University of Southern California Dissertations and Theses

Deformation behavior and microstructural evolution of nanocrystalline aluminum alloys and composites

(USC Thesis Other)

Deformation behavior and microstructural evolution of nanocrystalline aluminum alloys and composites

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

DEFORMATION BEHAVIOR AND MICROSTRUCTURAL EVOLUTION OF
NANOCRYSTALLINE ALUMINUM ALLOYS AND COMPOSITES

by

Byungmin Ahn

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
(MATERIALS SCIENCE)

August 2008

Copyright 2008 Byungmin Ahn

ii
Acknowledgements

First of all, I would like to thank my advisor, Professor Steven R. Nutt, for his
great guidance, encouragement and support during my study for Ph.D. I believe that this
work would not have been accomplished without his valuable advice. I sincerely respect
his enthusiasm for research and dedication to education. He has always provided
unlimited accessibility to discussion as well as immediate assistance in facilitating
research works. He has helped me developing my professional skills and knowledge.
I would also like to thank Professor Enrique J. Lavernia and Professor Julie M.
Schoenung at University of California, Davis for their support and advice in the research
program.
I would also like to thank my other committee members, Professor Edward Goo
and Professor Charles G. Sammis for their time and effort to evaluate my thesis.
I am sincerely grateful to Dr. Zonghoon Lee at Lawrence Berkeley National
Laboratory for his valuable advice and warm suggestion. My appreciation is extended to
Mr. Warren Haby, a laboratory manager at University of Southern California, for his
great advice and technical support facilitating experiments.
A special expression of appreciation is extended to my parents and sister for their
encouragement and endless love. In addition, I would like to appreciate my family-in-law
for their encouragement.
Finally, I would like to express my sincere gratitude to my wife, Megumi
Kawasaki, for her everlasting love and unstinting support, with best wishes for her first
delivery.
iii
Table of Contents

Acknowledgements

List of Tables

List of Figures

Abstract

Chapter 1. Introduction
References

Chapter 2. Literature Review for Nanocrystalline Materials
2.1. Definition
2.2. Synthesis
2.2.1. Overview
2.2.2. Cryomilling
2.2.2.1. Formation of nanostructure
2.2.2.2. Thermal stability
2.2.2.3. Consolidation
2.3. Mechanical properties
2.3.1. Yield strength
2.3.2. Inverse Hall-Petch effect
2.3.3. Ductility
2.3.4. Strain hardening
2.3.5. Strain rate sensitivity
2.4. Deformation mechanisms
2.4.1. Dislocation pile-up
2.4.2. Grain boundary sliding
2.4.3. Grain boundary rotation and grain coalescence
2.4.4. Shear band formation
2.4.5. Grain boundary dislocation creation and annihilation
References

Chapter 3. Novel Experimental Techniques
3.1. Nanoindentation
3.1.1. Motivation
3.1.2. Equipment and basic method
3.1.3. Indentation hardness and modulus
3.1.4. Load-displacement curves
3.1.5. Factors affecting nanoindentation data
ii
vi
vii
xiv
1
5
6
6
8
8
12
12
15
20
20
21
23
27
30
31
33
33
35
37
39
40
42
50
50
50
50
54
56
58
iv
3.2. Digital image correlation
3.3. Micro-tensile test
3.4. Sample preparations for electron microscopy
3.4.1. Tripod polishing for TEM
3.4.2. Cross section polishing for SEM
References

Chapter 4. Nanocrystalline Al Alloys via Cryomilling
4.1. Degassing parameter effects
4.1.1. Motivation
4.1.2. Degassing details and characterization procedures
4.1.3. Experimental results and discussion
4.1.4. Conclusions
4.2. Consolidation and forming method effects
4.2.1. Motivation
4.2.2. Materials and processing details
4.2.3. Mechanical tests and characterization procedures
4.2.4. Experimental results and discussion
4.2.5. Conclusions
4.3. Strain rate sensitivity
4.3.1. Motivation
4.3.2. Materials and processing details
4.3.3. Mechanical tests and characterization procedures
4.3.4. Experimental results and discussion
4.3.5. Conclusions
References

