failure compared to the as-received counterparts
An immediate thought would be to decrease the pressing temperature, however, in
practice the Al 7000 series alloys are often difficult to process by ECAP at room
temperature directly because of their very limited formability. An alternative solution is
to solid solute the alloys for certain time, followed immediately by water quenching, and
then conduct the ECAP at room temperature. This processing routine has been approved
to be feasible in a few Al-Zn-Mg alloys and achieve strengthening rather than softening .
Consequently, a similar approach to Al-7136 is tempting to strengthen the
material. However, the specific solid solution parameters may vary from previous reports
92
ENGINEERING STRAIN (%)
0 50 100 150 200 250 300 350
ENGINEERING STRESS ( MPa)
0
5
10
15
20
25
30
35
40
45
50
Al-7136
ECAP:4 P (473 K)
T=673 K
3.3 x 10
-2
s
-1
1.0 x 10
-2
s
-1
1.0 x 10
-3
s
-1
1.0 x 10
-4
s
-1
3.3 x 10
-5
s
-1
Figure 4.2. Engineering stress-strain rate curve of Al-7136 alloy after 4 passes of ECAP,
tested at 673 K and various strain rates.
ENGINEERING STRAIN ( %)
0 50 100 150 200 250 300
ENGINEERING STRESS (MPa)
0
5
10
15
20
25
30
35
40 Al-7136
ECAP: 4 P (473 K)
T= 703 K
1.0 x 10
-1
s
-1
1.0 x 10
-2
s
-1
1.0 x 10
-3
s
-1
1.0 x 10
-4
s
-1
Figure 4.3. Engineering stress-strain rate curve of Al-7136 alloy after 4 passes of ECAP,
tested at 703 K with various strain rates.
93
ENGINEERING STRAIN(%)
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
ENGINEERING STRESS (MPa)
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
Al-7136
ECAP: 473 K
T= 703 K
10
-2
s
-1
。
8-P
4-P
1-P
2-P
Figure 4.4.Engineering stress-strain rate curve of Al-7136 alloys after different passes of
ECAP, tested at 703 K.
ENGINEERING STRAIN (%)
0 100 200 300 400 500
ENGINEERING STRESS ( MPa)
0
4
8
12
16
20
24
Al 7136
T=673K
1.0 x 10
-2
s
-1
1.0 x 10
-3
s
-1
1.0 x 10
-4
s
-1
Figure 4.5. Engineering stress-strain rate curve of as-received Al-7136 alloys, tested at
673 K with various strain rates.
94
Table 4.1 Hardness of Al-7136 after solid solution at various temperatures,*denotes the
test was done 5 days after solid solution.
Solid Solution Time Solid Solution Temperature HV HV*
‐ ‐ 163.51 165.21
30 min 743 K 81.64 174.20
30 min 773 K 79.44 169.20
60 min 743 K 76.36 172.00
60 min 773 K 76.22 166.60
95
as the chemical compositions are different. For this reason, aging experiments were
conducted to assist the parameters selection. Traditionally, the solid solution temperature
for Al 7000 alloys were selected between 743 K and 773 K, thus these two temperatures
were chosen and thermal durations of 30 mins and 60 mins were selected as shown in the
Tab.4. 1. The furnace were heated to the pre-set temperature and kept stable for 20
minutes, the rods in the length of 64 mms were carefully packed with thin aluminum foils
to keep the possible oxidation to the minimum extent and put into the center of the
furnace for the desired time; The rods were immediately put into the water tank for
quenching after solid solution. Discs were sliced from the as-quenched rods
perpendicular to the cylindrical axis and mounted. Before Vickers hardness test ,
these specimens were polished with the abrasive paper as well as cloth to mirror-like
surface. Each of the hardness results was averaged from 16 separate tests randomly
distributed over the sample.
It is found solid solution treatment could effectively soften Al-7136 thus provides
the possibility of conducting ECAP on these solid solution treated samples at room
temperature. The minimum Hv measured was from the samples solid solution treated at
773 K for duration of 60 minutes; however, the difference in Hv is only marginal at 743
K for 60 miniutes. Considering the possible grain growth occurrence during the solid
solution, the lower 743 K is a better choice.
Furthermore, this processing routine including solid solution, quenching and
immediately room temperature ECAP processing was employed with different solid
solution treatment to testify the feasibility . Figure.2 shows two billets of alloy Al-7136
96
subjected to a solution treatment at 743 K, water quenching and then processing by
ECAP within ~10 minutes of the quenching operation. These two billets differ only in
the length of time used for the solution treatment where these times are (a) 30 and (b) 60
minutes, respectively. It is apparent from Fig 4.6 that a solution treatment
of 30 minutes is not satisfactory for this alloy and instead the billet shows segmentation
in Fig. 4.6(a) whereas in Fig. 4.6(b) the billet has a smooth surface when processed by
ECAP after a solution treatment for 60 minutes. These results demonstrate that
segmentation occurs in the Al-7136 alloy after a short solution treatment but that
segmentation may be avoided if the solution treatment is extended for a sufficiently long
time.
This is actually different from the previous reports, Chinh et al reported the
successful pressing of a similar Al-Zn-Mg alloy at room temperature of up to 7 passes,
the solid solution temperature was set at 743 K and for a duration of 30 minutes[81]. No
longer solid solution time was tried for that alloy. The exact same processing routine
resulted visual large shear bands in Al-7136. This could be contributed to the higher
alloying elements percentage in Al-7136, for the previous alloy, it contains 5.7 wt% Zn,
1.9 wt% Mg and 1.5wt% Cu, however, in Al-7136 , all these numbers almost doubled: it
contains 9.4 wt% Zn, 2.5 wt% Mg and 2.5 wt% Cu. It is thus very likely more
precipitates would form at natural ageing condition in the Al-7136 alloys, assuming
similar diffusion coefficients of the alloying elements in these alloys, the solution of these
precipitates takes a relatively longer time at any selected elevated temperature by
comparison with the lower concentration counterparts. Considering the issue of
97
Figure 4.6. Surface topography of billets after 1 ECAP pass at room temperature in the
case of Al-7136 solution treated at 743 K for a) 30 mins;b)60 mins.
98
processing by ECAP for age-hardenable alloys where the strengthening precipitates play
an important and often dominant role, the present results provide a clear demonstration
that it is imperative to develop an appropriate solution treatment strategy for every alloy.
4.1.2.1 Room temperature tensile test
A tensile specimen was also cut from the billet of the Al-7136 after ECAP at room
temperature (RT) through one pass and it was tested at room temperature and the strain
rate of 3.3 x 10
-4
s
-1
. The results are shown in Fig. 4.7. where data are presented for the
as-received condition, after 1 pass of ECAP at 473 K and after a solution treatment at 743
K for 1 hour and then pressing through 1 pass of ECAP at room temperature. These
results show there is an increase in the yield strength of the billet pressed at room
temperature by >100 MPa thereby confirming that lower temperature processing is
advantageous in retaining high strength in this alloy.
4.1.2.2. High temperature tensile test
Figure.4.8. showed the engineering stress-engineering strain curve of the R.T.
pressed sample, the sample achieved an elongation-to-failure of ~260% when tested at
673 K and the strain rate of 1.0 x 10
-3
s
-1
, as indicated, this is almost identical to that of
the sample subjected to 1-Pass ECAP at 473 K and tested at 703 K and the strain rate of
1.0 x 10
-2
s
-1
, however, the flow stress is higher in the R.T. pressed sample even at a
lower strain rate.
99
4.1.3. Pressing Al-7136 at 110 ℃ after solid solution
Despite the advantage of the room temperature pressing in retaining the strength,
it is comparatively difficult to process Al-7136 by more than 1 pass following the routine
above. As a result, it is worth trying if some compromise on the pressing temperature
could improve the workability of the material.
For this reason, Al-7136 rods were solid solute treated at 743 K for 1 hour and
water quenched to form the supersaturate solid solution( S.S.S.) , then immediately
pressed by ECAP at 110 ℃ (383 K) for 1 Pass. The as-pressed billets showed no visual
cracks or shear bands with smooth surface. It is thus concluded as a possible approach to
press the Al-Zn-Mg-Cu alloy.
4.1.3.1. Room temperature tensile test
Similarly, tensile specimen was cut from the as-pressed rods parallel to the
cylindrical axis and tested by Instron machine at room temperature and the strain rate of
3.3 x 10
-4
s
-1
, the results are shown in Fig.4.9 in the form of stress-strain curve. The
yield strength of the rods pressed at 383 K remained similar to the as-received material,
but the ultimate tensile strength (UTS) was apparently lowered by this processing routine,
the elongation-to-failure also decreased.
In order to separate the effect of ECAP on the mechanical properties, tensile tests
were also conducted on the samples subjected to solid solution treatment and a similar
thermal duration at 110 ℃. The testing parameters were identical to previous room
100
Figure.4.7. Room temperature tensile properties of Al-7136 Processed by ECAP pressed
under different conditions.
True Strain
0.0 .1.2.3
True Stress (MPa)
0
200
400
600
800
1000
As-received
ECAP: 473 K,1 Pass
Solid Solution: 743 K, 1h+ ECAP: R.T., 1 Pass
Al-7136
T = 296 K
0
= 3.3x10
-4
s
-1
True Strain
0.0 .1.2 .3.4 .5
True Stress(MPa)
0
200
400
600
800
1000
Solid Solution:743K 1h
Solid Solution:743K 1h+ ECAP:1 Pass, RT
As Recieved
Al-7136
T=296 K
0
=3.3x10
-4
s
-1
(a)
(b)
101
Figure 4.8 The high temperature stress –strain curve of the room temperature pressed
Al-7136.
Engineering Strain
0.0 .5 1.0 1.5 2.0 2.5 3.0
Engineering Stress (MPa)
0
10
20
30
40
50
Solid Solution:743 K,1 h+ECAP:1 Pass, RT; Tensile: T=673 K, =10
-3
S
-1
ECAP:1 Pass,473 K; Tensile: T=703 K, =10
-2
S
-1
Al - 7136
102
temperature test. It is found the rods with ECAP processing again indicated a lower
strength compared to that with ageing treatment, it suggested that like the one pressed at
473 K, ECAP at 383 K would also lead to the softening rather than strengthening of the
material.
4.1.3.2. High temperature tensile test
Al-7136 samples subjected to solid solution , water quenching followed by ECAP
at 383 K were also tested at high temperature, the results in the form of engineering
stress-engineering strain curves were again plotted in the following graphs.
In Figure. 4.10., 383 K-pressed samples were tested at 643 K at the strain rates of
1.0 x 10
-2
s
-1
and 1.0 x 10
-3
s
-1
by Instron machine. The one tested at 1.0 x 10
-3
s
-1
indicated a better engineering elongation-to-failure of ~310% comparing to the ~240%
achieved at 1.0 x 10
-2
s
-1
. The tensile test temperature was then shifted to 673 K , same
strain rates were used, again the one tested using the initial strain rate of 1.0 x 10
-3
s
-1
,
but the elongation-to-failure dropped to ~260%, the sample tested at the initial strain rate
of 1.0 x 10
-2
s
-1
showed a similar elongation-to-failure as that tested at the 643 K, but the
flow stress was lower.
4.1.4.Microstructures of studyAl-7136
4.1.4.1 .Mass spectrum Study
103
A typical mass spectrum from Al-7136 sample analyzed using a LEAP under
laser pulsing is shown in Fig. 4.11. Cr and Mn are trace elements in the alloy as is evident
in Table 4.2. However, their peaks are invisible in the mass spectrums obtained under
voltage pulsing because the two elements will ionize mainly in the +2 charge-state ions
during field evaporation and they are hard to resolve from the tails of the Al
+1
and Mg
+1
peaks. Nevertheless, under laser pulsingthe two elements can evaporate as +1 charge-
state ions due to the exchange charge-state [122]and accordingly their peaks are clearly
discernible in Fig.4.11. The specimen taper angle is an important specimen parameter
which is known to have a significant effect on the resolution of a mass spectrum obtained
under laser pulsing mode. Thus, a tip with a larger taper angle often offers improved
mass resolution [126]. The tips used for this experiment had a higher taper angle of >10°
and achieved a mass-resolution of M/ M 700 at the full-width half maximum (FWHM)
for most analyses under the laser pulsing. This resolution is higher than a resolution of
350 (FWHM) from a similar tip run under voltage pulsing. One important benefit is that
the trace element peaks of Si, V and Zr are all evident in the mass spectrum as shown in
Fig. 4.11.
4.1.4.2. X-Ray diffraction result
The magnified XRD patterns of as-received and as-pressed Al-7136 samples were
shown in Fig.4.12, The indexes of diffraction planes of the hexagonal η were also
indicated in the figures. . For all of the Al-7136 samples, besides Al reflections, there
104
(a)
True Strain
0.0 .1.2.3.4.5
True Stress(MPa)
0
200
400
600
800
1000
Solid Solution:743K 1h+ ECAP:1 Pass, RT
Solid Solution:743K 1h+ ECAP:1 Pass,383K
As Recieved
ECAP:1 Pass, 473K
Al - 7136
T=296 K
0
=3.3x10
-4
s
-1
(b)
True Strain
0.0 .1 .2 .3 .4 .5
True Stress(MPa)
0
200
400
600
800
1000
Solid Solution:743K 1h+ Aging:383K 20 min
Solid Solution 743K 1h+ ECAP:1 Pass,383K
Al - 7136
T=296 K
0
=3.3x10
-4
s
-1
Figure.4.9. Room Temperature Tensile properties of Al-7136 Processed by ECAP
Pressed under different conditions.
105
(a)
Engineering Strain
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Engineering Stress (MPa)
0
10
20
30
40
50
Al - 7136
Solid Solution:743K,1h
ECAP:1 Pass,383K
T=643 K
10
-2
s
-1 10
-3
s
-1
(b)
Engineering Strain
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Engineering Stress (MPa)
0
10
20
30
40
50
Al - 7136
Solid Solution:743K,1h
ECAP:1 Pass,383K
T=673 K
10
-2
s
-1
10
-3
s
-1
Figure.4 .10.High Temperature Tensile properties of Al-7136 Processed by ECAP
Pressed under different conditions.
106
appeared a broad peak at about 2 θ=20° and some other weak peaks. The broad peak at
about 20° corresponds to the G–P zones [127], and the other weak peaks whose positions
are a little lower than those of hexagonal η phase are from the metastable hexagonal η
′
phase, whose lattice parameters are a little larger than those of η phase [128].
Because the XRD sample has the same area involved in the reflection, the
intensity of the XRD pattern can be compared. Fig. 4.12 shows that the intensity of the
G–P zone broad peak of the 1-pass sample is larger than that of the as-received sample,
indicating that the volume fraction of the G–P zones in the 1-pass sample is larger than
that in the as-received sample.
This comparison was also made between the 1 Pass ECAP processed at different
temperatures. The phase peak at 20 degree is a lot intense in the one pressed at 473 K
than that processed by solid solution and room temperature pressing, it is suggesting
solid solution could successfully solute the precipitates from natural ageing , and also the
precipitation is moderate in the 1 pass room temperature sample compared to that directly
subjected to ECAP at 473 K for 1 pass.
4.1.4.3 . TEM study of Al-7136 subjected to ECAP
4.1.4.3.1. Microstructure of the as-received Al-7136
TEM Images of the as-received Al-7136 are shown in Fig.4.13. Grain size
measurement in a wider selection of TEM images was conducted , it is concluded that
107
the as-received alloy contained a reasonably equiaxed array of grains with a
comparatively low dislocation density. The average measured grain size is around 3.4
m. Like other Al-Zn-Mg-Cu alloys[20-22], the as-received Al-7136 also contained the
and ’ precipitates distributed homogeneously throughout the materials.
The plate shaped precipitates were observed within the grains, with the length of
around 600 to 800 nm and widths of around 200 nm; the needle-like ’ phases were
much smaller , with the length of less than 40 nm, mostly lying on the grain boundaries,
with a very small portion in the grains. At a even higher magnification, Al
3
Zr particles
at the diameters of ~25 nm were observed . This result is consistent with typical
microstructure of 7xxx Aluminum alloys with similar chemical compositions.[20-22]
4.1.4.3.2. Microstructure after ECAP through different number of
passes
The microstructure of Al-7136 after processing by ECAP through totals of 1, 4,8
passes were observed by TEM and shown in Figs.4.14-4.16.
Fig.4.14. shows the microstructure of the alloy after one single pass of ECAP at
473K. The grains were noticeably refined , however, most of the grains were also
elongated . Grain size measurements concluded the average length of the grains was ~1.8
m, while the average width was~0.25 m. The appearance of the precipitations changed
108
Figure.4.11 The mass spectrum of the as-received Al-7136
92.14 0.05 0.03 0.02 0.06 1.03 0.07 0.11 2.71 3.78 at%
84.93 0.1 0.05 0.05 0.2 2.5 0.15 0.12 2.5 9.4 wt%
Al Ti Cr Mn Zr Cu Fe Si Mg Zn
Table 4.2. The nominal compositions of Al-7136.
Mg
+2
Al
+2
Si
+2
Mg
+1
Zn
+2
Zr
+2
Cr
+1
Mn
+1
Cu
+1
Cu
+1
Zn
+1
Al
+1
AlH
2
+1
V
+2
109
significantly:the large rod-shaped precipitations were barely seen in the TEM images
but an evidence of fragmentations could be found, this is consistent with the observations
by Xu et al.[21] in Al-7034 alloy; meanwhile, there were many platelet precipitates with
the aspect ratio of ≤3:1 distributed homogeneously , the maximum length of these
precipitates was about 33 nm.
Fig.4.15 shows the microstructure of Al-7136 after 4 pass ECAP at 473K. Some
of the grains still kept the elongated shape with the aspect ratio around 3:1, but the
lengths of these grains decreased slightly to about 1 to 1.5 m, meanwhile, many of the
grains evolved into reasonable equiaxial shape with the measured size of ~0.5-0.8 m.
The grain boundaries were also less wavy and better defined. The microstructure of the
precipitates changed significantly, most of them changed into equiaxed shape , the largest
size was around 35-40 nm.
The microstructure of the alloy processed by 8 passes of ECAP at 473K is shown
in Fig.4.16. A reasonable equiaxed grain structure is found and the average grain size is
~300-400 nm. The precipitates remain equiaxed, despite some very fine particles, the
average precipitates size increased with the increasing number of passes. The maximum
precipitate size is around 65-70 nm. One thing has to be pointed out is there is no
alignment of precipitates in the 4-pass and 8-pass samples, which is considered to be the
evidence of fragmentation of large precipitates. It is thus concluded that fragmentation
only happened to large precipitates.
110
Figure.4.12 The XRD profile of Al-7136 in different conditions
Al‐7136
20 30 40 50
2-Theta(°)
0
50
100
150
200
250
300
350
400
Intensity (Arb. Units )
1‐P,473 K
1‐P,R.T
(b)
Al‐7136
20 30 40 50
2-Theta(°)
0
50
100
150
200
250
300
350
1‐P,473 K
Intensity (Arb. Units )
As‐Received
(a)
111
TEM investigation was also conducted on the room temperature pressed sample
as shown in Fig.4.17, the grains were no surprisingly in the elongated shape, the
measured average length of the grains was ~1.20 m and the measured average width
was ~0.15 m. The grain size is substantially smaller compared to the counterpart
pressed at 473 K directly for 1 pass. This should be contributed to the comparatively
lower kinetics at the lower pressing temperature, and is believed to be one of the reasons
of the superior room temperature mechanical performance. Another noticeable difference
between the room temperature 1-pass sample and the 473 K 1 –pass sample is the
precipitation. There is no platelet shaped precipitates in the room temperature 1-pass
sample, instead, much smaller spherical precipitates in the diameter of ~ 10 nm were
observed. Also the precipitates were in a much smaller number compared to the 473 K 1-
pass sample. This is again from the lower precipitation kinetics.
It is difficult to explain the softening of the materials based on the grain size
evolution. Like in other aluminum alloys, ECAP could effectively refine the grains, and
could achieve equiaxial grain structures of Al-7136 when the imposed strain reached a
critical level. Meanwhile, the non-equilibrium grains and wavy grain boundaries
suggested a much higher dislocation density after ECAP processing. Assumingly, Al-
7136 should be strengthened rather than softened. The overall softening suggested a very
important role of solid solutes and precipitates in the ECAP processing.
112
Figure.4.13. Microstructure of the as-received Al-7136
Figure.4.14. Microstructure of the 473 K ECAP 1-Pass Al-7136.