Chapter 5. Al Alloys with Bimodal Microstructure
5.1. Multi-scale concept
5.1.1. Microstructure
5.1.2. Tensile behavior
5.2. Strain localization
5.2.1. Motivation
5.2.2. Experimental procedures
5.2.3. Experimental results and discussion
5.2.4. Conclusions
References

Chapter 6. Al Nanocomposites Reinforced with B
4
C Particulates
6.1. Multi-scale concept
6.1.1. Microstructure
6.1.2. Compressive behavior
6.2. Microstructural evolution during cryomilling
6.2.1. Motivation

58
63
65
65
71
76
77
77
77
78
80
95
96
96
97
99
101
115
116
116
117
118
119
129
131
133
133
133
136
139
139
141
145
158
159
161
161
161
166
171
173
v
6.2.2. Experimental procedures
6.2.3. Experimental results and discussion
6.2.4. Conclusions
References

Chapter 7. Suggestions for Future Works

Bibliography

173
174
182
183
185
187

vi
List of Tables

Table 2.1 Leading review articles on nanocrystalline materials. 7
Table 4.1 Number-based mean grain size and aspect ratio of grains of the as-
cryomilled and degassed powders.
84
Table 4.2 Summary of material characteristics after forging. 89
Table 4.3 Impurity contents of the forged materials. 94
Table 4.4 Chemical analysis of rolled Al 5083 plates. Values in wt.%, except H
(ppm).
102
Table 4.5 Mean grain size and aspect ratio of rolled Al 5083 plates. 106
Table 4.6 Tensile properties of rolled Al 5083 plates. 110
Table 4.7 Fracture toughness of rolled Al 5083 plates. 111
Table 4.8 Mean grain size and aspect ratio of the as-cryomilled NC powder,
forged UFG plate, and standard Al 5083 plate (measured from TEM
micrographs and optical micrographs).
123
Table 6.1 Room temperature mechanical properties of the trimodal
nanocomposite, compared with a conventional CG material and a
cryomilled NC material.
167