113
4.1.4.4 Precipitate Morphology and Distribution
As a unique feature of the Al-7000 series alloys, precipitate plays an important
role influencing both the mechanical behavior at room temperature and that at the high
temperature, in fact, it is generally believed the change of the morphologies as well as the
distribution of the precipitates are the main reasons why the Al-Zn-Mg-Cu alloys
indicated a rather unique mechanical properties when compared with other Al alloys
subjected to identical or similar ECAP processings. The mircostructure study
revealed the grains of Al-7136 were still effectively refined to the ultranfine grain size
range, espeically after 8 passes of ECAP, the grains are mostly in the equiaxed shape.
This is similar to other alloys subjected to ECAP , however, Al-7136 , or more generally ,
the Al-Zn-Mg-Cu alloys exhibited a decrease in the strength rather than improvement, it
could not be explained by the grain refinement. Consequently, another noteable change of
the microstructure brought by ECAP, the changes of morphologies and distributions of
the precipitates should be carefully studied.
It is usually very difficult to understand the evolution of the precipitates ,
especially on the compositions of the precipitates by the TEM investigation alone. By far,
most of the reports only employed X-ray diffraction methods as well as TEM to
investigate on this issue. The modern 3-D atom probe tomography provided a powerful
tool to depict not only the morphology and distributions of the precipitates statically, but
most importanly, on the evolution of the precipitates. Meanwhile, TEM was also used in
this research.
114
Figure.4.15. Microstructure of the 473 K ECAP 4-Pass Al-7136.
Figure.4.16. Microstructure of the 473 K ECAP 8-Pass Al-7136.
115
Figure.4.17. Microstructure of the Solid solution plus room temperature ECAP 1-Pass
Al-7136.
116
4.1.4.4.1 Precipitates in the as-received alloy
Besides the large phase as well as smaller ’ phase observed in the TEM, the
as-received sample also contained densely distributed ultra-fine features with a size of ~5
nm and dark contrast as shown in Fig. 4.18a. Several spherical Al
3
Zr particles, with
diameters of ~25 nm, are marked with black arrows in the TEM micrograph. The
selected area diffraction pattern (SADP) along a <111>
Al
zone axis, as shown in Fig.
4.18b, contains weak diffractions from precipitates. The weak diffraction at 1/3 of
{422}
Al
, marked with a white arrow, is from GPII zones [129] and the weak diffractions
at 1/3 and 2/3 of {220}
Al
, marked with black arrows, probably come from the
precipitates [129,130]. The relatively higher intensity of GPII zone diffractions indicates
there is a significant GPII zone presence in the microstructure. The sharp diffraction spots
at 1/2{220}
Al
, marked with the two large hollow black arrows in Fig. 4.18b, are from the
Al
3
Zr particles in the matrix.
An atom probe tomography characterization of the as-received material confirmed
that small solute-rich features with a size <4 nm were resolved in the element maps. The
localized Mg-enriched regions are evident in the Mg map shown in Fig. 4.19a. These
solute-rich features are consistent with the fine features observed by TEM examination
although the size estimated from the TEM micrograph is slightly larger due to the strain
effect of coherent precipitates [131,132]. After the atoms in the matrix are removed,
these solute clusters are clearly visible with a very high number density of ~2.08 0.07
10
24
m
-3
and they are enriched with Mg, Zn and Cu as shown in Fig. 4.19b-d.
117
4.1.4.4.2 The precipitation microstructures of thermally-aged samples
After a short ageing of the as-received material at 200°C for 5 min, as shown in
Fig. 4.20a, fine platelets on {111}
-Al
dominate the microstructure with lengths of ~3-14
nm, thicknesses of ~0.6-1.2 nm and aspect ratios of length to thickness in the range from
~3:1 to ~12:1. After ageing for 20 min, shown in Fig. 4.20b, the platelet precipitates are
~3-25 nm long, ~0.6-4 nm thick and with aspect ratios in the range from ~3:1 to ~16:1.
After ageing for 40 min, as shown in Fig. 4.20c, some large precipitates are thicker and
others are longer with lengths of ~25-40 nm, thicknesses of ~1.2-5 nm and aspect ratios
in the range from ~5:1 to ~38:1.
The corresponding <110>
-Al
SADPs of Fig. 4.20a-c, as shown in Fig. 4.20d-f,
indicate that two types of precipitates, ’ and , co-exist in the microstructures. It is
known that precipitates can have up to 11 variants in Al-Zn-Mg alloys, designated
1
-
11
, with each having different orientation relationships with the matrix [133,134].
According to the orientation relationship observed in the SADPs in Figs 4.20d-f, the
precipitates in the thermally-aged samples belong only to
2
. After ageing for 5 min, the
intensity of diffraction marked by a white arrow is much stronger than the
diffraction marked by a black arrow in Fig. 4.20d. This indicates that fine precipitates
are highly dominant in the sample aged for 5 min. After ageing for 20 min, the
diffraction intensity of is stronger than for as observed by the weak diffractions near
{111}
-Al
marked with arrows in Fig. 4.20e. This implies the formation of a large
118
number of precipitates. After ageing for 40 min, platelet diffraction dominates as
shown in Fig. 4.20f, thereby indicating an additional increase in the volume fraction.
4.1.4.4.3. The precipitation microstructure and distribution in the
ECAP samples
The sample taken through 1 pass of ECAP contained platelet precipitates
primarily having low aspect ratios of ≤3:1 as shown in Fig. 4.21a. For this sample, the
largest precipitate in the matrix was ~33 0.6 nm long and ~12 0.6 nm thick with an
aspect ratio close to ~3:1. After 2 passes of ECAP, there was a significant change in the
precipitate microstructure and most of the precipitates were now equiaxed. After
additional processing through 4 and 8 passes, the precipitates remain equiaxed but their
average size increases steadily with increasing number of passes as shown in Figs 4.21b
and c. Thus, the diameters of the largest precipitates increase from ~37 2 nm after 4
passes to ~66 4 nm after 8 passes.
The indexing of weak diffractions in a <110>
-Al
SADP for the 1 pass sample, as
shown in Fig. 4.22a, confirmed that these platelets are primarily and precipitates.
However, the precipitates are dominant because the diffraction intensity of at near
2/3 of {220}
-Al
appears weaker than the intensity of the diffractions. Most of the
precipitates are of the
2
type and only a few belong to the
3
type according to the much
weaker diffraction intensity of
3
marked with a white arrow in Fig. 4.22a. One
important special feature of the reflections after 1 pass is that, as indicated by the
119
arrows, they are arced in a circumferential direction. This arcing is due to a rotation of
the precipitates relative to the matrix [82]. Interestingly, the reflections of Al
3
Zr close
to ½{220}
-Al
also deviate slightly from their typical reflection positions, indicating that
the severe plastic deformation imposed in 1 pass of ECAP rotates the precipitates
slightly from their original orientation in the matrix. It appears that, since it is difficult
for the dislocations to shear through the precipitates, there is an accumulation of
dislocations at the precipitate/matrix interfaces and this is effective in altering the
precipitate/matrix orientation.
After 4 passes of ECAP, as shown in Fig. 4.22b, most weak diffraction spots of
the precipitates scatter randomly with many spots situated, as shown in Fig. 4.22c, on the
white ring pattern of the phase but away from typical
2
and
3
reflections. This
indicates that these precipitates are phase with random orientations in the matrix. Thus,
the accumulated strain imposed by severe plastic deformation significantly alters the
orientations of the precipitates away from their initial preferred orientations. The
remainder of the diffraction spots situated between the rings of the phase, as marked
with red rings in Fig. 4.22d, are covered primarily by the diffraction rings of the S-phase
with a unit cell in the space group Cmcm, a = 0.400 nm, b = 0.923 nm and c = 0.714 nm
[135]. It is known that the S-phase is a high temperature phase in Al alloys with high Cu
content [136] and the phase is situated predominantly at the grain boundaries. It is clear
from these observations that 4 passes of ECAP causes significant fragmentation of the
large high temperature precipitates of the S-phase and increases their appearance within
120
Figure 4.18. TEM micrograph (a) and selected area diffraction pattern (SADP) (b) of as-
received Al-7136.
Figure 4.19. Element maps of an as-received Al-7136 sample, (a) Mg map, (b), (c) and (d)
are Mg, Cu and Zn map after removing matrix atoms.
As-received
Al
3
Zr
50 nm
(a) (b)
Zn Cu Mg
20 nm
(a) (b) (c) (d)
Mg
121
the matrix.Figs 4.23a-c show the relevant Mg maps for samples experiencing 1, 4 and 8
passes in ECAP, respectively, and it is apparent that the precipitate microstructure
develops significantly with increasing numbers of passes. After 1 pass in Fig. 4.23a, the
precipitates exhibit a wide size range of ~3-30 nm in their longest dimension and the
small precipitates having sizes in the range of ~3-5 nm are mostly blocky with low aspect
ratios of less than 1.6. Most of the larger received sample shown in Fig. 4.19, the
precipitates grow significantly after 1 pass of ECAP and their number density decreases
to ~1.2 0.1 10
23
m
-3
which compares with the as-received value of ~2.08 0.07 10
24
m
-3
. In the sample taken through 4 passes in Fig. 4.23b, only nine precipitates in
this condition appear elongated or platelet with thicknesses of ~5-10 nm and lengths
or diameters of ~15-40 nm. By comparison with the small solute-rich features of the as-
precipitates were observed in the analysed volume giving a number density of ~1.6 0.5
10
22
m
-3
which is nearly one order of magnitude lower than in the 1 pass sample and
two orders of magnitude lower than for the as-received condition. All precipitates are
equiaxed after 4 passes and the largest precipitates have a diameter of ~20 nm which is
smaller than the maximum diameter of ~30 nm observed by TEM on a larger scale in Fig.
4.21b. In addition, the smallest precipitates are ~8 nm in diameter which is larger than
the diameter of ~5 nm observed in the 1 pass sample. Interestingly, no small debris with
a size smaller than ~5 nm was observed in the microstructure.
In the 8 pass sample shown in Fig. 4.23c, the precipitate number density drops
significantly to ~2.3 1.3 10
21
m
-3
and there is further growth of large precipitate to
122
Figure 4.20 . TEM bright field (BF) images and corresponding SADPs of Al-7136
samples thermally aged at 200°C up to 40 mins. (a), (b) and (c) are BF images and (d),
(e), (f) are SADPs aged for 5, 20 and 40 mins, respectively.
5 min
(a)
20 nm
20 min
(b)
40 min
(c)
(d)
(f)
(e)
(f)
123
Figure 4.21 TEM BF micrographs of (a) 1-pass, (b) 4-pass and (c) 8-pass ECAP samples
of a Al-Zn-Mg-Cu (7136 ) alloy.
(a)
(b)
(c)
124
~40 nm. Examination by TEM, as shown in Fig. 4.21c, gave a similar precipitate
evolution with a largest size of ~70 nm and a smallest size of ~15 nm. Again, no small
debris with sizes <5 nm was observed in the analyzed volume.
The chemical compositions of the larger precipitates formed at different numbers
of passes are listed in Table 4.3 and it is apparent these precipitates are enriched with Mg,
Cu and Zn. The precipitates are not aluminium-free with 7-8 at% Al, 8-11 at% Cu, 42-47
at% Zn, 34-39 at% Mg and a ratio of (Zn+Al+Cu)/Mg 2. This contrasts with 8.45 at%
Al, 59.05 at% Zn and 32.46 at% Mg of precipitates in the Al-7108 alloy (T7) with a
trace Cu concentration of 0.007 at% measured by 3DAP [137] and Zn/Mg = 2 for
equilibrium in the ternary Al-Zn-Mg system [133]. It is probable that Al and Cu
primarily occupy some of the Zn sites in the typical phase (Zn
2
Mg) and they have
certain solubilities in the precipitation phase in the quaternary Al-Zn-Mg-Cu system. The
solute concentrations in the matrix decrease with increasing numbers of passes as shown
in Fig.4.24a and this matches the decrease in their number density as shown in Fig. 4.24 b.
This suggests that the precipitate evolution is in the growth regime.
4.2. High Pressure Torsion on Al-7136
4.2.1 Microhardness study of Al-7136 after HPT
A series of patterned Vickers hardness tests were conducted on samples
subjected to HPT at room temperature, for 1/8, 1/4, 1 , 2 and 4 turns, respectively.
The microhardness measurements were plotted in the form of color-coded
contour maps as shown in Fig.4.25, hardness test has been conducted following a grid
125
Figure 4.22. SADPs of Al-7136 samples experienced (a) 1-pass and (b) 4-pass ECAP at
200 °C, (c) the SADP of 4-pass sample superimposed with white artificial diffraction
rings of the precipitate and (d) the 4-pass SADP superimposed with red artificial rings
from S-phase.
4-pass
(a) (b)
1-pass
111
3
2
2
4-pass 4-pass
(c) (d)
126
Figure 4.23.Mg maps of Al-7136 alloy samples experienced with (a)1-pass, (b) 4-pass
and (c) 8-pass ECAP.
20 nm
(a) (b)
(c)
127
pattern with 0.3mm interval between adjacent points. The significance of the colors in
Fig.4.25 is defined explicitly in the small inset at the lower right and thus the plots
provide a direct and visual representation of the sample homogeneity.
Figure 4.25 a) depicts Al-7136 subjected to 1/8 turn of high pressure torsion at
room temperature with a pressure of 6 GPa. As shown in the color-coded map, the
majority of the test results fall into the range between 200-220, with a weaker region in
the center of the disk, where the average Vickers hardness is around 180-190. This
observation is consistent to earlier reports on the homogeneity of Al alloys, and could be
explained by the equivalent strain calculation of HPT, where the larger distance to the
center of the disk corresponds to a larger imposed strain.[28] It is also worth to mention
the softer region in the edge part of the disk, this inhomogeneity is probably related to the
very low turns of HPT: with a very short time of processing ( ~8 s) and comparatively
low imposed strain, the experimental result might be somehow off the theoretical
calculation result.
The Vickers hardness distribution of the sample subjected to ¼ turn of 6 GPa
room temperature compared to the one subjected to 1/8 turn of HPT, what is more
apparent is the shrinking of the soft regions: the outer softer region diminished as a result
of the larger imposed strain and longer processing time, what is more, the center softer
region is smaller, which indicated an overall homogeneity improvement over the whole
disc. This conclusion is again consistent to earlier reports on aluminum alloys,where the
area of the center soft region decreased as the turns of HPT increased. This could be
explained by the theoretical calculation of the imposed strain as well.
128
When N increased to 1, which corresponds to 1 turn of HPT at room temperature
at the pressure of 6 GPa, the Al-7136 disc was much more strengthened. From the plotted
contour map in figure.3 c), the Vickers microhardness of the whole disc was substantially
improved to ~250 for the majority of the disc, there is still a soft region in the center of
the disc, where the average hardness value is around 200-220. The maximum Vickers
hardness was found to be ~280 in this sample, which claimed very good strengthening
effect of HPT up to 1 pass for Al-7136 alloys.
As indicated by the contour map of the disc subjected to 2 turns of HPT at 6 GPa
in Figure 4.25 d), the measured Vickers microhardness changes in a different way:
instead of further strengthening , the overall microhardness decreased to ~ 240, the
decrease is more substantial at the outer region of the disc, which corresponds to a larger
imposed strain from the theoretical calculation. The center soft region of the disc
remained mostly unchanged as to the size as well as the hardness.
Figure 4.25 d) showed the hardness contour map of the Al-7136 disc subjected to
3 turns of HPT at the pressure of 6 GPa and room temperature, the trend of the hardness
decrease continued but is not as substantial as from ¼ turn sample to 1 turn sample. The
lowest Vickers microhardnees measured on the disc was around 200 and at the edge part
of the disc.
When N increased to 4, which corresponds to 4 turns of HPT at the pressure of 6
GPa and room temperature, the color-coded contour map was shown in the figure 4.25 d) .
It could be concluded from the map that the Vickers microhardness kept decreasing to
the saturated value to around 200, which means, for those parts with hardness more than
129
Zn, at% Mg Cu Al (Zn+Cu+Al)/Mg
1‐pass 45.4 0.5 38.6 0.4 8.6 0.1 7.4 0.1 1.6
4‐pass 42.6 0.5 38.4 0.3 11 0.2 8 0.2 1.6
8‐pass 49.5 0.4 34.6 0.3 8.1 0.1 7.7 0.1 1.9
Table 4.3 The compositions of large precipitates formed under different passes ECAP in
the Al-7136 alloy.
Figure 4.24. Matrix solute concentration evolution (a) and precipitate number density
evolution (b) in the Al-7136 alloy as a result of different number of ECAP processing.
1
10
100
1000
10000
Number density x 10
21
, m
-3
0-pass
1-pass
4-pass
8-pass
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Mg Cu Zn
Solute content, at%
as-recieved
1-pass
4-pass
8-pass
(a) (b)
0‐p 1‐p 4‐p 8‐p
130
200, there is a trend that the hardness kept decreasing; for those parts with the hardness
close to 200 , there is no further decrease of the hardness found.
To better understand the Vickers hardness contour maps and to evaluate the
extent of homogeneity quantitatively, histograms showing the number fraction of
measurement points having different values of the Vickers microhardness within
incremental segments of 10 Hv were constructed for the samples subjected to 1/8, ¼ ,1,2,
3, and 4 turns of HPT, respectively. This approach is illustrated in Figure 4.26 a) to e)
and the average of Hv is denoted in each of the plots by a vertical line, and plotted in
figure 4.26 f), furthermore, the average of Hv of the as-received Al-7136 was marked by
the dash line in figure 4.26 f) for comparison.
It is obvious that the average measured Vickers microhardness in all discs
subjected to HPT were higher than that of the as-received sample, HPT thus could
effectively strengthen Al-7136 alloys with even 1/8 turns of HPT at room temperature
and the pressure of 6 GPa. Additionally, as discussed above, for HPT up to 1 turn, the
disc were overall strengthened with the average Hv increased from ~210 in the 1/8 turn
sample to ~215 in the ¼ turn sample, and finally ~250 in the 1- turn sample.
With more than 1 turn of HPT, in this study , starting from 2 turns of HPT , there
is trend of hardness decreasing rather than increasing, in 2 turn sample, the average
measured Hv dropped to around 240, this value further dropped to around 235 in the 3-
turn sample, and around 230 in the 4-turn sample. This softening of the materials is quite
unique thus is interesting to study from the microstructure perspective.
131
Figure.4.25. Vickers Hardness (HV) distribution on Al-7136 discs after room temperature
HPT at 6 GPa.
(a) (b)
(c) (d)
132
Figure.4.25.Continued. Vickers Hardness (HV) distribution on Al-7136 discs after room
temperature HPT at 6 GPa.
(e)
(f)
133
One thing also worth mentioning is the hardness distribution information from the
hardness histogram in different samples. There is a narrowing in the width of the
distribution in Fig.4.26 b) in comparison with Fig.4.26 a) and this demonstrates directly
the greater homogeneity achieved when the numbers of turns increased from 1/8 to ¼.
However, this trend did not continue from ¼ turn sample to 1 turn sample, there is a
wider spread of the hardness distribution in the 1 turn sample in comparison with the ¼
turn sample. This hardness distribution remained similar in the 2- turn sample as well as
the 3- turn sample, then again narrowed in the 4-turn sample. The best homogeneity
achieved in all these samples is at the ¼ turn sample, where a largest number fraction of
more than 70% falls into the Vickers microhardness in between 210-220.
4.2.2. Microhardness study of Al-7136 after 1 pass room temperature
ECAP plus HPT processing
It has been shown that high pressure torsion could effectively strengthen Al-7136
alloy even with 1/8 turns at the pressure of 6 GPa according to the Vickers microhardness
test results. Considering the saturation grain size is smaller in HPT than in ECAP , there
is the possibility of introducing additional grain refinement by processing using ECAP
and then cutting disks and further processing by HPT. This type of approach was
confirmed using pure Ti where the grain size after ECAP was ~300 nm but this reduced
to ~200 nm through additional processing by HPT [138] . The general tendency for
achieving smaller grains in HPT is also confirmed by noting that, whereras the grain sizes
134
Figure.4.26. The histogram of the Vickers Hardness (HV) distribution on Al-7136 discs
after room temperature HPT at 6 GPa.