vii
List of Figures

Figure 2.1 Schematic illustration of strategies for synthesis of nanostructured
materials.
9
Figure 2.2 Schematic diagram of an attrition mill used for mechanical alloying
[53].
13
Figure 2.3 Ball-powder-ball collision of powder mixture during mechanical
alloying [53].
14
Figure 2.4 Schematic representation of grain refinement mechanism in
microscopic level during ball milling [56].
16
Figure 2.5 High-resolution TEM images of AlN dispersoids: (a) a single
dispersoid and (b) multiple disporsoids adjacent to grain boundary
[59].
18
Figure 2.6 Grain size variation as a function of annealing temperature for two
annealing times [60].
19
Figure 2.7 Young’s modulus as a function of porosity for NC Pd and Cu [26]. 22
Figure 2.8 Summary of experimental data from references for NC Cu [70]; A:
hardness divided by 3 [26], B: hardness divided by 3 [64], C: the
solid like is the Hall-Petch relation ship for the yield strength of Cu,
σ
y
(MPa) = 25.5 + 3478.5 d
-1/2
[71], D: tensile yield strength [70],
E: tensile yield strength [72], F: compressive yield strength [64], G:
tensile yield strength [64], H: tensile yield strength [73], I: yield
strength by miniaturized disk bend test [68], and J: Compressive
yield strength [74].
24
Figure 2.9 Schematic plot of the variation of yield stress as a function of grain
size from microcrystalline to NC regimes [13].
25
Figure 2.10 Comparison of yield strength versus elongation for (a) NC materials
and (b) UFG materials [96].
29
Figure 2.11 Stress-strain response of UFG materials at different strain rates: (a)
Cu, (b) Ni and (c) Al-4Cu-0.5Zr [111].
32
Figure 2.12 Braking up of dislocation pile-up: (a) microcrystalline regime and
(b) NC regime [16].
34
viii
Figure 2.13 Schematic illustration of grain boundary sliding model (a) initial
position of grains and (b) position after top layer has slid to right
[16].
36
Figure 2.14 Grain rotation and grain coalescence and during plastic deformation
leading to creation of elongated grains by annihilation of grain
boundary [16].
38
Figure 2.15 Grain boundary source-sink model [16]. 41
Figure 3.1 SEM image of an indentation made by Berkovich indenter tip and
(b) indentation parameters of the Berkovich tip [3].
52
Figure 3.2 Schematic plot of an indentation load-displacement curve showing
important parameters [2].
53
Figure 3.3 Schematic representation of a section through an indentation
showing various quantities used in the analysis [1].
55
Figure 3.4 Schematic examples of load-displacement curves for different
material responses and properties: (a) elastic solid, (b) brittle solid,
(c) ductile solid, (d) crystalline solid, (e) brittle solid with cracking
during loading, and (f) polymer exhibiting creep [3].
57
Figure 3.5 Schematic concept of image correlation technique showing planar
deformation from the original state (S) to the deformed state (S
1
)
[12].
60
Figure 3.6 DIC analysis showing horizontal strain of Al specimen during
deformation recorded at (a) the beginning and (b) the end of the
deformation. A single camera was used to measure the deformation.
[13].
62
Figure 3.7 Micro-tensile test module. 64
Figure 3.8 Experimental apparatus for dynamic observation of deformation,
showing (a) computer-controlled micro-tensile module attached to a
light microscope and (b) close view of the specimen on the module
with an objective lens focusing on the surface of specimen.
66
Figure 3.9 Tripod polisher. 68
Figure 3.10 Schematic of tripod polisher alignment for wedge shape production:
(a) first-side polishing and (b) second-side polishing. The point of
interest is represented by star.
69
ix
Figure 3.11 (a) TEM and (b) SEM micrographs of bulk nanocrystalline Al-Ti-
Cu alloy [14].
70
Figure 3.12 SEM micrographs of cross section of a multi-layered card edge
connector prepared by (a) the CP method and (b) the conventional
mechanical polishing [15].
73
Figure 3.13 SEM backscattered electron image of cryomilled Al powder cross-
sectioned by the CP method showing (a) pores between powder
particles and (b) nanocrystalline and coarse-grain structures as well
as sub-micron voids.
75
Figure 4.1 SEM backscattered electron image of cross section of as-cryomilled