Al-7136
HPT: 6 GPa, 1/8 turn
Vickers M icrohardness ( Hv)
140 160 180 200 220 240 260 280
Frequency ( Percentage)
0
10
20
30
40
50
60
70
80
90
100
Vickers Microhardness ( Hv)
140 160 180 200 220 240 260 280
0
10
20
30
40
50
60
70
80
90
100
Frequency ( Percentage)
Al-7136
HPT: 6 GPa, 1/4 turn
(a) (b)
V icke rs M ic rohard ness (H v )
140 16 0 1 80 200 220 240 26 0 2 80
Frequency (Percentage)
0
10
20
30
40
50
60
70
80
90
100
A l-7136
H P T : 6 G P a, 1 tu rn
V icke rs H a rdness ( H v)
14 0 160 18 0 200 2 2 0 2 4 0 260 2 8 0
Frequency ( Percentage)
0
10
20
30
40
50
60
70
80
90
10 0
Al-7 1 3 6
H P T : 6 G P a, 2 tu rn s
(d) (c)
135
Figure.4.26.Continued. The histogram of the Vickers Hardness (Hv) distribution on Al-
7136 discs after room temperature HPT at 6 GPa.
HPT Turns
.125 .250 1.000 2.000 3.000 4.000
Average Vickers Microhardness (Hv)
140
160
180
200
220
240
260
Al-7136
HPT: 6 GPa, R.T.
(g)
V ickers M icrohardness ( H v)
140160 180200 220240 260 280
Frequency ( Percentage)
0
10
20
30
40
50
60
70
80
90
10 0
Al-7 1 3 6
H P T: 6 G P a, 3 turns
V ic k ers M ic ro h a rdn es s ( H v )
14 0 160 180 2 0 0 220 24 0 2 60 280
Frequency ( Percentage)
0
10
20
30
40
50
60
70
80
90
100
Al- 7 1 3 6
H P T : 6 G P a, 4 tu rn s
(e) (f)
136
attained in ECAP area usually within the submicrometer range of 100-1000 nm, there are
now many reports documenting true nanometer grain sizes in various Al alloys[139,140],
Cu alloys [110] and much more.
Accordingly, the combination of ECAP and HPT for Al-7136 is also tempting in
this research. Previous investigation in ECAP found that although ECAP could be
successfully conducted on Al-7136 for up to 8 passes with no cracking and segmentations
at 473 K, the overall mechanical properties is rather deteriorated. However, the new
approach of conducting solid solution treatment on Al-7136 first, followed by water
quenching and immediately room temperature ECAP pressing could effectively improve
the strength. It is thus more promising to employ the room temperature ECAP in this
part of research to achieve finer grains and possibly, better strength.
The as-received Al-7136 were then solid solution treated at 743 K for 1 hour ,
water quenched and pressed by ECAP facility at room temperature for 1 pass. Disks were
sliced from the center of the as-pressed billets and carefully polished into decent
thickness for HPT processing. Again the pressure of HPT was selected at 6 GPa , the
processing were carried out at room temperature. 1 turn and 2 turn samples were
successfully achieved following this processing routine.The HPT processed disks were
polished into mirror-like surface and tested using the Vickers microhardness
machine. The results in the form of color-contoured maps were shown in Figure.4.27.
For better comparison , the 1 turn and 2 turn HPT only samples were replotted into the
same color scale as for these two samples. Vickers microhardness distribution
histograms were also plotted for each of the samples.
137
An immediate conclusion would be the pre-processing of room temperature
ECAP could effectively further strengthen the Al-7136 alloy. For 1 turn HPT only
sample ,the average Hv is around 252, this value siginificantly increased to around 275
for the one with room temperature ECAP pre-processing; for 2 turn HPT only sample, the
average Hv is around 242, again the Vickers microhardness increased in the sample with
ECAP pre-processing and HPT, the value is about 270. The more interesting , or less
intuitive result lies on the color-coded contour maps part, the hardness evolution in the
HPT only samples have been discussed in the previous results and concluded that the
hardness trended to decreased after 1 turns of HPT, however, this is not quite the truth for
the samples with combined room temperature ECAP and HPT processing, the vicrkcers
hardness evolved in two different fashions depended on the position of the tested point,
i.e. the distance of the tested point from the center of the disk. For the inner region
within 2/3 of the radius, the hardness increased drastically, by an average of around 40
Hv; However, for the remaining part of the disk , i.e. the outer ring shape region, the
hardness decreased by the margin of 40 Hv. It has also to be mentioned that the central
soft region almost diminished after 2 turns of HPT.
The hardness distribution histogram in another aspect, show some valuable
information about the homogeneity of the disks subjected to the processing. For the ones
with HPT only processing , the 2-turn sample indicated a slightly narrower spread , thus
better homogeneity compared to the 1-turn sample, this is also true for the ones with
room temperature pre-processing as well as HPT processing , one more turn of HPT also
narrowed the spread of the hardness distribution. Comparing the histograms of the
138
samples with and without room temperature ECAP pre-processing , it is found that the
additional ECAP pre-processing resulted in a much wider spread of the hardness
distribution: for the 1-turn samples, the HPT only one covered the histogram from 220 to
280, a 60 Hv range, but for the one with ECAP and HPT, the hardness distributed from
220 to 310 with a 90 Hv range; for the 2-turn samples, the HPT only one covered a Hv
range of 50 from 210 to 260, this range increased to 90 in the one with ECAP from 150 to
240.
4.2.3 Microhardness study of Al-7136 after 4 passes ECAP at 473 K
plus HPT processing
It could be concluded that high pressure torsion is an effective method in
strengthening the 7000 series Al alloys, the adding of room temperature ECAP for 1 Pass
in the processing routine could further strengthen the material. It could be easily assumed
that the additional straining from room temperature ECAP would cause additional work
hardening and possibly further grain refinement that would contribute to the strength
increase. The hardness evolution from 1 turn to 2 turn samples , onthe other hand, is
more complicated and less understandable. For HPT only samples , there is only a sole
trend of overall hardness decreasing from 1 turn to 2 turn samples regardless of the
position of the testing points, while for room temperature ECAP and HPT samples, the
evolution is more complicated and dependent on the distance from the center of the disc,
the inner part continued being strengthened but for the outer ring shape region, there is a
significant hardness decreasing.
139
(c) (d)
Figure.4.27. The contour maps and histograms of the Hv distribution on Al-7136 discs
after room temperature HPT at 6 GPa.
(a)
V ic k e r s M ic r ohar dnes s ( H v )
180 200 220 240 2 6 0 2 80 3 0 0 3 20
Frequency (Percentage)
0
10
20
30
40
50
60
70
80
90
10 0
A l - 7136
H P T : 6 G P a, 1 tu r n
(b)
(c)
V icke rs H a rd n e ss ( H v )
180 200 220 240 260 280 3 0 0 320
Frequency ( Percentage)
0
10
20
30
40
50
60
70
80
90
100
Al - 7 1 3 6
H P T : 6 G P a, 2 tu rn s
(d)
140
(g) (h)
Figure.4.27.Continued. The contour maps and histograms of the Hv distribution on Al-
7136 discs after room temperature HPT at 6 GPa.
(e) V ic k e rs H a rd n e s s ( H v )
18 0 2 0 0 220 240 2 6 0 2 80 3 0 0 3 2 0
Frequency ( Percentage)
0
10
20
30
40
50
60
A l - 7136
E C A P : S o lid S o lu tion T = 74 3 K ,
1 h + 1 P a s s ,R .T .
H P T : 6 G P a , 1 tu rn
(f)
(g)
V icke rs H a rdness ( H v )
18 0 200 220 240 26 0 2 8 0 300 320
Frequency ( Percentage)
0
10
20
30
40
50
60
Al- 7 1 3 6
EC AP: So lid So lu tio n T = 7 4 3 K ,
1 h + 1 P a ss,R .T .
H P T : 6 G P a, 2 tu rn s
(h)
141
The differences on the hardness evolution ,by assumption, is related to the grain
size and precipitates. It is already observed that room temperature ECAP by only 1 pass
would change the microstructure and the precipitate morphologies significantly, thus
would affect the combined influence on the hardness evolution. It is then becoming
tempting to study if the variation of the parameters of ECAP would affect the hardness
evolution.
For this reason , a different ECAP pre-processing was designed, the Al-7136
rods were pressed by ECAP facility at the elevated temperature of 473 K for 4 Passes.
The HPT processing was identical to that for the previous studies. Vickers
microhardness test were again taken on the well polished disks. The color-coded contour
maps shows the result in figure.4.28.
The first impression would be the overall strength of the material is much lower
than the one with room temperature ECAP pre-processing, the average Hv in the Al-7136
disk subjected to 473 K ECAP pre-processing and 1 turn HPT is about 212, compared to
~274 in the one with room temperature ECAP and 1 turn HPT. This is also true for the 2
turn samples, average Hv of the sample with 473 K ECAP and 2 turn HPT is around 220,
while this value is around 268 in the one with room temperature ECAP pre-processing.
In fact , the average hardness of the samples with 473 K ECAP pre-processing and HPT
is even lower than the counterparts with HPT processing only: the two values are 252
and 242 for 1 turn and 2 turn HPT samples, respectively. This result is fully expected, as
in the ECAP investigation , it was found that ECAP at 473 K for 4 passes would lead to a
much lower strength compared to the as-received material due to the coarsen precipitates
142
generated during the thermal-mechanical processing, but room temperature ECAP for 1
pass would strengthen the material benefiting from the slower kinetics at lower
temperature thus less and smaller precipitates. As a result, Al-7136 had completely
different microstructure before the HPT processing. Despite the generally strengthening
effect from the room temperature HPT processing, the strength softening was inherited
from the ECAP processing thus lead to a comparatively lower strength in the final 4-Pass
and 1-turn samples.
The hardness evolution is another focus. It is found from the color-coded contour
maps that the hardness evolutes differently in the samples with HPT only and room
temperature ECAP plus HPT. There is an overall hardness decrease for the HPT only
samples regardless of the position on the disk, but the for the latter one, the inner region
in the disk got further strengthened while the outer region was softened instead. In this set
of samples, unlike either of the two sets of the samples, it showed a hardness saturation in
the 1 turn disk at the value of ~220-230, the disk did not get further strengthened beyond
that level in the 2-turn disk , but the center soft region was much smaller and closer to the
saturation hardness. This also could be found from the hardness distribution histogram,
where the 1 turn sample showed a much wider spread due to the large center soft region,
it is much narrowed in the 2 turn sample, with more than 60% results fell into the 220-
230 range. It is then concluded ECAP at 473 K for 4 passes plus 2 turn HPT would
improve the homogeneity of the disk.
143
4.2.4. Microstructures of Al-7136 after HPT processing
Other than the microhardness study over the HPT disks, microstructure study is
very important in understanding the influence of HPT on the material. It also provides
valuable information in understanding the mechanical behaviors of the materials. For
materials subjected to severe plastic deformation , the grains are usually very small ,
generally within the ultrafine grain range (100 -1000 nm); High pressure torsion could
produce even smaller grain size compared to ECAP, sometimes the materials subjected to
HPT are true nano materials with the grain size within 100 nm. Transmission electron
microscopy is the most popular tool in the microstructure study of severely plastically
deformed materials due to its outstanding resolution.
In HPT processed materials, due to the nature of HPT, the strain imposed on the
material is dependent on the position on the disk, i.e. the distance from the center of the
disc. It also could be observed from the hardness test result that the center part and the
peripheral part sometimes showed different hardness evolution trends. In this study , the
disk is partitioned into 2 parts, the center part and the peripheral part for TEM
investigation purpose. The center part is defined as the inner disk with a radius as 1/3 of
that of the HPT disk, the peripheral part is defined as the ring-shaped region outside the
center part on the HPT disk. For the center part TEM samples, the 3 mm diameter disc
were punched from the center region after careful polishing. For the peripheral part
samples , the thin disc were punched from the ring-shaped region. All these partitions are
schematically shown in Figure.4.29 . However, It has to be pointed out that due to the
144
(a) (b)
(c) (d)
Figure.4.28. The contour maps and histograms of the Hv distribution on Al-7136 discs
after 4 passes ECAP at 473 K plus HPT processing
120 140 160 18 0 2 0 0 2 2 0 2 40
0
20
40
60
80
100
Al - 7 1 3 6
EC AP: 4 Pa s s , T = 4 7 3 K
HP T : 6 G P a , 2 tu r n s
Frequency ( Percentage)
V ick e r s M ic ro h a rd n e s s ( H v )
V icke rs M ic ro h a rd n e ss (H v)
12 0 1 4 0 1 6 0 1 80 2 0 0 220 240
0
20
40
60
80
10 0
Frequency ( Percentage)
A l- 713 6
E C A P :4 P a s s , T = 473 K
H P T : 6 G P a, 1 tu rn
(a)
(b)
(c) (d)
145
intrinsic limitation of using electropolishing method in the final preparation of TEM
samples, there is no way to expect even samples in the same region observed with the
identical distances from the center of the disc. Despite the limitation, this kind of partition
still benefited in better understanding the microstructure evolution of the HPT disks and
would help in explaining the hardness result.
4.2.4.1.Microstructure of the as-received Al-7136
TEM images of as-received Al-7136 were shown in the ECAP investigation part,
it was found that the as-received alloys contained a reasonably equiaxed array of grains
with a comparatively low dislocation density. The average measured grains is about 3.4
m.The plate shaped precipitates were observed within the grains, with the length of
around 600 to 800 nm and widths of around 200 nm; the needle-like ’ phases were
much smaller , with the length of less than 40 nm, mostly lying on the grain boundaries.
For the 1 turn HPT sample, TEM investigations were both done on the center and
the peripheral part as shown in Figure 4.30 and 4.31. For the center part, the grains are in
almost equiaxed shape with the average grain size of around 200 nm, there still are some
elongated grains with the lengths of about 250 nm and widths of about 100 nm. The
contrast within the grains suggests a high dislocation density within the grains. There are
aligned precipitate arrays with the precipitate size of ~ 10-20 nm marked by the white
arrow in Figure.4.30 b). The alignment of small precipitates usually is considered as the
evidence of fragmentations of the large precipitates, which were observed in the as-
146
(a) Center Part of the 1 turn HPT sample
Figure.4.30 TEM images of the center part of the 1 turn HPT Al-7136 sample
Center Part
Peripheral Part
Figure.4.29 Partition of the HPT disk for TEM investigation purpose
TEM Sample
147
(b) Center part of the 1 turn HPT sample
(c) Center part of the 1 turn HPT sample
Figure.4.30.Continued. TEM images of the center part of the 1 turn HPT Al-7136
sample
148
received Al-7136 also and characterized as phase. These fragmentations possibly will
serve as barriers for the dislocation motions. Because the dislocations could not cut
through the very small size precipitates, they will pile up against the fragmentations of
the large precipitates and eventually form grain boundaries. In Figure 4.30 c), the white
circle marked the triple junctions of the grain boundaries with the small precipitates
pinning the grain boundaries.
In the 1 turn peripheral part, the grains are in better equiaxed shape with the
average measured grain size of 120 nm, which is smaller than that of the grains in the
center part. There are no aligned precipitate arrays in the size of ~ 10 nm observed.
However, there are slightly larger precipitates distributed along the grain boundaries.
This difference between the center part and the peripheral part is related to the difference
of the imposed strain as discussed in earlier chapters. The peripheral part was subjected
to a larger imposed strain compared to the center part, If we assume the debris from the
fragmentation of the large precipitates still existed , the more severe straining would
place the debris in more random positions with no obvious alignment, similar thing is
observed for ECAP samples, where despite precipitate alignments were observed in the
1-pass samples, a much more random distribution of the precipitates was observed in
samples subjected to further ECAP processing. These less agglomerated precipitates
would thus provide more barrier sites for dislocation motions. Meanwhile, the higher
imposed strain would increase the dislocation density of the materials, these dislocations
again piled up against the precipitates and formed smaller grains.
149
Figure.4.31 TEM images of the peripheral part of the 1 turn HPT Al-7136 sample
(a)
Figure.4.32 TEM images of the center part of the 2 turn HPT Al-7136 sample
150
(b)
Figure.4.32.Continued.TEM images of the center part of the 2 turn HPT Al-7136 sample
Figure.4.33 TEM images of the peripheral part of the 2 turn HPT Al-7136 sample
151
The grain size for either the center or the peripheral part of the 1-turn sample is
much smaller compared to the as-received coarse grained materials. It is therefore
reasonable that Al-7136 after 1 turn HPT showed a higher hardness than the as-received
materials.In 2 turn samples, for the center part, compared to the 1 turn center part TEM
images, the grains are in better equiaxed shape suggesting a further straining from the
HPT processing. The average measured grain size is around 150 nm , which is smaller
compared to the 1 turn counterpart. There still are a lot of precipitates existing and
distributed mainly along the grain boundaries, but the size of these precipitates is larger ,
in the range of 25-30 nm. As to the peripheral part of the sample, the measured average
grain size is about 100 nm, again slightly smaller than 1 turn peripheral part. There are
precipitates in the same size as those in the center part, and also distributed mainly on the
grain boundaries. These images are suggesting a growing or agglomeration of the fine
precipitates on the grain boundaries from 1 turn sample to 2 turn sample.
In 4 turn samples, despite the TEM images were taken both on the center and
peripheral part, the center part showed a much larger 500-700 nm grain size, indicating a
dichotomy to the hardness contour map. This is probably from the limitation in the TEM
sample preparation method and the nature of HPT processing, theoretically the strain is
always zero no matter how many turns of HPT were conducted on the sample,
therefore if the thin area of the TEM specimen is too close to the center of the HPT
disc, a much less strained part would be observed with comparatively coarse grains .
152
(a)
(b)
Figure.4.34 TEM images of the peripheral part of the 4 turn HPT Al-7136 sample
153
(c)
Figure.4.34.Continued.TEM images of the peripheral part of the 4 turn HPT Al-7136
sample
154
For the peripheral part, the grains size fell into two categories, there are smaller
grains with the average measured grain size of around 90-100 nm , as well as the larger
grains with the average measured grain size of around 170 nm, the maximum grain
size observed is around 300 nm, the grain size of the larger size group is obviously larger
even than that of the 1 turn sample. This suggests a dynamic recovery or
recrystallization of the grains as no large grains were observed in 1 turn or 2 turn samples
no matter for the center or peripheral part.
It is not only the grain size that varies in the peripheral part of the 4-turn sample,
so is the morphology of the precipitates. The precipitates kept evolving from 2 turn to 4
turn samples, however, unlike in previous TEM images, there are also two different
kinds of precipitate morphologies, the precipitates still distributed along the grain
boundaries, but in two different sizes , for most of the precipitates, the precipitate
diameter is around 10 nm, which is close to the size of the precipitates in the 1 turn
sample but smaller than that in the 2 turn sample; for a very small portion of the
precipitates, the sizes are abnormally large – close to 40 nm, and these large precipitates
are connected one by one along the grain boundaries.
4.2.4.2 Microstructure of Al-7136 subjected to R.T. ECAP plus HPT
In the ECAP part, the advantages of ECAP at room temperature has been
discussed and it is agreed that despite the difficulty of processing , room temperature
ECAP could effectively control the precipitates occurrence and lead to a smaller grain
155
size compared to ECAP at elevated temperature. Also in the hardness test part, it has
been proven that the pre-processing of ECAP at room temperature of 1 pass could
effectively enhance the strengthening effect from room temperature HPT. TEM study on
the HPT samples with room temperature pre-processing would reveal the reason for the
better strengthening.
In the ECAP plus 1 turn HPT center part, surprisingly, the grains are in less
developed shape compared to the one without ECAP pre-processing. A measurement
concludes the grains with an aspect ratio of about 2:1, the length and widths are about
800 nm and 400 nm , respectively. This is smaller than the one with room temperature
ECAP only, but much larger than that of 1 turn HPT only. The development of the finer
grain structure is more difficult for the samples subjected to 1 pass room temperature
ECAP. The precipitates are in similar morphology and distribution as that in the 1 pass
room temperature ECAP sample.
The microstructure of the peripheral part of the 1 turn sample looks a lot
different to the center part, the grains are in well-defined equiaxed shape and much
smaller. The average measured grain size is about 80 nm, this indicates that it is possible
to use room temperature plus HPT processing to achieve true nanostructured Al-7136.
The precipitates are in very small size ( around 10 nm) and distributed on the grain
boundaries, this is consistent to earlier observations.