Al powder showing both NC and relatively CG structures with
nano-size voids.
81
Figure 4.2 TEM microstructure of as-cryomilled powder: (a) a bright-field
image showing mostly NC grains with arrows indicating the
boundary with a relatively CG region, (b) corresponding SAD
pattern; (c) histogram of the grain size distribution; (d) histogram of
the grain aspect ratio distribution.
82
Figure 4.3 TEM microstructures of the powder degassed at the highest
temperature (max. T = 500 ºC) showing neighboring CG and UFG
regions; (a) bright field image with corresponding SAD pattern and
(b) dark field image.
85
Figure 4.4 TEM bright field images and graphs of the measured grain size
distribution of cryomilled powder degassed at different
temperatures: (a) 415, (b) 440 and (c) 500 ºC.
87
Figure 4.5 Optical micrographs of as-CIPped powder degassed previously at
(a) low (415 ºC), (b) medium (440 ºC) and (c) high temperatures
(500 ºC).
88
Figure 4.6 Microstructure of forged material made from CIPped powder
degassed at the highest temperature (max. T = 500 ºC): (a) optical
(CG regions lighter-toned), and (b) bright field TEM (showing a
CG band perpendicular to the forging axis).
91
Figure 4.7 Histograms of the measured grain size distribution of the as-forged
materials previously degassed at different temperatures: (a) 415, (b)
440 and (c) 500 ºC.
92
x
Figure 4.8 Optical micrographs (unetched) of cryomilled Al 5083 plate
showing pores (black) viewed normal to the rolling direction: (a)
HIP/extruded and (b) QI forged.
103
Figure 4.9 TEM micrographs, with grain size distributions, of HIP/extruded
cryomilled Al 5083 plate viewed in three directions with respect to
processing: (a) rolling, (b) extrusion, and (c) transverse axes.
104
Figure 4.10 TEM micrographs, with grain size distributions, of QI-forged
cryomilled Al 5083 plate viewed in three directions with respect to
processing: (a) normal to forging and rolling, (b) rolling, and (c)
forging (transverse) axes.
107
Figure 4.11 Optical micrographs (etched) of standard Al 5083 H131 plate:
viewed normal to the primary rolling direction (R1): (a) normal to
the transverse axis (R2) and (b) parallel to the transverse axis.
108
Figure 4.12 SEM fractographs for crack surfaces in the plane of rolled
cryomilled Al 5083 plates: (a) HIP/extruded (T
R
) and (b) QI forged
(F
R
).
113
Figure 4.13 SEM fractograph for a crack in the plane (F
L
) of standard Al 5083
H131 plate.
114
Figure 4.14 TEM bright field micrographs of (a) cryomilled NC powder and (b)
forged UFG plate with corresponding SAD patterns, respectively.
120
Figure 4.15 Histograms of the grain size distribution of (a) cryomilled NC
powder and (b) forged UFG plate when viewed normal to the
forging direction.
122
Figure 4.16 Load vs. indentation depth curves performed at a loading rate of
1,000 µN/s for: (a) conventional CG plate, (b) UFG plate and (c)
NC powder.
124
Figure 4.17 Hardness variations of the cryomilled NC powder, forged UFG and
conventional CG plate with respect to the loading rates measured by
nanoindentation.
126
Figure 4.18 Compression behavior of the UFG and CG plates: (a) strain rate
sensitivity exponent and (b) true stress-strain behavior with respect
to different strain rates.
127
xi
Figure 5.1 Schematic of synthesis of a multi-scale grain structured material,
incorporating CG material into cryomilled NC material to improve
ductility and toughness.
134
Figure 5.2 Optical micrographs of as-extruded bimodal alloys containing 15,
30 and 50% CG content, (a), (b) and (c), respectively, when viewed
from normal and transverse to the extrusion direction. (brighter tone
indicates CG regions in darker NC matrix).
135
Figure 5.3 TEM micrographs showing global microstructure of as-extruded
bimodal alloy containing 50% CG: (a) a bright field image and (b) a
dark field image.
137
Figure 5.4 Comparison of tensile stress-strain curves between experimental
behavior [15] and calculated behavior by FEM simulation [18] of
bimodal materials as well as 100% NC and conventional CG
materials.
138
Figure 5.5 Geometry of the double-notched tensile specimen showing
experimental scheme and (b) stress-strain curve of the notched
specimen showing the points interrupted for hardness