The TEM image of the ECAP plus 2 turn HPT center part is similar to that one
of the ECAP plus one turn peripheral part, the grains are in equiaxed shape with an
average measured grain size of ~70-80 nm, there are also small precipitates distributed on
156
Figure.4.35 TEM images of the center part of R.T.ECAP plus 1turn HPT Al-7136
sample
(a)
Figure.4.36 TEM images of the peripheral part of R.T.ECAP plus 1 turn HPT Al-7136
sample
157
(b)
Figure.4.36.Continued. TEM images of the peripheral part of R.T.ECAP plus 1 turn HPT
Al-7136 sample
Figure.4.37 TEM images of the center part of R.T.ECAP plus 2 turn HPT Al-7136
sample
158
(a)
(b)
Figure.4.38. TEM images of the peripheral part of R.T.ECAP plus 2 turn HPT Al-7136
sample.
159
the grain boundaries with the diameter of around 10 nm. There are no precipitates visable
within the grains.
The grain size is extremely small in the peripheral part of the ECAP plus 2 turn
HPT sample, the measured average grain size is less than 40 nm, almost 1/2 of the size of
the already very fine grain size in the ECAP plus 1 turn HPT peripheral part.
Interestingly , the precipitates are in the similar size , with the diameter about 10 nm,
there are even more precipitates on the GBs. Considering the increased GB to volume
ratio, the precipitates are much more than that with larger grains. This is more clear to
look at Figure 4.38 b), where obviously inhomogeneous straining occurred. There is a
fine grain zone with the grain size of 40 nm in between two comparatively coarse grain
zone with the grain size of about 80 nm, it is clear that in the coarse grain zone has fewer
precipitates on the grain boundaries than the fine grain zone, despite the precipitates are
in similar size.
160
5. DISCUSSION
5.1 The effect of Severe Plastic Deformation on precipitates in Al-7136
Despite processing Al-7136 at 473 K is feasible and easy to conduct at the
temperature of 473 K, including the production of billets with smooth surfaces after the
pressing operation, the nature of this super-saturated alloy determines the temperature
sensitiveness when processing by ECAP. This approach leads generally to inferior
mechanical properties for Al-7136 alloy. A detailed microstructure study of the Al-7136
alloy is given in chapter 4 for samples processed by ECAP at 473 K. However, these
results show that, although the grains are significantly refined from ~3.4 m to ~300-400
nm in samples processed through 8 passes at 473 K, the precipitate morphologies are also
severely influenced by the pressing operation. Consequently, the effect of performing
ECAP on the Al-7136 at an elevated temperature is similar to the results reported earlier
for the Al-7034 alloy[78], where high temperature pressing led to a reduction in the
overall strength of the material.
The strength of a supersaturated polycrystalline Al alloy generally constitutes of
three parts: the grain size part, the solid solution strengthening part and the second phase
or precipitate strengthening part. The well known Hall-Petch equation concludes that the
yield strength of a polycrystalline materials is proportional to the inverse square root of
the average grain size, a smaller grain size would correspond to a higher strength of the
material. However, this is obviously not the case for Al-7136 alloy processed by ECAP
161
at 473 K , the grains are almost equiaxed in the 8 pass sample with a ultrafine grain size
of around 300-400 nm , but the yield strength of the 8 pass sample is ~250 MPa, 100
MPa less than that of the coarse-grained counterpart. Consequently, the overall softening
behavior is more dependent on the last two factors, the solid solution part and the second
phase strengthening part.
As an age-hardenable Al alloy, Al-7136 contains excessive alloying elements:
there are 9.4 wt% Zn , 2.5 wt% Mg and 2.5 % Cu in Al-7136 . This complicated
composition would consequently introduce a very complicated precipitation process. For
Al-Zn-Mg-Cu alloys, during ageing treatments, the precipitation process follows a fixed
transformation sequence dominated by the precipitation kinetics and thermodynamics so
that the precipitates generally have a simple orientation relationship with the matrix and
this determines their unique morphology such as rods or platelets [134].
The evidence to date suggests that processing by ECAP has a relatively
complicated effect on precipitate evolution. when the ECAP is conducted at higher
temperatures, such as 473 K, precipitation is promoted in an Al-Zn-Mg alloy [79]. In an
Al-Zn-Mg-Cu (Al-7034) alloy, processing by ECAP at 473 K produced spherical
precipitates by fragmentation of the pre-existing larger platelet precipitates [21]. To date,
only fragmentation of the larger precipitates has been identified as a mechanism
dominating precipitate evolution for materials containing pre-existing large precipitates
[21,82] and accordingly it is not clear at present whether other mechanisms may be
involved and whether they are equally important in precipitate evolution. Thus, a more
162
complete understanding of the precipitation behaviour during ECAP is an essential
prerequisite for successfully understanding the unique mechanical behavior of Al-7136
alloy.
5.1.1. Precipitation during ECAP of the Al-7136
According to Chapter 4 ,after 1 pass of ECAP at 473 K, the precipitation
microstructure evolves from a mixture of fine coherent GP zones and precipitates in
the as-received alloy to a mixture dominated by platelet but with some remaining .
In addition, only the precipitates are present after 4 and 8 passes. This observation
shows that is stable at 473 K in the alloy and this agrees with DSC measurements
showing thermal stability of in an Al-Zn-Mg (Al-7050) alloy in the temperature range
from 473 to 510 K [141,142]. Among the 11 variants of precipitates designated as
1
-
11
[133,134],
2
platelets are dominant after 1 pass according to the SADP shown in Fig.
4.22(a). In practice, it should be noted that
2
and have similar orientation
relationships with the matrix and similar morphologies and it is generally believed that
2
platelets are formed by direct transformation of [134]. On the basis of the
precipitation observed in different ECAP samples, it is concluded that precipitation
during ECAP probably follows the same sequence as in a conventional ageing treatment
(i.e. GP zones ).
163
5.1.2. The effect of ECAP on the precipitation kinetics
The precipitation microstructures formed in ECAP are significantly different from
those formed in thermal ageing based on the TEM and APT characterisations. According
to the relative diffraction intensities observed near 2/3 of {022} in the SADPs shown in
Figs 4.20d and 4.22a, the equilibrium precipitates are dominant after 1 pass of ECAP
whereas metastable intermediate precipitates are dominant after thermally ageing for 5
min despite the fact that 1 pass of ECAP has a similar effective thermal duration.
Furthermore, the size of the precipitates in the ECAP sample is much larger than in the
thermally-aged sample. The larger plates after 1 pass of ECAP are up to ~12 nm thick
and ~33 nm long with an aspect ratio <3:1 whereas the platelets in the sample
thermally-aged for 5 min are <1.2 nm thick and <14 nm long with aspect ratios of ~3:1-
12:1. As the precipitates are
2
-type, they should be in the form of hexagonal or rounded
plates [143]. Based on these precipitate dimensions, the volume of a large precipitate
after 1 pass of ECAP is estimated as more than 50 times larger than in the sample
thermally-aged for 5 min. Thus, from the size differences of the largest precipitates
formed under these two conditions, precipitation under ECAP is estimated to occur
approximately 50 times faster than in conventional thermal-ageing at 473 K. This is in
agreement with an earlier report suggesting that SPD promotes the precipitation process
[79].
The high density of mobile dislocations plays an important role for precipitate
evolution during ECAP. Firstly, the strain effect of mobile dislocations, in combination
164
with the increased interfacial areas due to dislocations shearing through the coherent
precipitates, effectively promotes the dissolution of small metastable precipitates [144].
Secondly, the dislocations serve as fast diffusion paths for solute atoms in the matrix,
thereby permitting rapid diffusion to the stable precipitates to assist in their growth.
Thirdly, severe shear deformation may bring two well-separated precipitates closer
together and promote their coalescence. In an earlier report, a kinetic Monte Carlo
simulation in a 7xxx Al alloy without the presence of mobile dislocations identified
coalescence as an important mechanism for larger GP zone growth [145]. Thus, if a
metastable precipitate is located extremely close to a neighbouring stable precipitate,
the metastable precipitate may merge with the precipitate and this will lead to the
growth of the large platelet at the expense of the small coherent precipitates. As a
result, the growth of large precipitates will become proportional to the decrement of the
number of small coherent precipitates and this is consistent with the significant drop in
the precipitate number density shown in Fig.4.24b.
Processing by the alternative SPD process of high-pressure torsion is known to
increase the concentration of vacancies and consequently this increases the diffusion
coefficient of solute atoms within the matrix [146]. If there is a similar increase in
vacancy concentration in ECAP, this will assist the rapid development of the
precipitation microstructure in the Al-Zn-Mg-Cu alloy.
Fragmentation of the larger platelets plays a major role in precipitate evolution
and this process is responsible for the spheroidization of precipitates during ECAP
165
[21,82]. If fragmentation of the small precipitates also occurs during ECAP, the growth
of the small precipitates will be restrained and essentially no small platelets will have the
capability to grow into larger precipitates. This is in contradiction with the many larger
platelets, more than ~25 nm long, observed after 1 pass and shown in Fig. 4.21a.
Accordingly, it is concluded that the fragmentation of small platelets <25 nm is not
important in the precipitate evolution occurring during ECAP. Careful examination by
APT confirmed that all large platelets have an hexagonal shape which is a characteristic
feature of the perfect plate reported elsewhere [147]. Figure 5.1 shows seven pairs of
close-neighbour precipitates out of 80 precipitates inside an analyzed volume of the
sample taken through 1 pass where these precipitates are probably undergoing
coalescence. Indeed, if the growth of large precipitate is predominantly via coalescence
rather than by the dissolution of smaller precipitates, no small solute cluster debris is
expected to exist at points remote from these larger precipitates. This conclusion is fully
consistent with the present APT and TEM observations.
5.1.3. The effect of ECAP on precipitate orientation and morphology
Processing by ECAP alters the orientation of the precipitates in the matrix so
that it deviates from the typical morphology observed in thermally-aged samples. A
single pass in ECAP produces a limited rotation of the larger precipitates and this rotation
is by about 2° away from the initial orientation based on the curved reflections of in
166
the SADP after 1 pass in Fig. 4.22a. The severe shear deformation arising from multiple
passes of ECAP further moves the precipitates away from their original orientations
according to the SADP in Fig. 4.22b. This change in the precipitate orientation
introduces an important consequence for the morphological evolution of the
precipitates.
After 1 pass of ECAP the platelets have lower aspect ratios by comparison with
the platelets in the thermally-aged samples. In particular, the smaller precipitates have
aspect ratios of <2:1 and this is significantly lower than for the large platelets where the
aspect ratios are from 3:1 to 4:1 as shown in Fig. 4.21a. Such morphology characteristics
are consistent with the orientation changes that occur in 1 pass of ECAP. It is reasonable
to assume that SPD produces the same orientation change in small platelets as in the
larger platelets. Thus, the change of precipitate orientation should increase the interfacial
energy between the precipitates and the matrix due to the change in the atomic
configuration at the low energy interfaces between the basal planes and the matrix.
Since the small precipitates have larger relative surface areas than the larger precipitates,
they are also more sensitive to the change in interface energy. It follows that, by
adopting low aspect ratios, these small platelets are able to effectively reduce their
surface area and accordingly small platelets with low aspect ratios are favoured
thermodynamically.
The equiaxed precipitates present after multiple passes of ECAP are due to their
new orientations within the matrix. Since the new orientations deviate from the typical
167
orientation for platelets, the atomic configurations at the interphase interfaces of / -
Al are significantly different from those of typical platelets. The missing low-energy
configuration interface between the basal plane of and {111} of the -Al matrix
increases the interfacial energy between and the matrix. Thus, the precipitates
become more favourable for isotropic growth and they evolve into an equiaxed
morphology. Although fragmentation of the larger precipitates may directly reduce the
average precipitate size, the increase in their number density and the change in their
aspect ratio are effective in maintaining an overall equiaxed morphology for these
fragmented precipitates. Furthermore, if the fragmented precipitates retained their
original orientations, it follows that they would maintain their anisotropic growth mode
and thus it would be difficult or impossible to evolve into the larger equiaxed precipitates.
It is concluded that the growth of the large equiaxed precipitates observed in this
investigation provides a clear demonstration that the change in the precipitate orientation
induced by SPD is directly responsible for the precipitates in isotropic growth and their
equiaxed morphology.
This experimental investigation has revealed a significant effect in conducting
ECAP at 473 K on the evolution of the precipitate microstructure in Al-7136. Firstly,
ECAP accelerates the precipitation kinetics by a factor of about 50 times faster than in
conventional ageing treatments. Secondly, it affects the precipitation thermodynamics by
promoting isotropic growth and forming equiaxed precipitates in the alloy. To control
the precipitation microstructure effectively for an optimisation of the strengthening
168
Figure 5.1. . Multiple view examinations of 7 pairs of close neighbour precipitates
adjacent to each other inside the 1-pass ECAP Al-7136 sample shown in Fig. 4.23a. (a),
(b), (c) shows 3 cases of close neighbour small precipitates, (d), (e), (f) and (g) shows a
small precipitate touching a larger one.
20 nm
(a)
(b)
(c)
(d)
(e)
(f)
(g)
169
effect arising from the precipitation of fine , further research is now needed to obtain a
more complete understanding of the effect of ECAP at lower temperatures on the
evolution of precipitates.
5.2 Segregation and depletion of alloying elements at grain boundaries
during SPD
The effect of ECAP on the evolution of the precipitates has been extensively
discussed in the previous paragraph, it is obvious there are fragmentations of the large
existing precipitates as well as agglomeration and growth of the small precipitates
resulted from the severe plastic deformation. The intensive imposed strain meanwhile
introduced a great number of dislocations into the materials that served as fast diffusion
paths for the small precipitates that promoted the precipitates growth. The agglomeration
of the small precipitates on one side decreased the number of precipitates that served as
barriers for dislocation motions , on the other side, increased the precipitates’ size , which
would inversely affect the role of these precipitates as barriers as well. It is thus
concluded that the promotion of precipitate growth is one of the important reasons that
lead to the softening of the materials subjected to ECAP at 473 K.
As discussed earlier, solid solution is another factor that possibly leads to the
softening of the material. It is usually very difficult to study the solid solution of the
material , traditionally, x-ray diffraction profile analysis is an approach to study the
lattice distortion that reflects the solid solution, there is no direct approach that could
170
visualize the solid solution within the matrix , not to mention the chemical composition
evolution of the grain boundaries.
The grain boundaries (GB) in UFG materials is an important feature of the
microstructure and they are more predominant than in conventional coarse-grained
materials. Indeed, these boundaries are responsible for the unique properties
demonstrated by UFG materials. Over the last 10 years there has been intensive
researches to understand the GB structural details and the texture effects of UFG
materials [62,148,149]. By contrast, little research has been conducted to tackle the
precise chemical information of the GBs and the role of any solute segregation. A better
understanding of solute segregation at the GBs will assist in understanding the
mechanical properties associated with the boundary regions and this is important both in
delineating the stability of the UFG structures and in determining the associated
deformation mechanisms. Furthermore, information on the precise solute segregation at
GBs is essential in obtaining an understanding of the solute balance in the material and
the significance of any unique precipitation behaviour in these UFG and nano-grain
materials. Attaining detailed information on grain boundary chemistry has long been a
very challenging task for material scientists [150,151]. However, the use of a three-
dimensional atom probe provides a unique tool which is sufficiently powerful to yield
quantitative measurements on the chemical details of the grain boundaries [121,152,153].
Furthermore, the new generation atom probe offers a larger view area of ~80 80 nm
2
which significantly improves the analytical statistics of the quantitative measurements
171
[154]. In the present investigation, modern atom probe tomography (APT) was employed
to characterize the grain boundaries in Al-7136 samples processed through different
numbers of ECAP passes.
Figure 5.2 shows the reconstructed elemental maps of an analysed volume from
an Al- 7136 alloy sample processed through 4 passes of ECAP. There is clear evidence
for a grain boundary across the analysed volume and a precipitate within the matrix. The
Mg and Cu maps shown in (a) and (b) demonstrate the clear segregation of Mg and Cu
atoms at the GB but the Zn map shown in (c) reveals a depletion of Zn near the GB
region. Trace element maps of (d) Si, (e) Cr and (f) Mn, with their discernible peaks in
the mass spectrum of the analysed volume, reveal no clear segregation of these elements
at the GB.
Using a selection box with its z axis parallel to the GB normal as shown in Fig.
5.2(a), the composition profile across the GB was obtained and is shown in Fig.5.3. The
Mg concentration at the GB was measured as high as 1.2 0.1 at% which is about 6
times higher than the value in the matrix (0.2 0.05 at%) as shown in Fig. 5.3. The peak
Cu concentration at the GB is 0.8 0.1 at% which is about 8 times higher than the value
measured in the matrix (0.1 0.05 at%). The Zn depletes at the GB to a level of 0.3
0.08 at% which corresponds to about one-third of the Zn concentration in the matrix
(1.0 0.1 at%). Measurements show the Zn depletion region is about 10 2 nm wide
across the GB and this is much wider than the Mg or Cu enrichment regions with widths
of about 4 1 and 5 1 nm, respectively. Solute excess is considered as a better
172
Figure 5.2. Elemental maps of Al-7136 alloy sample after 4 passes of ECAP processing:
(a) Mg containing a selection box, (b) Cu, (c) Zn, (d) Si, (e) Cr and (f) Mn maps.
20nm
(a)
(b) (c)
(d) (e) (f)
173
parameter to evaluate the solute segregation at the GB [150,151]. The Mg and Cu
excesses at the GB were measure as 1.0 0.06 atom/nm
2
and 1.06 0.06 atom
/nm
2
,respectively, suggesting that they both have an equality strong tendency to segregate
at the GB.
Figure 5.4 shows the Mg, Cu and Zn maps of an analyzed volume from Al-716
sample after 8 passes of ECAP processing. The analyzed volume contains four visible
grains and they are separated by six GBs. Similar to the GB observed after 4 passes of
ECAP, Cu and Mg segregate at the boundaries. The three grain boundaries marked GB1,
GB2 and GB3 in Fig. 5.4(a) were selected to give careful quantitative measurements.
Their composition profiles across each GB were obtained using a bin box containing
9000 atoms and the results are shown in Fig.5.5. The peak concentrations of Mg at the
three GBs look essentially the same with a value of about 1.8 0.12 at% which is about
10 times hgher than the Mg concentration in the matrix (0.16 0.1 at%). By contrast, the
peak concentrations of Cu measured at the three GBs have different values. Thus, the
peak concentration of Cu at GB1 is 1.2 0.11 at% Cu or about 20 times the
concentration in the matrix (0.06 0.01 at%) whereas the peak concentrations of Cu are
0.8 0.11 at% at GB2, 13 times the concentration in the matrix, and 1.0 at% at GB3, 17
times the matrix concentration.
The composition profiles show that Zn has a similar level of depletion at the three
GB regions, with a minimum value of ~ 0.45 0.08 at% which is close to one-half of the
concentration in the matrix. (~0.79 0.01 at%). It is worth noting that the peak
174
Figure 5.3. Composition profiles from a selection box with the z axis parallel to the grain
boundary normal from Al-7136 alloy sample after 4 passes of ECAP as shown in
Fig. 5. 2(a).
(a) (b) (c)
Figure 5.4 Elemental maps of a polycrystalline region in Al-7136 sample after
8 passes of ECAP. (a) Mg, (b) Cu and (c) Zn maps.
GB2
GB3
GB1
Mg
Cu Zn
20 nm
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
20 30 40 50 60
Distance, nm
Concentration, at%
Mg
Cu
Zn
175
concentration positions of Mg and Cu are very close to each other in the profiles from
two of the GBs (GB1 and GB3) with a small angle of ~25° to the shank axis of the tip
sample. A small narrow Zn enrichment peak was observed in the central region across the
GBs. For horizontal GBs, such as GB2 in Fig.5.4 and the GB in Fig.5.2, Mg peaks were
observed with an offset of ~2 nm in front of the Cu peaks. This is consistent with the
lower evaporation field of Mg by comparison with the Al matrix and with the higher
evaporation field of Cu [121]. This field difference may cause Cu to systematically
evaporate later than Mg.
To avoid the influence of such field evaporation artefacts, the grain boundary
excess is a better value for evaluating solute segregation at the GBs. The Mg and Cu
excesses at GB1 are 1.2 0.03 and 1.29 0.03 atom/nm
2
, respectively, and the Mg and
Cu excesses at GB2 are 0.89 0.06 and 1.0 0.06 atom/nm
2
, respectively. For GB3 there
is an Mg excess of 1.56 0.07 atom/ nm
2
and a Cu excess of 1.4 0.07 atom/nm
2
. It
appears, therefore, that all three grain boundaries have slightly different solute excess
values.
The results obtained in this research provide important information on the solute
segregation behaviour at GBs in the Al-Zn-Mg-Cu alloy after processing by ECAP. Mg
and Cu show higher segregation levels at the GBs and Zn is observed with a clear
depletion zone on either side of the GBs. Careful measurements of the chemistry of a GB
containing a large precipitate (GB3) indicates that a weak and narrow Zn peak, as is
visible in Fig. 5.5(c), will become more evident by moving the measurement region
176
closer to the large precipitate (i.e. ~1-3 nm from the outer edge of the large boundary
precipitate). The increase in Zn concentration with the decrease of the distance from the
outer edge of the precipitate suggests that the GB probably serves as a faster diffusion
path for Zn atoms to the boundary precipitates. Certainly the growth of the larger
boundary precipitates, with a Zn concentration of ~49.5 0.4 at% as reported earlier
[155], will require the diffusion of large numbers of Zn atoms to the precipitate. The
diffusion data for Mg, Cu and Zn in polycrystalline Al alloys is readily available
[156,157] and indicates that Zn diffusion in Al at 473 K is about 3 times faster than Mg
and about 20 time faster than Cu. The grain boundary diffusion of Zn [158] can be at
least two orders of magnitude faster than the grain boundary diffusion of Mg [65] in Al
alloys. The faster diffusion of Zn along the GB and from the matrix is consistent with the
observation of wider Zn depletion regions on either side of the GBs.
The excesses of Mg and Cu measured in the three GBs of the same specimen have
different values. If it is assumed that the fraction of high-angle grain boundaries increases
with increasing numbers of passes in ECAP, as noted earlier in both pure Al and an Al
alloy [62,148,149], the different segregation levels observed at different GBs may be
attributed to the different structural characteristics of the boundaries. Therefore, it is
suggesting that the increase in the number of passes or the decrease of the grain size may
have a positive influence on the solute segregation at the grain boundaries of the Al alloy.
The strong solute-solute interaction may be responsible for the segregation of multiple
solute elements to a single GB. The depletion of Zn indicates that the Zn-Mg interaction
177
and Zn-Cu interaction are not sufficiently strong to promote the segregation of Zn to the
GB but if the Mg-Cu interaction is strong it may enhance the segregation of these two
elements at the GBs.
Not all solutes show the same segregation tendency to the GBs in the Al-7136
alloy and clearly the GB chemistry, such as the Zn/Mg ratio, is significantly different
from the matrix. For coarse grain materials with a low GB area to volume ratio, the GB
segregation will not have a significant effect on the matrix chemistry. However, when the
grain size is reduced to the nano-scale, the preferential segregation of certain elements at
such a high level of 2 atom/nm
2
will influence the solute segregation at GBs and this
must be taken into account to understand and control the extent of precipitation in these
materials.
5.3. Discussion of the effect of ECAP on Al-7136
Based on the discussions above, it could be concluded that the softening of Al-
7136 came mostly from 2 parts: the precipitates growth and the grain boundary
segregations of the alloying elements. The kinetics of precipitates growth under ECAP
at 473 K is 50 times faster than that with only annealing at 473 K, dislocations and grain
boundaries served as fast diffusion paths that accelerate the growth. On the other side, 3-
D atom tomography also revealed that the excessive alloying elements of Mg , Cu tend
to segregate at the grain boundaries , also, there is a depletion zone adjacent to the grain
boundaries suggesting a fast diffusion of Zn at the grain boundaries, the segregation of
178
alloying elements would lower the solid solutes in the matrix , which consequently soften
the materials. Based on these result, the increased diffusion coefficients from the high
temperature combined with the promotion of diffusion of the ECAP would substantially
increase the growth of the precipitations , which would also made the coarsening of the
grains much easier. Consequently, the high temperature performance of Al-7136
subjected to ECAP is a lot worse than that of the as-received one. It is also worth
mentioning room temperature ECAP pressing following the solid solution could
strengthening the materials, this is due to the fact that a slower kinetics at the room
temperature, despite the driving force from the supersaturation of the alloying elements is
even stronger from the solid solution , the precipitates are a lot smaller in the 1-pass room
temperature ECAP pressing sample and also less segregation at the grain boundaries,
assumingly. These features would help retaining the strength of the material, the
strengthening of the yield strength was from the grain refinement according to the TEM
study. However, if the diffusion became easier, for example, tested at high temperature,
the diffusion coefficient is a lot larger, the dislocations and defects introduced by ECAP
would also promote the growth of the precipitates, that is why the room temperature
processed Al-7136 showed a poor high temperature performance.
5.4. Discussion of the effect of HPT on Al-7136
As shown in chapter 4, high pressure torsion could be successfully conducted on
the Al-7136 discs at room temperature without cracking. The Vickers hardness test
179
Figure 5.5. Composition profiles obtained using a selection box of 20 20 40 nm with
the z axis parallel to the normal of each grain boundary in the sample after 8 passes of
ECAP as shown in Fig.5. 4: the composition profiles from (a) GB1, (b) GB2 and (c) GB3.
Top Bottom
(a)
(b)
(c)
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
0 10203040
Distance, nm
Concentration, at%
Mg
Cu
Zn
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
0 102030 40
Distance, nm
Concentration, at%
Mg
Cu
Zn
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
0 10203040
Distance, nm
Concentration, at%
Mg
Cu
Zn
GB1
GB2
GB3
180
revealed that HPT could effectively strengthen the materials after even 1/8 turn of HPT at
the pressure of 6 GPa. A pre-processing of ECAP at room temperature of 1 pass indicated
a further strengthening effect and lead to an even smaller grain size of the materials.
However, the hardness evolution is rather complicated and needs to be discussed.
The Severe Plastic Deformation of supersaturated solid solution could be
considered as a balance between deformation-induced disordering and deformation-
accelerated diffusion bringing the alloys closer to equilibrium. It is a well-known fact that
an increase of the concentration of solute component leads to solid solution hardening of
a material. In contrast, a decrease of grain size leads to Hall-Petch hardening. Based on
the results obtained and other literature reports [159,160], it can be expected that both
processes compete during HPT.
As shown in the previous section, 3-D atom tomography result confirmed that
there is a segregation of the alloying elements at the grain boundaries, the accelerated
precipitate growth suggested that dislocations also served as fast diffusion paths, If we
understand that HPT is a even more intense deformation method compared to ECAP, and
usually leads to higher dislocation densities and smaller minimum grain size achievable,
it is reasonable to believe that all these characteristics of segregation and diffusion still
exist in the samples after High Pressure Torsion. It has to be noted that the 3-D atom
probe technology could not be applied to the HPT samples as it has a minimum size limit
for the specimen, the HPT sample is too thin to be made into the 3-D APT specimens.
Nevertheless, the intrinsic feature that SPD could promote the precipitate growth
and segregation has been well-studied in the ECAP part, there are also other reports
181
suggesting the softening of Aluminum alloys after HPT processing at room temperature.
Al-Zn, Al-Mg and Al-Zn-Mg alloys were selected as the experimental materials in those
reports[159,160].
As observed in the Al-7136 alloys, there were significant grain refinements
reported in these alloys, meanwhile, the supersaturated solid solution decomposition also
occurred, Fig.5.6 depicted the decomposition in Al-Zn alloy[159]. As the alloys actually
evolved from a more equilibrium grain structure to a less equilibrium one, meanwhile,
the phase structure after the decomposition is more equilibrium state, HPT at room
temperature should rather be considered as a warm deformation processing at moderate
strains than a cold work.
It has also to be pointed out that the extent of the decomposition of Zn and Mg is
different even in the same HPT conditions in these reports. As shown in Fig 5.6 , almost
all of the excessive Al-Zn solid solution decomposed in the Al-Zn alloys, this is different
in Al-Zn-Mg alloys and Al-Mg alloys, where only part of the excessive solid s
decomposed. The bulk diffusion coefficients of Zn and Mg in all these alloys were
calculated based on the grain size and concentration information, , in Al-Zn and Al-Mg
alloys, the calculated bulk diffusion coefficients are 10
-15
m
2
s
-1
for Zn and 10
-17
m
2
s
-1
for Mg, in Al-Zn-Mg alloys, the calculated bulk diffusion coefficients is 10
-17
m
2
s
-1
for
Zn and Mg. Earlier reports on the bulk diffusion coefficients of Zn in Al are available
both on polycrystals [161] and single crystals[162].Despite neither of them gave out the
room temperature diffusion coefficients, the predictions using the Arrhenius equation
182
Figure.5.6.Concentration of Zn in (Al) solid solutions in various Al-Zn alloys
before and after HPT [159]
183
based on the activation energy and temperature information for both the poycrystal and
single crystal are similar at the value of around 1.0x10
-23
m
2
s
-1
. One may notice the
significant difference between the values from experiments on alloys subjected to HPT
and this prediction. The situation is similar for the Mg bulk diffusion in Al[163], the
extrapolated diffusion coefficient for 300 K is around 1.7x 10
-24
m
2
s
-1
, which is also a lot
less than that calculated based on the diffusion paths in the alloys subjected to HPT. It
also could be concluded that generally Mg diffuses at a slower speed compared to Zn in
Al matrix, this could probably explain the distinct behaviors of Mg and Zn solid solution
decomposition in these alloys. However, the difference of about 8 orders of magnitude
for both Zn and Mg diffusion coefficients is more important, the bulk diffusion should be
considered as frozen and not to count for the decomposition of the solid solutions during
HPT.
GB diffusion is another possible diffusion mechanism for Zn and Mg diffusion in
Al matrix in this case. It is actually quite reasonable if one understands that the Zn and
Mg solute solutions in Al matrix should have very strong interactions with the
dislocation cores. These interactions would play very important roles in forming the new
GBs, consequently, the newly formed GBs would be enriched in the Zn and Mg elements.
Then, it is very likely GB diffusion would be favored over the bulk diffusion for these
alloys subjected to HPT. The calculation based on the GB diffusion mechanism were
also carried out on for Al-Zn, Al-Mg and Al-Mg-Zn alloys subjected to HPT [159,160].
The calculated D value of ~(0.5-1) x 10
-24
m
3
s
-1
is very close to the available data for
184
65
Zn tracer GB diffusion at 300 K (~(1-3) x 10
-24
m
3
s
-1
) [164-167]. From Other reports
[168,169] , the extrapolation of the data for the sD for Mg GB diffusion in Al to 300 K
is around 5 x 10
-28
m
3
s
-1
. Similar to bulk diffusion, Mg GB diffusion in Al matrix is
slower than Zn GB diffusion in Al matrix. One may link this fact to the slower
decomposition of supersaturated solid solutions in Al-Mg alloys compared to that in Al-
Zn alloys deformed under the same HPT conditions[159,160].
Even though it seems the previous assumptions that GB diffusion dominates
could explain the decomposition of Zn-rich and Mg-rich solid solutions in these alloys
during HPT, there is one very important factor that has been ignored – high pressure.
Generally high pressure has a negative influence on the diffusion. Extrapolation of data
for Mg bulk diffusion in Al–Mg alloys [170] to 5 GPa yields D (300 K, 5 GPa) = 10
-29
–
10
-28
m
2
s
-1
. This is 4–5 orders of magnitude less than that at atmospheric pressure. The
data for Zn GB diffusion under high pressures in Al polycrystals [171] demonstrate that
by extrapolation to 6 GPa the GB diffusivity decreases by 3.5–4.7 orders of magnitude
depending on the GB misorientation angle. As of these, there must be some other changes
brought by the HPT processing that cause a much faster diffusion. Defects or vacancies
introduced by the heavy straining from HPT probably would assist the diffusion. There
are reports using residual resistivity measurements to estimate the vacancy production
during cold work [172]. In metals the residual resistivity attributed to point defects
increases with strain according to laws that vary somewhat with the type of stress–strain
curve. From the residual resistivity per Frenkel pair, deduced from irradiation
185
experiments, it was concluded that atomic concentrations of 10
-5
–10
-4
are reached for
strains =1 [173,174]. This is indeed significant and comparable to the vacancy
concentration comparable to the equilibrium value [172]. The deformation dependence
of the atomic concentration c of point defects may be written as
where is the shear modulus and the stress value. This relationship is obeyed fairly
well in fcc metals [164,175,176].
There have been reports claiming the production of additional vacancies during
SPD [146]. In Fe-Cu system. Kiritani et al. directly observed the formation of vacancy
clusters in the form of stacking fault tetrahedra by the in situ elongation of thin Au, Cu,
Ni and Al foils in the column of a TEM instrument [177]. Ex situ measurements of
positron lifetime after HPT deformation of Cu and Mg alloys reveal the presence of small
microvoids with a size of 4–5 vacancies [178-182]. Ungar et al. [183] and Zehetbauer et
al. [184] recently proposed the a method using the diffuse background scattering of X-
rays that would be employed in in-situ measurement of the concentration of excess
vacancies during SPD. In Fig.5.7, the kinetics of softening exactly follows the
decompositions of solid solution measured by the changing of lattice parameters in Al.
One can conclude that GB diffusion accelerated by fluxes of vacancies produced
due to SPD can explain the full decomposition of supersaturated solid solution in the
Al–Zn alloys and its partial decomposition in the Al–Mg and Al-Mg-Zn alloys. The solid
186
solution hardening would be consequently weakened, this is more pronounced in the Al-
Zn alloys but also applied to Al-Zn-Mg alloys as well as Al-Mg alloys. This softening
could be stronger than work hardening, Hall-Petch hardening and precipitation hardening,
which means SPD or more specifically HPT may lead to the softening of the super-
saturated materials. This is exactly what happened to Al-7136 during HPT.
Based on all these discussions above , it is possible to explain the complicated
hardness evolution assuming the facts that severe plastic deformation methods including
high pressure torsion could introduce dislocations and defects that assist the grain
boundary diffusion, also segregation of the alloying elements at the grain boundaries
exists in the HPT samples and the extent of segregation is higher with more turns of HPT
and smaller grain sizes.
For the HPT only samples, the overall hardness increased at lower turns of
HPT , this is because substantial grain refinement occurred , meanwhile, because of the
existence of the large precipitates in the as-received materials, precipitate fragmentation
happened as shown in the TEM images, both of them would lead to the strengthening of
the materials, on the other side, the imposed strain is not large enough to cause
substantial segregations at the grain boundaries. Overall, the Al-7136 disc was
strengthened. With the strain increasing, more dislocations and defects would be
introduced into the materials, the segregations would also be more pronounced, the
fragments of the large precipitates, usually located at the grain boundaries because of the
pinning effect at lower turns would in turn grow into bigger size, furthermore , the
187
Fig. 5.7. Dependence of: (a) hardness and (b) lattice spacing on deformation
degree (number of torsions) for Al–30 wt.% Zn alloy.[185]
188
further grain size refinement is only moderate from 1 turn to 2 turn samples,
consequently, the decrease of the hardness was found. When the size of the precipitates
surpassed a critical value, the fragmentation would again occurred : in the 4 turn samples ,
there are very large precipitates distributed on the grain boundaries probably due to an
inhomogeneous straining suggesting the continuous growth of the precipitates without
fragmentation, for most of the precipitates, they are in the size of around 10 nm and
distributed on the grain boundaries, no obvious alignment of the precipitates was found,
this is because fragmentation now occurred on much smaller precipitates, the fragments
would be much less thus more difficult to indentify as an alignment. The large grains
found in the peripheral part of the 4- turn samples is from the dynamic recrystallization,
which happened when the imposed strain is sufficiently high.
For the room temperature ECAP plus HPT samples, solid solution treatment
would increase the alloying element in the matrix ,which would lead to a lower stacking
fault energy. As previously reported, materials with lower stacking fault energy evolves
at a slower speed, this is why for the 1 turn sample, in the center part, most of the grains
are less developed and larger than that of the HPT only sample. The strengthening at the
1 turn center part is mostly from the room temperature ECAP, for which an average
Vickers hardness value measured from 16 random points on the disc is around 210 Hv,
this is close to that of the 1 turn center part. As the strain increases , the center part was
much refined , the Hall-Petch factor played a dominant role in the strength of the material,
this is why the center part of the materials is strengthened rather than softened, for the
189
peripheral part, despite the grains were also refined, according to earlier investigations,
the segregation at the GB as well as the promotion of the diffusion is much more
pronounced , this is also proved by the TEM investigation, the hardness dropped by a
large margin- this should be correlated to the much higher alloying elements
compositions in the materials after solid solution, the driving forces are consequently
higher , if we assume the equilibrium status in the matrix is identical, there would be
more precipitates or segregations at the GBs, which lead to a drastic decrease of the
hardness.
This could also explain the HPT samples with 4 pass ECAP pre-processing at 473
K, according to the atom probe study , the solid solute in the 4 Pass ECAP at 473 K
sample is much lower than that of the as-received alloys, close to the equilibrium state.
As a result, although similar defects or dislocations would be introduced into the
materials by HPT, the segregation or promoted diffusion would be much less, simply
because the driving force is much lower and the excessive alloying element is much less
within the matrix , the softening would be in a less extent compared to other HPT
samples. However, grain refinement would still occurred, that is why the materials was
continuously strengthened from 1 turn to 2 turn sample. It has to be noted that this part of
conclusion needs to be more carefully studied using TEM or other microstructure study
methods.
190
6. SUMMARY AND CONCLUSIONS
In the current research , the supersaturated Al-Zn-Mg-Cu alloy Al-7136 is
extensively studied to investigate the effect of severe plastic deformation on the
mechanical properties and microstructure of this specific Aluminum alloy. The unique
mechanical behavior of the materials subjected to severe plastic deformations including
ECAP and HPT were depicted using room temperature tensile tests, high temperature
tensile tests as well as Vickers microhardness study. In order to understand the
mechanisms behind the mechanical behaviors, microstructure studies using TEM, 3-D
Atom probe and XRD have been conducted on materials in different conditions. The
results obtained by the testing and conclusions acquired from the discussions are
described as follows.
1) The as-received Al-7136 in natural ageing condition was subjected to ECAP using
route B
C
at the elevated temperature of 473 K for up to 8 passes. The measured
yield stresses of the Al-7136 alloy are significantly lower after ECAP by
comparison with the as-received material and, in addition, there is relatively little
change in the yield stresses between 1 and 8 passes. By contrast, there is a
tendency for the flow stress to decrease with increasing numbers of ECAP passes.
There is no simple relationship between the number of passes and the extent of the
uniform elongation.
2) Room temperature of 1 pass could be successfully conducted on Al-7136
immediately after a specific solid solution treatment at 743 K for 1 hour followed
by water quenching , the required solid solution duration is longer than that reported
191
in similar Al-Zn-Mg alloys, this is correlated to the higher alloying elements
containment in Al-7136. Room temperature tensile test on the room temperature 1
Pass sample revealed an increase in the yield strength of the billet pressed at room
temperature by >100 MPa thereby confirming that lower temperature processing is
advantageous in retaining high strength in this alloy.
3) It is also feasible to process Al-7136 at 383 K for 1 pass following the identical
solid solution treatment as that for the room temperature ECAP, room temperature
tensile test indicated that despite yield strength could be retained, the UTS of the
as-pressed materials is worse than that of the as-received material.
4) High temperature tensile tests were conducted on the Al-7136 samples subjected to
ECAP at 473 K. First, tensile tests at 673 K were done on 4-P Al-7136 samples at
different strain rates, the results suggested an optimum strain rate of 10
-2
s
-1
, with
which a maximum elongation-to-failure of around 300% was achieved. At 703 K,
10
-2
s
-1
still is the optimum strain rate, while the maximum elongation-to-failure
achieved on Al-7136 sample is slightly smaller, of about 280%;Secondly, materials
subjected to different passes of ECAP at 473 K were tested at 743 K and the strain
rate of 10
-2
s
-1,
no elongation-to-failure larger than 280% was achieved, there is no
correlation between the number of passes and the elongation-to-failure found.
5) As-received Al-7136 were also tested at 703 K and different strain rates. the
maximum elongation-to-failure achieved is at the strain rate of 1.0 x 10
-3
s
-1
with
the value of 480%, larger than the maximum one achieved on the as-pressed
samples.