measurements.
143
Figure 5.6 Microstructures of the as-forged material showing a multi-scale
grain structure: (a) and (b) optical micrographs (etched), (c) a bright
field TEM image, and (d) a dark field TEM image.
146
Figure 5.7 Nominal stress-strain behavior of the notched bimodal specimen
with 50% CG, compared with engineering stress-strain curves of a
100% NC [15] and a conventional Al 5083 alloy [15].
147
Figure 5.8 Successive optical micrographs recorded around a notch during
tensile deformation: (a) initial state before deformation started (t
1
=
0 s); (b) voids nucleated short time after yielding (t
2
= 80 s); (c)
fracture started (t
3
= 100 s); and (d) immediately before rupture of
the specimen (t
4
= 110 s).
149
Figure 5.9 Strain fields around a notch measured by DIC at different strains:
(a) large strain (1.25×10
-2
, 30 second interval), (b) intermediate
strain (2.51×10
-3
, 6 second interval), and (c) micro-strain (4.18×10
-
4
, 1 second interval).
150
Figure 5.10 Optical micrographs and corresponding DIC results showing
inhomogeneous strain measured from two locations: (a) near the
notches and (b) central part between the notches.
152
xii
Figure 5.11 (a) Optical micrograph and (b) corresponding strain field showing
strain concentration (arrows) in CG regions adjacent to porosity
(circles).
154
Figure 5.12 Vickers hardness variation measured from both UFG and CG
regions at two locations (near the notches and between the notches)
with respect to the nominal tensile stress.
156
Figure 5.13 SEM micrographs of tensile fracture surfaces of the notched
specimen showing two different surface morphologies comprised of
small and large dimples, when viewed from directions (a) normal to
the surface and (b) tilted from the surface.
157
Figure 6.1 Schematic of synthesis of a multi-scale trimodal composite,
incorporating CG material into cryomilled NC and B
4
C to improve
ductility and toughness.
163
Figure 6.2 Optical micrographs of the as-extruded trimodal composite, when
viewed from (a) normal to the extrusion direction and (b) transverse
to the extrusion direction. CG Al regions appear brighter tone in
gray NC Al matrix containing black B
4
C particles.
164
Figure 6.3 TEM micrographs of the as-extruded trimodal composite: (a) bright
field image showing global view from normal to the extrusion
direction, (b) selected area diffraction (SAD) patterns of NC and
B
4
C from their interface, and (c) and (d) bright field images
showing B
4
C particles primarily residing in NC Al matrix without
direct contact with CG regions.
165
Figure 6.4 True stress-strain curve for the annealed trimodal composite tested
at ambient temperature at a strain rate of 10
-3
s
-1
.
168
Figure 6.5 TEM bright field image showing clear interface between B
4
C
particles and NC matrix as well as corresponding SAD patterns
from NC regions and a B
4
C particle, respectively.
170
Figure 6.6 (a) True stress-strain curves for the trimodal composite tested at
various elevated temperatures and (b) corresponding temperature
dependence of the yield strength.
172
Figure 6.7 SEM images showing the morphology of the powders: (a) as-
received gas-atomized Al 5083 powder and (b) Al/B
4
C composite
powder cryomilled for 8 hours.
175
xiii
Figure 6.8 Optical micrographs of Al/B
4
C composite powders showing cross
sectional morphology of powders cryomilled for (a) 1 hour, (b) 2
hours, (c) 4 hours, (d) 6 hours, (e) 8 hours, and (f) 10 hours,
respectively.
176
Figure 6.9 Variation in the mean particle size during cryomilling with respect
to the milling time, determined statistically from optical
micrographs: (a) the Al/B
4
C composite powder and (b) B
4
C
particles. The error bars represent the standard deviation.
178
Figure 6.10 Optical micrographs of Al/B
4
C composite powders showing
distribution of B
4
C particles within the NC Al matrix: powders
cryomilled for (a) 1 hour, (b) 2 hours, (c) 4 hours, (d) 6 hours, (e) 8
hours, and (f) 10 hours, respectively.
179
Figure 6.11 Variation in the mean grain size of Al 5083 during cryomilling with
respect to the milling time and corresponding TEM bright field
images showing microstructure of powders cryomilled for 2 hours
and 8 hours.
181

xiv
Abstract

Nanocrystalline or ultrafine-grained Al alloys are often produced by severe plastic
deformation methods and exhibit remarkably enhanced strength and hardness compared
to conventional coarse-grained materials, resulting in great potential for structural
applications. To achieve nanocrystalline structure, grains were refined by cryomilling
(mechanical milling at cryogenic temperature) pre-alloyed powders. Cryomilling
provides capability for rapid grain refinement and synthesis of commercial quantities
(30 −40 kg). The cryomilled powder was primarily consolidated by hot or cold isostatic
pressing in general. Secondary consolidation was achieved by extrusion or forging.
Alternatively, quasi-isostatic forging was applied either as an initial consolidation or as a
further deformation step.
To improve insufficient ductility and toughness of nanocrystalline materials, an
intelligent design with microstructural modification was introduced by generation of
multiple size scales. A bimodal grain structure consisting of nanocrystalline grains and
inclusions of coarse-grained material was produced by consolidation of blended powders.
The resulting materials exhibited enhanced ductility compared to 100% nanocrystalline
materials, with only moderate decreases in strength. A similar process was used to
produce hybrid trimodal microstructures comprised of regions of nanocrystalline and
coarse grains, as well as hard ceramic particles, providing super-high compressive
strength.
For cryomilled nanocrystalline Al alloys, effects of degassing temperature were
investigated in terms of microstructural evolution. Higher degassing temperatures
xv
resulted in higher density and lower hydrogen content, which can reduce loss of
toughness in consolidated materials. Different consolidation methods were compared
with regard to the relation between the microstructures and mechanical properties. Quasi-
isostatic forging led to greater and more isotropic fracture toughness, compared with
other processing routes. Strain rate sensitivity in room temperature deformation was
examined as a function of grain size using nanoindentation. Negative strain rate
sensitivity was observed in nanocrystalline and ultrafine-grained materials, while a
conventional alloy was strain rate insensitive.
For multi-scale materials, local displacements in bimodal materials during tensile
deformation were measured by digital image correlation. Inhomogeneous strain behavior
was observed between nanocrystalline and coarse-grained regions and attributed to
differences in dislocation plasticity. In the Al matrix nanocomposite with hybrid
microstructures, microstructural evolution of the composite powder with boron-carbide
reinforcements was investigated as a function of milling time.
1
Chapter 1. Introduction