192
6) The 383 K 1-Pass Al-7136 was tensile tested at 643 K and 673 K using strain rates
of 10
-2
s
-1
and 10
-3
s
-1
, the maximum elongation-to-failure is around 310 %
achieved at 643 K and the stain rate of 10
-3
s
-1
. Similarly , the room temperature 1-
pass Al-7136 was tested at 673 K and the strain rate of 10
-3
s
-1
, the elongation-to-
failure is around 260%. Both of these two values are substantially lower than that
achieved on the as-received material.
7) Microstructure study on the ECAP samples using TEM, XRD and 3-D atom probe
technology revealed that ECAP could effectively refine the grain size, an equiaxial
structure could be achieved after 4 passes of ECAP at 473 K or above, the
minimum grain size achieved after 8 passes of ECAP at the 473 K ECAP is 300-
400 nm compared to the 3.4 m grain size in the as-received materials. The grain
size in the room temperature 1-pass sample is substantially smaller than that of the
sample subjected to 1 pass ECAP at 473 K, which is consistent to the room
temperature tensile test result.
8) The morphology and distribution of the precipitates in Al-7136 subjected to ECAP
at 473 K have been studied in this research using 3-D atom tomography, it is
suggesting processing by ECAP of Al-7136 alloy at 473 K changes the precipitate
orientations in the matrix. The consequent change in interfacial energy at the / -
Al interfaces promotes precipitation and leads to equiaxed particles in multiple
pass samples. The precipitate evolution occurring during ECAP at 473 K is about
50 times faster than in conventional ageing treatments at the same temperature. The
high density of mobile dislocations produced by ECAP promotes the dissolution of
193
small metastable precipitates and the formation of large precipitates by
coalescence. The major solute concentrations within the matrix decrease with
increasing numbers of ECAP passes thereby suggesting that precipitate evolution is
in the growth regime. Fragmentations of the precipitates was observed but is only
important to the large precipitates with the size > 25 nm.
9) Three-dimensional atom probe tomography was also used to characterize GB
segregation in Al-7136 processed under multiple passes by ECAP. The results show
Mg and Cu segregate strongly at GBs with an excess in the range of 0.89-1.5 atom/
nm
2
for Mg and 1.0 - 1.4 atom/nm
2
for Cu. The main alloying element of Zn
exhibits no segregation at the GB and instead it gives a depletion zone with a width
of ~10 nm along the boundaries. The depletion zone suggested GBs are fast
diffusion paths of Zn to form precipitates on the grain boundaries.
10) High pressure torsion was conducted on as-received Al-7136 as well as Al-7136
subjected to room temperature 1 Pass ECAP preprocessing and 473 K 4 Pass ECAP
preprocessing for different turns at room temperature . No cracking was found.
11) Patterned Vickers hardness tests were carried on each of the discs subjected to HPT
processing, the results were plotted into color-coded contour maps shown in the
result part. The hardness evolution indicated by the maps showed different trends
for materials subjected to different processing routines. For HPT only samples, the
discs were overall strengthened in samples subjected to low turns of HPT, starting
from 1 turn samples, the materials showed a trend of softening .For room
temperature ECAP plus HPT samples, the materials were again overall strengthened
194
in the 1 turn sample, but the hardness evolution varied from 1 turn sample to 2 turn
sample dependent on the distance to the center of the disc: for the inner region
within 2/3 of the radius, the hardness increased drastically, by an average of
around 40 Hv; However, for the remaining part of the disk , i.e. the outer ring
shape region, the hardness decreased by the margin of 40 Hv; for the 473 K 4 Pass
ECAP plus HPT samples, the discs showed a hardness saturation in the 1 turn disk
at the value of ~220-230, from 1 turn to 2 turn samples, there were no softening
found.
12) TEM study on the microstructure of the center and peripheral part of most HPT
discs showed HPT has a significant grain refinement effect on Al-7136, the
minimum grain size achievable with HPT only processing is about 100 nm. For
room temperature ECAP plus HPT samples, extremely fine grain size of less than
40 nm was achieved in the peripheral.It has to be mentioned that sample with room
temperature ECAP preprocessing showed a slower grain size evolution , this is due
to the lower stacking fault energy from the solid solution treatment. As to the
precipitate part, there are precipitate fragmentations in the 1 turn HPT only sample,
from 1 turn to 2 turn sample, the precipitate grew into bigger size, from 2 turn to 4
turn part, as the precipitate continuously grew, fragmentations of the precipitates
again occurred, most of the precipitates were in the similar size as that in the 1 turn
sample. All of the precipitates observed in the HPT samples were distributed on the
grain boundaries, suggesting GBs played a very important role in the precipitations
during HPT processing.
195
13) All of the hardness evolution could be explained based on the promotion of the
precipitates, segregations of the alloying elements at the GBs and that extent of
segregation is higher with more turns of HPT and smaller grain sizes. The
discussions on the diffusion coefficients during HPT processing suggested that
dislocations and defects introduced by the HPT assisted the diffusion and
consequently lead to the unique mechanical behavior of the supersaturated Al-7136.
196
REFERENCES
[1] K.Kubota , M.Mabuchi and K.Higashi, J Mater Sci 34 (1999) 2255.
[2] A.Bussiba ,A.BenArtzy , A.Shtechman, S.Ifergan and M.Kupiec, Mater Sci Eng
A302 (2001) 56
[3] W-J.Kim, S.W.Chung, C.S.Chung and D.Kum, Acta Mater 49 (2001) 3337.
[4] J.C.Tan, M.J.Tan,Mater Sci Eng A339(2003),81.
[5] A.Galiyev,R. Kaibyshev, Scripta Mater 51(2004) 89.
[6] R.Z.Valiev ,T.G.Langdon,Prog.Mater. Sci. 51 (2006) 881.
[7] R.Z.Valiev, R.K.Islamgaliev and I.V.Alexandrov, Prog Mater Sci 45 (2000) 103.
[8] A.P. Zhilyaev, T.G. Langdon .Prog Mater Sci 53 (2008) 893.
[9] R.Z.Valiev,D.A.Salimonenko,N.K.Teenev,P.B.Berbon and T.G.Langdon, Scripta
Mater 37 ( 1997) 1945.
[10] Z.Horita, M.Furukawa, M.Nemoto,A.J.Barnes and T.G Langdon, Acta Mater 48
(2000) 3633.
[11] V.N.Perevezentsev,V.N.Chuvil’deev, A.N.Sysoev, V.I.Kopylov and T.G. Langdon,
Phys Metals Metall 94 (2002) S45.
[12] R.Kaibyshev, K.Shipilova, F.Musin and Y.Motohashi, Mater Sci Tech 21 (2005)
408.
[13] C.Xu , T.G.Langdon, Scripta Mater 48 (2003) 1.
[14] C.J.Luis Perez, Scripta Mater 50 (2004) 387.
[15] R. Luri , C.J.Luis, J.Leon and M.A.Sebastain, J. Manufacturing Sci. Eng.128 (2006)
860.
[16] A.V.Nagasekhar, W.Wei, T.H.Yip and G.Chen, Adv.Eng.Mater 9 (2007) 573.
[17] T.G.Langdon,Mater Sci Forum 15 (1999) 304.
197
[18] S.Komura, P.B.Berbon, M.Furukawa, Z.Horita, M.Nemoto and T.G. Langdon,
Scripta Mater. 38 (1998) 1851.
[19] Y.Iwahashi, Z.Horita, M.Nemoto and T.G. Langdon, Metall Mater Trans A29 (1998)
2503.
[20] L.J.Zheng, H.X.Li, M.F.Hashmi, C.Q.Chen , Y.Zhang and M.G.Zeng. J. Mater
Processing Tech. 171 (2006) 100.
[21] C.Xu, M.Furukawa, Z.Horita and T.G. Langdon, Acta Mater 53 (2005) 749.
[22] C.Xu, W.Dixon, M.Furukawa, Z.Horita and T.G. Langdon,Mater Letters 57 (2003)
3588.
[23] G.Sakai, Z.Horita and T.G.Langdon, Mater Sci & Eng A393 (2005) 344.
[24] A.P.Zhilyaev,K.Oh-ishi,T.G.Langdon and T.R.McNelley, Mater Sci. Eng, A410-411
(2005) 277.
[25] Z.Horita, T.G.Langdon, Mater Sci.Eng. A410-411 (2005) 422.
[26] C.Xu,Z.Horita and T.G.Langdon,Acta Mater 56 (2008) 5168.
[27] M.Y.Murashkin,A.R.Kil’Mametov and R.Z.Valiev,Phy. Metal. Metallography 106
(2008) 90.
[28] A.P.Zhilyaev, G.V.Nurislamova, B.K.Kim, M.D.Baró, J.A.Szpunar and T.G.
Langdon, Acta Mater 51(2003) 753.
[29] A.P.Zhilyaev, T.R.McNelley and T.G.Langdon, J Mater Sci 42 (2007) 1517.
[30] H.S.Kim,S.I.Hong,Y.S.Lee,A.A.Dubravina and I.V.Alexandrov, J. Mater Process
Tech. 142 (2003) 334.
[31] K.Edalati,T.Fujioka and T.G.Langdon,Mater. Trans.50 (2009) 44
[32] R.Z.Valiev,Y. Estrin,Z. Horita ,T.G. Langdon ,M.J. Zehetbauer and Y.T. Zhu,
JOM 58(4) (2006);33.
[33] Y.Saito,N. Tsuji, H.Utsunomiya, T.Sakai and R.G.Hong. Scripta Mater 39 (1998)
1221.
[34] R.Z.Valiev, A.V.Korznikov and R.R.Mulyukov, Mater Sci Eng A168 (1993) 141.
198
[35] R.Z.Valiev, Mater Sci 21 (1996) 369.
[36] M.V.Markushev, C.C.Bampton , M.Y.Murashkin and D.A.Hardwick, Mater Sci
Eng.A 234 (1997) 927.
[37] Z.Horita, T.Fujinami, M.Nemoto and T.G.Langdon,Metall Mater Trans A31 (2000)
691.
[38] N.Saito, M.Mabuchi, M.Naksnishi, I.Shigematsu, G.Yamauchi and M.Nakamura, J
Mater Sci 36 (2001) 3229.
[39] S.Lee S, P.B.Berbon, M.Furukawa, Z.Horita, M.Nemoto, N.K.Tsenev,R.Z.Valiev
and T.G.Langdon, Mater. Sci. Eng. A272 (1999) 63.
[40] Y.Iwahashi , J.T.Wang, Z.Horita, M.Nemoto and T.G.Langdon. Scripta Mater
35(1996) 143.
[41] S.Lee,A.Utsunomiya,H.Akamatsu, K,Neishi, M.Furukawa, Z.Horita and T.G.
Langdon,Acta Mater.50 (2002) 553.
[42] S.Ota, H.Akamatsu, K.Neishi, M.Furukawa, Z.Horita and T.G.
Langdon,Mater.Trans.43(2002) 2364.
[43] G.Sakai,Z.Horita and T.G.Langdon,Mater.Trans.45(2004) 3079.
[44] Y.Miyahara,Z.horita and T.G.Langdon,Mater.Sci.Eng.A420(2006) 240.
[45] K.Nakashima,Z.Horita,M.Nemoto and T.G. Langdon,Acta Mater. 46 (1998)1589.
[46] I.V.Aleksandrov,G.I.Raab,L.O.Shestakova,A.R.Kil’mametovand R.Z.Valiev,
Phys.Metals. Metall.93(2002) 493.
[47] V.M.Segal, V.I.Reznikov, A.E.Drobyshevskiy and V.I.Kopylov,Russian Metall 1
(1981) 99.
[48] C.Xu , T.G.Langdon, J Mater Sci 42 (2007) 1542
[49] Y.Iwahashi, M.Furukawa, Z.Horita, M.Nemoto and T.G.
Langdon,Metall.Mater.Trans 29A (1998) 2245 .
[50] M.Furukawa,Y.Iwahashi,Z.Horita,M.Nemoto and T.G.Langdon,Mater. Sci.
Eng.A257 (1998) 328.
[51] Y.Iwahashi,Z.Horita,M.Nemoto and T.G.Langdon,Acta Mater.46 (1998) 3317.
199
[52] K.Oh-ishi, Z.Horita, M.Furukawa, M.Nemoto and T.G. Langdon,Metall.Mater.Trans
A29 (1998) 2011.
[53] M.Mabuchi and K.Higashi,J.Mater.Sci.Lett.17(1998) 215.
[54] S.L.Semiatin,V.M.Segal,R.E.Goforth,N.D.Frey and D.P.Delo,Metall.Mater.Trans
A30 (1999) 1425.
[55] A.Yamashita ,D. Yamaguchi, Z.Horita and T.G.Langdon, Mater Sci Eng A287
(2000) 100.
[56] N.H.Ahmadeev, R.Z.Valiev,V.I. Kopylov, R.R.Mulyukov, Russian Metally 5 (1992)
96.
[57] J.T.Wang,Z.Horita,M.Furukawa, M.Nemoto, N.K.Tsenev, R.Z.Valiev, Y.Ma and
T.G. Langdon, J Mater Res.8 (1993) 2810.
[58] Z.Horita ,D.J.Smith, M.Furukawa, M.Nemoto,R.Z.Valiev and T.G. Langdon, J
Mater Res 11 (1996) 1880.
[59] Z.Horita, D.J.Smith, M.Furukawa, M.Nemoto, R.Z.Valiev and T.G.
Langdon,Mater.Charact.37 (1996) 285.
[60] K.Oh-ishi,Z.Horita,D.J.Smith and T.G.Langdon,J.Mater.Res.16(2001)583.
[61] M.Furukawa,Y.Iwahashi,Z.Horita,M.Nemoto,N.K.Tsenev,R.Z.Valiev and T.G.
Langdon,Acta Mater.45(1997)4751.
[62] S.D.Terhune, D.L.Swisher, K.Oh-ishi, Z.Horita, T.G.Langdon and T.R.McNelley.
Metall Mater Trans A33 (2002) 2173.
[63] T.Watanabe. Res Mech 11(1984) 47.
[64] M.Furukawa, Z.Horita and T.G. Langdon, J Mater Sci 2005;40:909.
[65] T.Fujita,Z.Horita, T.G.Langdon. Mater Sci Eng A371 (2004) 241.
[66] Y.T.Zhu,T.G. Langdon. JOM 56 (2004) 58.
[67] T.Mukai, M.Yamanoi, H.Watanabe, K.Higashi. Scripta Mater. 45 (2001) 89.
[68] T.G.Langdon, Metall Trans A13 (1982) 689.
200
[69] J.Pilling,N.Ridley, Superplasticity in Crystalline Solids, The Institute of Metals,
London 1989.
[70] M.Furukawa,Z. Horita, M.Nemoto, R.Z.Valiev and T.G.Langdon. Acta Mater
44(1996) 4619.
[71] J.E.Hatch, Aluminum: Properties and Physical Metallurgy, Am.Soc.Met., Metal
Parks,Ohio,1985:356.
[72] M.Nemoto ,Z. Horita, M.Furukawa and T.G. Langdon, Mater Sci Forum 304-
306(1999);59.
[73] J.T.Wang,Y.Iwahashi,Z.Horita,M.Furukawa,M.Nemoto,R.Z.Valiev,etal.:Acta.Mater.
,44(1996) 2973.
[74] M.Furukawa,A.Utsunomiya,K.Matsubara,Z.Horita and T.G.Langdon,Acta. Mater,49
(2001) 3829.
[75] D.G.Morris,M.A.Munoz-Morris,Acta.Mater. 50(2002),4047.
[76] V.V.Stolyarov, L.Zeipper, B.Mingler and M.Zehetbauer,Mater. Sci. Eng.
A,476( 2008), 98.
[77] R.B.Figueiredo, T.G.Langdon,Mater. Sci. Eng. A503 (2009),141.
[78] C.Xu,M.Furukawa,Z.Horita and T.G.Langdon, Acta Mater 51 (2003) 6139 .
[79] J.Gubicza,I.Schiller,N.Q.Chinh,J.Illy,Z.Horita and T.G.Langdon.Mater Sci
Eng.A.460-461(2007) 77.
[80] Y.H.Zhao,X.Z.Liao,Z.Jin,R.Z.Valiev and Y.T.Zhu,Acta Mater 52 (2004) 4589
[81] N.Q.Chinh, J.Gubicza, T.Czeppe, J.Lendvai, C.Xu, R.Z.Valiev and T.G.
Langdon,Mater Sci.Eng A 2009,In press.
[82] M.Murayama, Z.Horita, K.Hono, Acta Mater 49(2001) 21.
[83] K.Oh-ishi, Y.Hashi, A.Sadakata, K.Kaneko, Z.Horita and T.G.Langdon,Mater. Sci.
Forum 396-402(2002) 333.
[84] C.Y.Nam, J.H.Han, Y.H.Chung and M.C.Shin, Mater Sci Eng A347(2003) 253.
201
[85] K.Stiller, P.J.Warren, V.Hansen , J.Angenete and J.Gjonnes, Mater Sci Eng A270
(1999) 55.
[86] S.K.Maloney, K.Hono, I.J.Polmear and S.P.Ringer, Scripta Mater 41(1999) 1031.
[87] X.Li , M.J.Starink, Mater Sci Forum 331-337(2000) 1071.
[88] J.C.Huang, I.C.Hsiao, T.D.Wang, B.Y.Lou, Scripta Mater 43(2000) 213.
[89] P.W.Bridgman, J Appl Phys 14 (1943) 273.
[90] R.Z.Valiev, Yu.V.Ivanisenko, E.F.Rauch, B.Baudelet, Acta Mater 44(1996) 4705.
[91] F.Wetscher , A.Vorhauer, R.Stock and A.Pippan,Mater Sci Eng A387-389 (2004)
809.
[92] F.Wetscher ,R. Pippan, S.Sturm, F.Kauffmann,C. Scheu and G.Dehm, Metall Mater
Trans A37(2006) 1963.
[93] N.H.Polakowski, E.J.Ripling,Strength and Structure of Engineering Materials.
Englewood Cliffs, NJ: Prentice-Hall; 1966.
[94] A.Vorhauer, R.Pippan, Scripta Mater 51(2004) 921.
[95] H.Jiang,Y.T. Zhu,D.P. Butt,I.V Alexandrov and T.C.Lowe,Mater Sci Eng A290
(2000) 128.
[96] Z.Yang, U.Welzel, Mater Lett 59 (2005) 3406.
[97] A.P. Zhilyaev, S.Lee, G.V.Nurislamova,R.Z.Valiev and T.G.Langdon, Scripta Mater
44 (2001) 2753.
[98] Y.Estrin, A.Molotnikov, C.Davies, R.Lapovok, J Mech Phys Solid 56(2008) 1186.
[99] L.Tóth, A.Mlinari and Y.Estrin, J Eng Mater Tech 124(2002) 71.
[100] P. McKenzie, R.Lapovok and Y. Estrin Y, Acta Mater 55(2007) 2985.
[101] Y.H. Zhao,X.Z. Liao, Y.T.Zhu,Z.Horita and T.G. Langdon,Mater Sci Eng A410-
411( 2005) 188.
[102] F.A. Mohamed, Acta Mater 51 (2003) 4107.
202
[103] R.Islamgaliev, N.F.Yunusova , I.N.Sabirov , A.V.Sergueeva and
R.Z.Valiev ,Mater Sci Eng A 319-321(2001) 877.
[104] Senkov ON, Froes FH, Stolyarov VV, Valiev RZ, Liu J. Nanostruct Mater
1998;10:691.
[105] V.V.Stolyarov , L.O.Shestakova, Y.T.Zhu and R.Z.Valiev, Nanostruct Mater
12(1999) 923.
[106] Yu.Ivanisenko, I.MacLaren, X.Sauvage, R.Z.Valiev and H.J. Fecht, Acta Mater
54(2006) 1659.
[107] O.V.Rybal’chenko, S.V.Dobatkin,L.M. Kaputkina, G.I.Raab and N.A.Krasilnikov,
Mater Sci Eng A244-248 (2004) 387.
[108] S.D. Prokoshkin, I.Y.Khmelevskaya, S.V.Dobatkin, I.B. Trubitsyna,E.V.
Tatyanin, and V.V.Stolyarov, and E.A.Prokofiev, Acta Mater 53 (2005) 2703.
[109] V.V. Tcherdyntsev,S.D. Kaloshkin, D.V.Gunderov,E.A. Afonina, I.G.Brodova
and V.V.Stolyarov,Y.V. Baldokhin,E.V.Shelekhov and I.A.Tomilin, Mater Sci
Eng A375-377 (2004) 888.
[110] Y.H.Zhao, Y.T.Zhu, X.Z.Liao, Z.Horita and T.G.Langdon, Mater Sci Eng A463
(2007) 22.
[111] Y.H.Zhao, Z.Horita,T.G. Langdon and Y.T.Zhu, Mater Sci Eng A474 (2008) 342.
[112] Y.H.Zhao YH, Zhu YT, Liao XZ, Horita Z, Langdon TG. Appl Phys Lett 89
(2006) 121906.
[113] A.P.Zhilyaev, B.K.Kim, J.A.Szpunar, M.D.Baró and T.G.Langdon, Mater Sci
Eng A391 (2005) 377.
[114] A.P.Zhilyaev, J.Gubicza, G.V.Nurislamova, A.Révész, S.Suriñach ,M.D.Baró and
T.Ungar, Phys Status Solidi A198 ( 2003) 263.
[115] N.Krasilnikov, W.Lojkowski, Z.Pakiela and R.Z.Valiev. Mater Sci Eng A397
(2005) 330.
[116] A.P.Zhilyaev, G.V.Nurislamova, M.D.Baró, R.Z.Valiev and T.G.Langdon, Metall
Mater Trans A33 (2002) 1865.
203
[117] B.Yang, H.Vehoft,A.Hohenwarter,M.Hafok and R.Pippan, Scripta Mater.58(2008)
790.
[118] A.A.Mazilkin,O.A.Kogtenkova,B.B.Straumal,R.Z.Valiev and B.Baretzky, Defect
and Diffusion Forum 237-240(2005) 739.
[119] T.F.Kelly and M.K.Miller, Rev. Sci. Instrum. 78 (2007), 031101.
[120] D.Blavette,A. Bostel, J.M.Sarrau, B.Deconihout and A. Menand, Nature 363
(1993) 432.
[121] M.K.Miller,A. Cerezo,M.G. Hetherlington and G.D.W.Smith, Atom Probe Field
Ion Microscopy,Oxford Science, Oxford, (1996) 377.
[122] G.Sha and S.P. Ringer, Ultramicroscopy 109 (2009) 580.
[123] Y.Ma ,M. Furukawa,Z. Horita, M.Nemoto,R.Z. Valiev and T.G.Langdon, Mater.
Trans. 371 (1996)336.
[124] S.X.McFadden, R.S.Mishra ,R.Z. Valiev,A.P.Zhilyaev and A.K.Mukherjee,
Nature 398 (1999) 684.
[125] T.G.Langdon,M. Furukawa,M. Nemoto and Z.Horita, JOM 52 (2000) 30.
[126] A.Cerezo,P.H.Clifton,A. Gomberg and G.D.W.Smith, Ultramicroscopy 107
(2007) 720.
[127] Y.H.Zhao ,X.Z. Liao,Z. Jin,R.Z. Valiev and Y.T.Zhu, In: Y.T.Zhu,
T.G.Langdon,R.Z. Valiev ,S.L. Semiatin, D.H.Shin, T.C.Lowe , editors.
Ultrafine-grained materials III. The Minerals, Metals and Materials Society (2004)
511.
[128] L.F.Mondolfo,N.A. Gjostein and D.W.Levinson,Trans. Am. Inst. Min. (Metall)
Engrs. 206 (1956)1378.
[129] L.K.Berg ,J. Gjønnes,V. Hansen,X.Z. Li,M. Knutson-Wedel, G.Waterloo,D.
Schryvers and L.R.Wallenberg , Acta. Mater. 49 (2001) 3443.
[130] X.Z.Li ,V. Hansen,J. Gjønnes and L.R.Wallenberg, Acta. Mater. 47 (1999) 2651.
[131] R.M.Allen and J.B.Vander Sande , Metall Trans A 9 (1978) 1251.
[132] G.Sha and A. Cerezo, Acta. Mater. 52 (2004) 4503.
204
[133] J.Gjønnes and C.J. Simesen, Acta. Metall. 18 (1970) 881.
[134] H.P.Degischer ,W. Lacom,A.Zahra and C. Zahra, Z. Mettalkd 71 (1980)231.
[135] H.Perlitz and A.Westgren,Arkiv för Kemi, Mineralogy och Geologi 6B (1943) 1.
[136] D.J.Strawbridge,W.Hume-Rothery and A.T. Little , J. Inst. Metals. 74 (1948) 191.
[137] W.Lefebvre ,F. Danoix,G.Da Costa,F. De Geuser ,H. Hallem,A.Deschamps and
M.Dumont ,Surf. Interface. Anal. 39 (2007) 206.
[138] V.V.Stolyarov,Y.T. Zhu,T.C. Lowe,R.K. Islamgaliev and R.Z.Valiev ,
Nanostruct. Mater. 11 (1999) 947.
[139] S.V.Dobatkin In:Y.T. Zhu,T.G. Langdon,R.S. Mishra,S.L. Semiatin ,M.J.
Saran,T.C. Lowe, editors. Ultrafine grained material II. Warrendale, PA: The
Minerals, Metals and Materials Society (2002) 83.
[140] D.Fatay ,E. Bastarash ,K. Nyilas,S. Dobatkin, J.Gubicza and T.Ungar, Z.
Metallkd. 94 (2003) 842.
[141] M.Murayama,Z.Horita and K.Hono, Acta. Mater. 49(2001) 21.
[142] J.Buha,R.N. Lumley and A.G. Crosky, Mater. Sci. Eng. A 49 (2008) 1.
[143] N.Ryum ,Z. Metallkd. 66 (1975) 338.
[144] P.T.Loo,T.J. Bastow,A.J. Hill,J. da Costa Teixeira and C.R.Hutchinson. In:
J.Hirsch,B.Strotzki , G.Gottstein editors. Aluminium Alloys – Their Physical and
Mechanical Properties, Vol. 1 (2008) 751.
[145] G.Sha and A.Cerezo, Acta. Mater. 53 (2005) 907.
[146] X.Sauvage,F.Wetscher and P.Pareige,Acta. Mater. 53 (2005) 2127.
[147] G.Sha and A.Cerezo, Surf. Interface. Anal. 36 (2004) 564.
[148] O.V.Mishin ,D. Juul Jensen and N. Hansen, Mater. Sci. Eng. A 342 (2003) 320.
[149] A.A.Salem, T.G.Langdon, T.R.McNelley ,S.R.Kalidindi and S.L.Semiatin ,
Metall. Mater. Trans.A 37 (2006) 2879.
[150] B.W.Krakauer and D.N. Seidman,Phys. Rev.B 48 (1993) 6724.
205
[151] B.W.Krakauer and D.N.Seidman,Acta. Mater. 46 (1998) 6145.
[152] A.Cerezo, P.H.Clifton,S.Lozano-Perez,P.Panayi ,G. Sha and G.D.W.Smith ,
Microsc. Microanal. 13 (2007) 408.
[153] D.Blavette,E.Cadel ,C.Pereige,B.Deconihout and P.Caron, Microsc. Microanal.
13 (2007) 464.
[154] T.F.Kelly ,T.T. Gribb , J.D. Martens, R.L. Shepard, J.D. Wiener, S.A. Kunicki,
R.M. Ulfig, D.R. Lenz, E.M. Strennen, E. Oltman, J.H. Bunton, D.R. Strait:
Microsc. Microanal. 10 (2004) 373.
[155] G.Sha,Y.B.Wang,X.Z.Liao,Z.C.Duan,S.P.Ringer and T.G. Langdon, Acta.
Mater. 57 (2009) 3123.
[156] P.Doig and J.W.Edington ,Phil. Mag. 28 (1973) 961.
[157] A.W.Nicholls and I.P. Jones, J. Phys. Chem. Solids 44 (1983) 671.
[158] D.L.Beke,I.Goedeny,G.Erdelyi and F.J. Kedves, Phil. Mag. A 56 (1987) 659.
[159] B.B.Straumal,B.Baretzky,A.A.Mazilkin,F.Phillipp,O.A.Kogtenkova,M.N.Volkov
and R.Z.Valiev,Acta. Mater. 52 (2004) 4469.
[160] A.A.Mazilkin,O.A.Kogtenkova ,B.B.Straumal ,R.Z.Valiev and B. Baretzky ,
Defect and Diffusion Forum 237-240(2005) 739.
[161] N.L.Peterson, S.J.Rothman, Phys. Rev. B 1(1970) 3264.
[162] I.Godeny, D.L.Beke,F.J.Kedves, Phys. Stat. Sol. A 13 (1972) K155.
[163] S.J. Rothman, N.L. Peterson, L.J. Nowicki and L.C. Robinson: Phys. Stat. Sol. B
63(1974) K29.
[164] G.Saada, Acta. Met. 9 (1961) 965.
[165] D.L.Beke,I.Goedeny and F.J.Kedves ,Phil. Mag. A 47 (1983) 281.
[166] D.L.Beke,I.Goedeny and F.J.Kedves, Trans. Jap. Inst. Met. Suppl. 27 (1986) 649.
[167] P.Zieba ,A.Pawlowski and W.Gust , Def. Diff. Forum 194 (2001) 1759.
206
[168] T.Fujita ,H.Hasegawa, Z.Horita and T.G.Langdon,Def. Diff. Forum 194 (2001)
1205.
[169] T.Fujita ,Z.Horita and T.G. Langdon ,Phil. Mag A 82 (2002) 2249.
[170] Y.Minamino ,Y.Toshimi ,A. Shinomura ,M.Shimada,M. Koizumi ,N.Ogawa ,J.
Takahashi and H.Kimura , J. Mater. Sci. 18 (1983) 2679.
[171] G.Erdelyi ,W. Lojkowski ,D.L. Beke,I. Godeny and F.J.Kedves , Philos. Mag. A
56 (1987) 673.
[172] J.Friedel , Dislocations. Oxford: Pergamon Press; 1964.
[173] T.H.Blewitt , R.R.Coltman ,J.K. Redman, J. Appl. Phys. 28 (1954) 651.
[174] M.Winterberger , Acta. Metall. 7 (1959) 549.
[175] G.Saada , Acta. Metall. 9 (1961) 166.
[176] G.Saada , Acta. Metall. 10 (1962) 551.
[177] M.Kiritani ,K. Yasunaga, Y.Matsukawa and M.Komatsu , Radiat. Effects. Def.
Sol. 157 (2002) 3.
[178] J.Cizek ,I.Prochazka ,B.Smola ,I.Stulikova ,R.Kuzel ,Z.Matej ,V.Cherkaska,R.K.I
smgaliev and O.Kulyasova ,Acta .Phys. Polon. A 107 (2005)738.
[179] J.Cizek ,I. Prochazka ,R. Kuzel ,Z.Matej ,V. Cherkaska ,M. Cieslar ,B.Smola ,I.
Stulikova , G.Brauer , W.Anwand ,R.K.Islamagaliev and
O.Kulyasova ,Acta .Phys. Polon. A 107 (2005) 745.
[180] R.Kuzel ,Z. Matej ,V. Cherkaska ,J. Pesicka ,J. Cızek ,I. Prochazka and
R.K.Islamgaliev , J. Alloys Compd. 378 (2004) 242.
[181] J.Cizek ,I. Prochazka ,G.Brauer ,W.Anwand ,R.Kuzcaronel ,M.Cieslar , and
R.K.Islamgaliev , Phys. Stat. Sol. A 195 (2003) 335.
[182] J.Cizek ,I. Prochazka ,M.Cieslar ,R. Kuzel ,J. Kuriplach ,F. Chmelik , I.Stulikova.,
F.Becvar and O.Melikhova , Phys. Rev. B. 65 (2002) 094106.
[183] T.Ungar, E.Schafler ,P. Hanak ,S. Bernstorff and M. Zehetbauer , Z. Metallkd. 96
(2005) 578.
[184] M.Zehetbauer ,E. Schafler and T. Ungar, Z. Metallkd. 96 (2005) 1044.
207
[185] A.A.Mazilkin,B.B.Straumal, O.A.Kogtenkova, R.Z. Valiev and B.Baretzky, Acta.
Materialia 54 (2006) 3933.
208
ALPHABETIZE BIBLIOGRAPHY
Ahmadeev N.H., Valiev R.Z.,Kopylov V.I. and Mulyukov R.R., Russian Metally 5 (1992)
96.
Aleksandrov I.V., Raab G.I. ,Shestakova L.O. ,Kil' mametov A.R. and Valiev R.Z., Phys,
Metals.Metall.93(2002) 493.
Allen R.M. and Vander Sande J.B. , Metall Trans A 9 (1978) 1251.
Beke D.L.,Goedeny I.,Erdelyi G. and Kedves F.J.,Phil. Mag. A 56 (1987) 659.
Beke D.L.,Goedeny I.and Kedves F.J.,Phil. Mag. A 47 (1983) 281.
Beke D.L.,Goedeny I.and Kedves F.J.,Trans. Jap. Inst. Met. Suppl. 27 (1986) 649.
Berg L.K., Gjønnes J., Hansen V., Li X.Z., Knutson-Wedel M., Waterloo G., Schryvers
D. and Wallenberg L.R., Acta. Mater. 49 (2001) 3443.
Blavette D., Bostel A., Sarrau J.M., Deconihout B. and Menand A., Nature 363 (1993)
432.
Blavette D.,Cadel E., Pereige C., Deconihout B. and Caron P., Microsc. Microanal. 13
(2007) 464.
Blewitt T.H., Coltman R.R, Redman JK. J. Appl. Phys. 28 (1954) 651.
Bridgman P.W.,J. Appl. Phys. 14 (1943) 273.
Buha J., Lumley R.N. and Crosky A.G., Mater. Sci. Eng. A 49 (2008) 1.
Bussiba A., BenArtzy A. ,Shtechman A., Ifergan S. and Kupiec M.,Mater Sci Eng A302
(2001) 56.
Cerezo A., Clifton P.H., Gomberg A. and Smith G.D.W., Ultramicroscopy 107 (2007)
720.
Cerezo A., Clifton P.H.,Lozano-Perez S., Panayi P., Sha G. and Smith G.D.W., Microsc.
Microanal. 13 (2007) 408.
Chinh N.Q.,Gubicza J. Czeppe T.,Lendvai J.,Xu C., Valiev R.Z. and Langdon T.G.,
Mater Sci.Eng A 2009, In press.
Cizek J., Prochazka I., Kuzel R.,Matej Z., Cherkaska V., Cieslar M.,Smola B., Stulikova
I.,Brauer G.,Anwand W.,Islamagaliev R.K.and Kulyasova O.,Acta .Phys. Polon. A 107
(2005) 745.
209
Cizek J., Prochazka I.,Brauer G.,Anwand W.,Kuzcaronel R.,Cieslar M., and Islamgaliev
R.K., Phys. Stat. Sol. A 195 (2003) 335.
Cizek J., Prochazka I.,Cieslar M., Kuzel R., Kuriplach J., Chmelik F. ,Stulikova I.,Becvar
F. and Melikhova O.,Phys. Rev. B. 65 (2002) 094106.
Cizek J., Prochazka I.,Smola B.,Stulikova I.,Kuzel R,Matej Z.,Cherkaska V.,Ismgaliev
R.K. and Kulyasova O. ,Acta .Phys. Polon. A 107 (2005)738.
Degischer H.P., Lacom W., Zahra A. and Zahra C., Mettalkd 71 (1980)231.
Dobatkin S.V. In: Zhu Y.T., Langdon T.G., Mishra R.S., Semiatin S.L., Saran M.J.,
Lowe T.C., editors. Ultrafine grained material II. Warrendale, PA: The Minerals, Metals
and Materials Society (2002) 83.
Doig P. and Edington J.W.,Phil. Mag. 28 (1973) 961.
Eckert J., Holzer J.C. and Johnson W.L., J. Appl. Phys. 73 (1993) 131.
Edalati K.,Fujioka T.and Langdon T.G.,Mater. Trans.50 (2009) 44.
Erdelyi G., Lojkowski W., Beke D.L., Godeny I. and Kedves F.J., Philos. Mag. A 56
(1987) 673.
Estrin Y.,Molotnikov A., Davies C. and Lapovok R., J. Mech. Phys. Solid. 56 (2008)
1186.
Fatay D., Bastarash E., Nyilas K., Dobatkin S., Gubicza J. and Ungar T., Z. Metallkd. 94
(2003) 842.
Figueiredo,R.B. and Langdon T.G.,Mater. Sci. Eng. A503 (2009),141.
Friedel J., Dislocations. Oxford: Pergamon Press; 1964.
Fujita T.,Hasegawa H., Horita Z. and Langdon T.G.,Def. Diff. Forum 194 (2001) 1205.
Fujita T.,Horita Z. and Langdon T.G., Mater. Sci. Eng. A 371 (2004) 241.
Fujita T.,Horita Z. and Langdon T.G.,Phil. Mag A 82 (2002) 2249.
Furukawa M, Horita Z., Nemoto M., Valiev R.Z. and Langdon T.G., Acta. Mater.
44(1996) 4619.
Furukawa M.,Horita Z. and Langdon T.G., J. Mater. Sci. 2005;40:909.
Furukawa M.,Iwahashi Y.,Horita Z.,Nemoto M. and LangdonT.G., Mater. Sci. Eng.A
257 (1998) 328.
Furukawa M.,Iwahashi Y.,Horita Z.,Nemoto M.,Tsenev N.K.,Valiev R.Z. and Langdon
T.G.,Acta. Mater.45(1997)4751.
210
Furukawa M.,Utsunomiya A.,Matsubara K.,Horita Z. and Langdon T.G.,Acta. Mater. 49
(2001) 3829.
Galiyev A. and Kaibyshev R., Scripta Mater. 51(2004) 89.
Gjønnes J. and Simesen C.J., Acta. Metall. 18 (1970) 881.
Godeny I.,Beke D.L.and Kerdves F.J., Phys. Stat. Sol. A 13 (1972) K155.
Gubicza J.,Schiller I.,Chinh N.Q.,Illy J.,Horita Z. and Langdon T.G.,Mater. Sci.
Eng.A.460-461(2007) 77.
Hatch J.E., Aluminum: Properties and Physical Metallurgy, Am.Soc.Met., Metal
Parks,Ohio,1985:356.
Horita Z. and Langdon T.G., Mater Sci Eng A410-411(2005) 422.
Horita Z., Fujinami T.,Nemoto M. and Langdon T.G.,Metall. Mater. Trans. A31 (2000)
691.
Horita Z.,Furukawa M.,Nemoto M.,Barnes A.J. and Langdon T.G., Acta. Mater. 48 (2000)
3633.
Horita Z.,Smith D.J., Furukawa M., Nemoto M.,Valiev R.Z. and Langdon T.G., J. Mater.
Res. 11 (1996) 1880.
Horita Z.,Smith D.J.,Furukawa M.,Nemoto M.,Valiev R.Z. and Langdon
T.G.,Mater.Charact.37 (1996) 285.
Huang J.C., Hsiao I.C., Wang T.D. and .Lou B.Y., Scripta Mater. 43(2000) 213.
Huang J.Y, Yu Y.D., Wu Y.K., Li D.X. and Ye H.Q., Acta. Mater.45 (1997) 113.
Islamgaliev R., Yunusova N.F. , Sabirov I.N. , Sergueeva A.V. and Valiev R.Z. ,Mater.
Sci. Eng .A319-321(2001) 877.
Ivanisenko Y., Lojkowski W., Valiev R.Z. and Fecht H-J. Acta Mater 51 (2003) 5555.
Ivanisenko Y., MacLaren I., Sauvage X.,Valiev R.Z. and Fecht H.J., Acta Mater. 54(2006)
1659.
Iwahashi Y. , Wang J.T., Horita Z., Nemoto M. and Langdon T.G., Scripta Mater.
35(1996) 143.
Iwahashi Y., Horita Z.,Nemoto M. and Langdon T.G., Metall. Mater. Trans. A29 (1998)
2503.
Iwahashi Y.,Furukawa M.,Horita Z.,Nemoto M. and Langdon T.G.,Metall..Mater.Trans
29A (1998) 2245 .
211
Iwahashi Y.,Horita Z.,Nemoto M.and Langdon T.G.,Acta. Mater.46 (1998) 3317.
Jiang H., Zhu Y.T., Butt D.P.,Alexandrov I.V. and Lowe T.C.,Mater. Sci. Eng. A290
(2000) 128.
Kaibyshev R.,Shipilova K., Musin F. and Motohashi Y., Mater. Sci. Tech. 21 (2005) 408.
Kelly T.F. and Miller M.K., Rev. Sci. Instrum. 78 (2007), 031101.
Kelly T.F., Gribb T.T., Martens J.D.,Shepard R.L.,Wiener J.D.,Kunicki S.A.,Ulfig
R.M.,Lenz D.R.,Strennen E.M.,Oltman E., Bunton J.H.,Strait D.R.: Microsc. Microanal.
10 (2004) 373.
Kim H.S.,Hong S.I.,Lee Y.S.,Dubravina A.A. and Alexandrov I.V., J. Mater. Process.
Tech. 142 (2003) 334.
Kim W-J., Chung S.W., Chung C.S. and Kum D., Acta. Mater. 49 (2001) 3337.
Kiritani M., Yasunaga K., Matsukawa Y. and Komatsu M., Radiat. Effects. Def. Sol. 157
(2002) 3.
Komura S., Berbon P.B.,Furukawa M.,Horita Z.,Nemoto M. and Langdon T.G., Scripta
Mater. 38 (1998) 1851.
Krakauer B.W. and Seidman D.N.,Acta. Mater. 46 (1998) 6145.
Krakauer B.W. and Seidman D.N.,Phys. Rev.B 48 (1993) 6724.
Krasilnikov N., Lojkowski W., Pakiela Z. and Valiev R.Z.. Mater. Sci. Eng. A397 (2005)
330.
Kubota K.,Mabuchi M. and Higashi K., J. Mater. Sci. 34 (1999) 2255.
Kuzel R., Matej Z., Cherkaska V., Pesicka J., Cızek J., Prochazka I, and Islamgaliev R.K.,
J. Alloys Compd. 378 (2004) 242.
Langdon T.G., Furukawa M., Nemoto M. and Horita Z., JOM 52 (2000) 30.
Langdon T.G., Metall. Trans. A 13 (1982) 689.
Langdon T.G.,Mater. Sci. Forum 15 (1999) 304.
Lee S., Berbon P.B.,Furukawa M., Horita Z., Nemoto M.,Tsenev N.K., Valiev R.Z. and
Langdon T.G., Mater. Sci. Eng. A272 (1999) 63.
Lee S.,Utsunomiya A.,Akamatsu H., Neishi K.,Furukawa M.,Horita Z.and Langdon
T.G.,Acta. Mater.50 (2002) 553.
212
Lefebvre W., Danoix F., Da Costa G., De Geuser F., Hallem H., Deschamps A. and
Dumont M.,Surf. Interface. Anal. 39 (2007) 206.
Li X. and Starink M.J., Mater. Sci. Forum 331-337(2000) 1071.
Li X.Z., Hansen V., Gjønnes J. and Wallenberg L.R., Acta. Mater. 47 (1999) 2651.
Loo P.T., Bastow T.J., Hill A.J., da Costa Teixeira J. and Hutchinson C.R.. In: Hirsch J,
Strotzki B., Gottstein G. editors. Aluminium Alloys – Their Physical and Mechanical
Properties, Vol. 1 (2008) 751.
Luis Perez C.J., Scripta Mater. 50 (2004) 387.
Luri R., Luis C.J., Leon J. and Sebastain M.A., J. Manufact. Sci.& Eng.128 (2006) 860.
Ma Y., Furukawa M., Horita Z., Nemoto M,. Valiev R.Z. and Langdon T.G., Mater.
Trans. 371 (1996)336.
Mabuchi M. and Higashi K. ,J.Mater.Sci.Lett.17(1998) 215.
Maloney M.K.,Hono K., Polmear I.J. and Ringer S.P., Scripta Mater. 41(1999) 1031.
Markushev M.V., Bampton C.C.,Murashkin M.Y. and Hardwick D.A., Mater. Sci. Eng.
A 234 (1997) 927.
Mazilkin A.A.,Kogtenkova O.A.,Straumal B.B.,Valiev R.Z. and Baretzky B., Defect and
Diffusion Forum 237-240(2005) 739.
Mazilkin A.A.,Straumal B.B. ,Kogtenkova O.A., Valiev R.Z. and Baretzky B., Acta.
Materialia 54 (2006) 3933.
McFadden S.X., Mishra R.S., Valiev R.Z., Zhilyaev A.P. and Mukherjee A.K., Nature
398 (1999) 684.
McKenzie P., Lapovok R. and Estrin Y., Acta Mater 55(2007) 2985.
Miller M.K., Cerezo A., Hetherlington M.G. and Smith G.D.W., Atom Probe Field Ion
Microscopy,Oxford Science, Oxford, (1996) 377.
Minamino Y., Toshimi Y., Shinomura A., Shimada M., Koizumi M.,Ogawa N.,
Takahashi J. and Kimura H., J. Mater. Sci. 18 (1983) 2679.
Mishin O.V., Juul Jensen D. and Hansen N., Mater. Sci. Eng. A 342 (2003) 320.
Miyahara Y.,Horita Z. and Langdon T.G.,Mater.Sci.Eng.A420(2006) 240.
Mohamed F.A., Acta. Mater. 51 (2003) 4107.
213
Mondolfo L.F., Gjostein N.A. and Levinson D.W.,Trans. Am. Inst. Min. (Metall) Engrs.
206 (1956)1378.
Morris D.G. and Munoz-Morris M.A.,Acta.Mater. 50(2002),4047.
Mukai T.,Yamanoi M., Watanabe H. and Higashi K., Scripta Mater. 45 (2001) 89.
Mukhopadhyay A.K., Yang Q.B. and Singh S.R., Acta. Metall. Mater. 42 (1994) 3083.
Murashkin M.Y.,Kil’Mametov A.R. and Valiev R.Z.,Phy. Metal. Metallography 106
(2008) 90.
Murayama M., Horita Z. and Hono K., Acta. Mater. 49(2001) 21.
Nagasekhar A.V.,Wei W., Yip T.H. and Chen G., Adv.Eng.Mater 9 (2007) 573.
Nakashima K.,Horita Z.,Nemoto M. and Langdon T.G.,Acta. Mater. 46 (1998)1589.
Nam C.Y., Han J.H., Chung Y.H. and Shin M.C., Mater. Sci. Eng. A 347(2003) 253.
Nazarov A.A., Romanov A.E. and Valiev R.Z., Acta. Metall. Mater. 41 (1993) 1033.
Nemoto M.,Horita Z., Furukawa M. and Langdon T.G., Mater Sci Forum 304-
306(1999);59.
Nicholls A.W. and Jones I.P., J. Phys. Chem. Solids 44 (1983) 671.
Oh-ishi K.,Hashi Y., Sadakata A.,Kaneko K., Horita Z. and Langdon T.G.,Mater. Sci.
Forum 396-402(2002) 333.
Oh-ishi K.,Horita Z.,Furukawa M.,Nemoto M. and Langdon T.G.,Metall.Mater.Trans
A29 (1998) 2011.
Oh-ishi K.,Horita Z.,Smith D.J. and Langdon T.G.,J.Mater.Res.16(2001)583.
Ota S.,Akamatsu H.,Neishi K.,Furukawa M.,Horita Z. and Langdon
T.G. ,Mater.Trans.43(2002) 2364.
Perevezentsev V.N.,Chuvil’deev V.N., Sysoev A.N.,Kopylov V.I. and Langdon T.G.,
Phys. Metals. Metall. 94 (2002) S45.
Perlitz H. and Westgren A., Arkiv för Kemi, Mineralogy och Geologi 6B (1943) 1.
Pilling J.,Ridley N., Superplasticity in Crystalline Solids, The Institute of Metals, London
1989.
Polakowski N.H. and Ripling E.J.,Strength and structure of engineering materials.
Englewood Cliffs, NJ: Prentice-Hall; 1966.
214
Prokoshkin S.D., Khmelevskaya I.Y.,Dobatkin S.V., Trubitsyna I.B.,Tatyanin
E.V.,Stolyarov V.V. and Prokofiev E.A., Acta Mater. 53 (2005) 2703.
Rothman S.J., Peterson N.L., Nowicki L.J. and Robinson L.C., Phys. Stat. Sol. B 63
(1974) K29.
Rybalchenko O.V., Dobatkin S.V.,Kaputkina L.M., Raab G.I. and Krasilnikov N.A.,
Mater. Sci. Eng. A244-248 (2004) 387.
Ryum N. ,Z. Metallkd. 66 (1975) 338.
Saada G., Acta. Met. 9 (1961) 965.
Saada G., Acta. Metall. 10 (1962) 551.
Saada G., Acta. Metall. 9 (1961) 166.
Saito N.,Mabuchi M., Naksnishi M., Shigematsu I., Yamauchi G. and Nakamura M., J.
Mater. Sci. 36 (2001) 3229.
Saito Y.,Tsuji N., Utsunomiya H., Sakai T. and Hong R.G., Scripta Mater. 39 (1998)
1221.
Sakai G.,Horita Z. and Langdon T.G.,Mater. Sci. Eng. A393 (2005) 344.
Sakai G.,Horita Z. and Langdon T.G.,Mater.Trans.45(2004) 3079.
Salem A.A., Langdon T.G.,McNelley T.R.,Kalidindi S.R. and Semiatin S.L., Metall.
Mater. Trans.A 37 (2006) 2879.
Sauvage X.,Wetscher F. and Pareige P.,Acta. Mater. 53 (2005) 2127.
Segal V.M., Reznikov V.I., Drobyshevskiy A.E. and Kopylov V.I.,Russian Metall. 1
(1981) 99.
Semiatin S.L.,Segal V.M.,Goforth R.E.,Frey N.D.and DeLo D.P. ,Metall.Mater.Trans
A30 (1999) 1425.
Senkov O.N., Froes F.H., Stolyarov V.V. and Valiev R.Z., Nanostruct. Mater. 10 (1998)
691.
Sha G. and Cerezo A., Acta. Mater. 52 (2004) 4503.
Sha G. and Cerezo A., Acta. Mater. 53 (2005) 907.
Sha G. and Cerezo A., Surf. Interface. Anal. 36 (2004) 564.
Sha G. and Ringer S.P., Ultramicroscopy 109 (2009) 580.
215
Sha G.,Wang Y.B., Liao X.Z.,Duan Z.C.,Ringer S.P. and Langdon T.G., Acta. Mater.
57 (2009) 3123.
Shabashov V.A., Korshunov L.G., Mukoseev A.G., Sagaradze V.V.,Makarov A.V.,
Pilyugin V.P, Novikov S.I. and Vildanova N.F., Mater. Sci. Eng. A 346 (2003) 196.
Shabashov V.A., Mukoseev A.G. and Sagaradze V.V., Mater. Sci. Eng. A 307 (2001) 91.
Stiller K., Warren P.J., Hansen V. , Angenete J. and Gjonnes J. , Mater. Sci. Eng. A270
(1999) 55.
Stolyarov V.V. , Shestakova L.O., Zhu Y.T.and Valiev R.Z., Nanostruct. Mater. 12(1999)
923.
Stolyarov V.V., Zeipper L., Mingler B.and Zehetbauer M.,Mater. Sci. Eng. A,476( 2008),
98.
Stolyarov V.V., Zhu Y.T., Lowe T.C., Islamgaliev R.K. and Valiev R.Z., Nanostruct.
Mater. 11 (1999) 947.
Straumal B.B.,Baretzky B.,Mazilkin A.A.,Phillipp F.,Kogtenkova O.A.,Volkov M.N.and
Valiev R.Z.,Acta. Mater. 52 (2004) 4469.
Strawbridge D.J., Hume-Rothery W. and Little A.T., J. Inst. Metals. 74 (1948) 191.
Tan J.C. and Tan M.J.,Mater. Sci. Eng. A 339 (2003) 81.
Tcherdyntsev V.V.,Kaloshkin S.D., Gunderov D.V.,Afonina E.A., Brodova I.G. and
Stolyarov V.V, Baldokhin Y.V., Shelekhov E.V. and Tomilin I.A., Mater. Sci. Eng.
A375-377 (2004) 888.
Terhune S.D., Swisher D.L., Oh-ishi K., Horita Z., Langdon T.G.and T.R.McNelley.
Metall Mater Trans A33 (2002) 2173.
Tóth L., Mlinari A. and Estrin Y., J. Eng. Mater. Tech. 124(2002) 71.
Ungar T., Schafler E., Hanak P., Bernstorff S. and Zehetbauer M., Z. Metallkd. 96 (2005)
578.
Valiev R.Z. and Langdon T.G.,Prog. Mater. Sci. 51 (2006) 881.
Valiev R.Z., Islamgaliev R.K. and Alexandrov I.V., Prog. Mater. Sci. 45 (2000) 103
Valiev R.Z., Ivanisenko Y.V., Rauch E.F. and Baudelet B., Acta Mater. 44(1996) 4705.
Valiev R.Z., Korznikov A.V. and Mulyukov R.R., Mater. Sci. Eng. A168 (1993) 141.
Valiev R.Z., Mater. Sci. 21 (1996) 369.
216
Valiev R.Z.,Estrin Y.,Horita Z. ,Langdon T.G , Zehetbauer M.J. and Zhu Y.T., JOM 58
(2006);33.
Valiev R.Z.,Islamgaliev R.K. and Alexandrov I.V., Prog. Mater. Sci. 45 (2000) 103.
Valiev R.Z.,Salimonenko D.A.,Teenev N.K.,Berbon P.B.and Langdon T.G., Scripta
Mater. 37 ( 1997) 1945.
Vorhauer A. and Pippan R., Scripta Mater. 51(2004) 921.
Wang J.T.,Horita Z.,Furukawa M., Nemoto M., Tsenev N.K., Valiev R.Z., Ma Y. and
Langdon T.G., J. Mater. Res.8 (1993) 2810.
Wang J.T.,Iwahashi Y.,Horita Z.,Furukawa M.,Nemoto M.,Valiev R.Z. and Langdon
T.G., Acta.Mater.,44(1996) 2973.
Watanabe T., Res. Mech. 11(1984) 47.
Wetscher F. , Vorhauer A., Stock R. and Pippan A.,Mater. Sci. Eng. A 387-389 (2004)
809.
Wetscher F. ,Pippan R., Sturm S., Kauffmann F.,Scheu C. and Dehm G., Metall. Mater.
Trans. A37(2006) 1963.
Winterberger M., Acta. Metall. 7 (1959) 549.
Xu C. and Langdon T.G., Scripta Mater. 48 (2003) 1.
Xu C. and Langdon T.G., J. Mater. Sci. 42 (2007) 1542.
Xu C., Dixon W., Furukawa M., Horita Z. and Langdon T.G. ,Mater. Letters 57 (2003)
3588.
Xu C.,Furukawa M., Horita Z. and Langdon T.G., Acta. Mate.r 53 (2005) 749.
Xu C.,Furukawa M.,Horita Z. and Langdon T.G., Acta. Mater. 51 (2003) 6139 .
Xu C.Horita Z. and Langdon T.G.,Acta.Mater. 56 (2008) 5168.
Yamashita A. ,Yamaguchi D., Horita Z. and Langdon T.G., Mater. Sci. Eng. A287 (2000)
100.
Yang B., Vehoft H.,Hohenwarter A.,Hafok M. and Pippan R., Scripta Mater.58(2008)
790.
Yang Z. and Welzel U., Mater. Lett. 59 (2005) 3406.
Zehetbauer M., Schafler E. and Ungar T., Z. Metallkd. 96 (2005) 1044.
217
Zhao Y.H, Zhu Y.T., Liao X.Z., Horita Z. and Langdon T.G., Appl Phys. Lett. 89 (2006)
121906.
Zhao Y.H., Horita Z.,Langdon T.G. and Zhu Y.T., Mater. Sci. Eng. A474 (2008) 342.
Zhao Y.H., Liao X.Z., Jin Z., Valiev R.Z. and Zhu Y.T., In: Zhu Y.T., Langdon T.G.,
Valiev R.Z., Semiatin S.L., Shin D.H., Lowe T.C., editors. Ultrafine-grained materials III.
The Minerals, Metals and Materials Society (2004) 511.
Zhao Y.H., Liao X.Z., Zhu Y.T.,Horita Z. and Langdon T.G.,Mater. Sci. Eng. A410-
411( 2005) 188.
Zhao Y.H., Zhu Y.T., Liao X.Z., Horita Z. and Langdon T.G., Mater. Sci. Eng. A 463
(2007) 22.
Zhao Y.H.,Liao X.Z.,Jin Z.,Valiev R.Z.and Zhu Y.T.,Acta. Mater. 52 (2004) 4589.
Zheng L.J., Li H.X., Hashmi M.F., Chen C.Q., Zhang Y. and Zeng M.G.. J. Mater.
Processing Tech. 171 (2006) 100.
Zhilyaev A.P. and Langdon T.G., Prog. Mater. Sci. 53 (2008) 893.
Zhilyaev A.P., Kim B.K., Szpunar J.A., Baró M.D. and Langdon T.G., Mater Sci Eng
A391 (2005) 377.
Zhilyaev A.P., Gubicza J., Nurislamova G.V., Révész A., Suriñach S. Baró M.D. and
Ungar T. Phys. Status Solidi. A 198 ( 2003) 263.
Zhilyaev A.P., Lee S., Nurislamova G.V.,Valiev R.Z. and Langdon T.G., Scripta Mater.
44 (2001) 2753.
Zhilyaev A.P., McNelley T.R. and Langdon T.G., J. Mater. Sci. 42 (2007) 1517.
Zhilyaev A.P., Nurislamova G.V., Baró M.D., Valiev R.Z. and Langdon T.G., Metall.
Mater. Trans. A33 (2002) 1865.
Zhilyaev A.P., Nurislamova G.V., Kim B.K., Baró M.D., Szpunar J.A. and Langdon T.G.,
Acta. Mater. 51(2003) 753.
Zhilyaev A.P.,Oh-ishi K.,Langdon T.G. and McNelley T.R., Mater. Sci. Eng, A 410-411
(2005) 277.
Zhu Y.T. and Langdon T.G., JOM 56 (2004) 58.
Zieba P., Pawlowski A. and Gust W., Def. Diff. Forum 194 (2001) 1759.
Abstract (if available)
Abstract
The super-saturated Al-Zn-Mg-Cu alloys or 7000 series Al alloys are increasingly replacing the 2000 series Al alloys in aerospace applications due to its outstanding strength. The potential of further strengthening the 7000 series Al alloys using SPD methods makes it tempting to conduct ECAP as well as HPT on the selected Al-7136 alloy.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Microstructural evolution and mechanical properties of a copper-zirconium alloy processed by severe plastic deformation
PDF
Microstructural evolution and mechanical properties of an aluminum alloy processed by equal-channel angular pressing and high-pressure torsion
PDF
Properties of a ZK60 magnesium alloy processed by severe plastic deformation
PDF
Mechanical behavior and deformation of nanotwinned metallic alloys
PDF
Tailoring compositional and microstructural complexity in nanostructured alloys
PDF
Mechanical behavior and microstructure optimization of ultrafine-grained aluminum alloys and nanocomposites
PDF
Thermal properties of silicon carbide and combustion mechanisms of aluminum nanoparticle
PDF
Synthesis, characterization, and mechanical properties of nanoporous foams
PDF
Mechanical behavior of materials in extreme conditions: a focus on creep plasticity
PDF
Expanding sputtering synthesis domains for the discovery of novel film microstructures and morphologies
PDF
Modeling of bicontinuous nanoporous materials and investigation of their mechanical and hydraulic properties
PDF
Atomistic modeling of the mechanical properties of metallic glasses, nanoglasses, and their composites
Asset Metadata
Creator
Duan, Zhichao
(author)
Core Title
Investigation of the influence of severe plastic deformation on the microstructure and mechanical properties of Al-7136 alloy
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Mechanical Engineering
Degree Conferral Date
2009-08
Publication Date
08/04/2009
Defense Date
06/15/2009
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Al alloy,ECAP,HPT,mechanical properties,microstructure,OAI-PMH Harvest,severe plastic deformation
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Langdon, Terence G. (
committee chair
), Goo, Edward K. (
committee member
), Hodge, Andrea M. (
committee member
)
Creator Email
kvduan@gmail.com,zduan@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m2465
Unique identifier
UC1469895
Identifier
etd-Duan-3063 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-183884 (legacy record id),usctheses-m2465 (legacy record id)
Legacy Identifier
etd-Duan-3063.pdf
Dmrecord
183884
Document Type
Dissertation
Rights
Duan, Zhichao
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
Al alloy
ECAP
HPT
mechanical properties
microstructure
severe plastic deformation