For the past few decades, interest has increasingly focused on nanostructured
materials due to their particular structure and the related novel physical, chemical and
mechanical properties [1 −5]. Nanostructured materials (or nanocrystalline materials,
synonymously) are materials with the characteristic length scale of typically less than 100
nm. Nanocrystalline (NC) materials are composed of nanometer-sized crystallites with
different crystallographic orientations. Because of the extremely fine grain sizes in NC
materials, a large portion of the volume is associated with grain boundary regions. For
instance, grain boundary area accounts for approximately 50% of the volume in a
material with 5 nm grain size. Also, NC materials exhibit anomalously improved
strengths heretofore unattainable with conventional materials, providing great potential
for structural materials in engineering applications. When the grain size of Al alloys is
reduced down to nanometer range of ~10 nm, yield strength increases up to three times
that of the strongest conventional Al alloys [6]. Therefore, the deformation mechanisms
must be substantially different in a nanoscale regime, while dislocation movements
predominantly govern the deformation in conventional coarse-grained materials.
The synthesis, characterization and processing of such NC materials are part of an
emerging and rapidly growing field referred to as nanotechnology. A critical research and
development issue in the field of NC materials emphasizes research aimed at the
fabrication of bulk nanostructured materials (BNMs) with engineered properties and
technological functions based on scientific discoveries, ultimately aimed at the generation
2
of materials with controlled microstructural characteristics. A wide variety of techniques
have been proposed to synthesize nanostructured materials [1 −5].
Cryomilling is one of several techniques to synthesize NC structures, and the
merits of cryomilling include (i) cryogenic temperature, which minimizes heat generation
during milling so that recovery and recrystallization are suppressed, (ii) rapid grain
refinement and capability for production of large batches of refined powder, (iii) reduced
oxygen contamination from the atmosphere, and (iv) high thermal stability against grain
growth because of large nitrogen content as a result of milling in liquid nitrogen [7,8].
During milling, the powder particles are repeatedly sheared, fractured and cold-welded,
and severe plastic deformation effects the formation of nanostructures [7,8]. Generally,
cryomilling is aided by the addition of an organic processing control agent (PCA), often
stearic acid (C
18
H
36
O
2
), which is necessary to mediate cold welding of particles. However,
PCA can lead to the presence of hydrogen, which is detrimental to the ductility of the
final products. Therefore, the cryomilled powder must be hot-vacuum degassed to
remove the PCA including hydrogen and oxygen, although grain growth then becomes a
concern [9].
Properties of NC materials are dramatically changed by the presence of porosity
in the materials. Porosity-free BNMs are required for understanding the intrinsic
structures and properties of this new class of materials. However, most commonly used
approaches to obtain NC materials involve materials in the form of powder in some stage
of the processing, and it is still a formidable challenge to produce bulk compact samples
without losing the nanostructure by consolidation of powders with NC structure.
Therefore, the method used to consolidate a cryomilled powder is critical to the retention
3
of superior strength along with acceptable tensile ductility in the bulk product. As a
primary consolidation technique, hot isostatic pressing (HIP) has been preferred to
consolidate the degassed cryomilled powder, producing bulk materials with virtually full
density. However, the HIP often results in undesired grain growth due to extensive
diffusion in the high temperature process. Cold isostatic pressing (CIP) has been
investigated not only to avoid this problem, but also to reduce cost and time to achieve
high temperature. In terms of mechanical properties, despite such approaches, both as-
HIPped and as-CIPped materials typically exhibit low ductility due to the presence of
prior particle boundaries (PPBs) within the structure, which are deleterious to ductility
and toughness. Therefore, a further deformation step which introduces additional shear to
break up the PPBs is required, such as forging or extrusion.
The strain rate sensitivity (SRS) of solid materials has important implications in
understanding the mechanical behavior and deformation mechanisms of NC materials.
Some unusual deformation behavior observed in NC materials is often associated with
the SRS. Although many experiments have revealed effects of strain rates on deformation
behavior in conventional CG materials, there is limited knowledge on the strain sensitive
mechanical properties of NC/UFG materials.
BNMs often exhibit limited ductility and reduced toughness compared to
conventional materials. As an effort to improve ductility and toughness of NC Al alloys
in the present study, a bimodal microstructure was produced that consisted of regions of
NC and CG microstructures. The presence of CGs (micron- to submicron-sized grains)
within the cryomilled NC structure reportedly can enhance the ductility [10,11].
4
Some questions have arisen regarding the distribution of strain during plastic
deformation of materials with discrete regions of two different grains sizes. However,
accurate measurements of strain, especially micro-strain for the non-uniform bimodal
microstructure, have been unachievable using conventional measurement techniques.
Digital image correlation (DIC) is a relatively new technique used for non-contacting
displacement/strain field measurements, offering the ability to analyze non-uniform full-
field deformation with sub-pixel resolution in digital images and to visualize the actual
strain distributions within materials [12 −16]. The DIC technique has proven to be an
effective method for mapping displacements, and such a direct measurement method
lends itself to development and validation of numerical or analytical models.
The strength of cryomilled al alloys can be increased even further by the addition
of a ceramic reinforcement, such as a boron-carbide (B
4
C), leading to an ultimate tensile
strength of more than 1 GPa [17]. B
4
C particulates are blended with Al powder, and the
mixture is cryomilled, forming a composite powder with a good interfacial bonding
between the reinforcing phase and Al matrix. Also, a similar approach to the bimodal
microstructure has been made to improve the ductility and toughness by addition of CG
material, characterized by a hybrid trimodal nanocomposite. The development of
microstructures in the Al nanocomposite has been investigated in terms of the
microstructural evolution during cryomilling and the relationship between structure and
properties.

5
References
[1] H. Gleiter, Prog. Mater. Sci. 33 (1989) 223.

[2] R. Birringer, Mater. Sci. Eng. A 117 (1989) 33.

[3] C. Suryanarayana, Intl. Mat. R. 40 (1995) 41.

[4] K. Lu, Mater. Sci. Eng. R 16 (1996) 161.

[5] H. Gleiter, Acta Mater. 48 (2000) 1.

[6] A.L. Greer, Nature 368 (1994) 688.

[7] H.J. Fecht, Nanostruct. Mater. 6 (1995) 33.

[8] D.B. Witkin and E.J. Lavernia, Prog. Mater. Sci. 51 (2006) 1.

[9] B. Ahn, A.P. Newbery, E.J. Lavernia and S.R. Nutt, Mater. Sci. Eng. A 463 (2007)
61.

[10] V.L. Tellkamp, A. Melmed and E.J. Lavernia, Metall. Mater. Trans. 32A (2001)
2335.

[11] D. Witkin, Z. Lee, R. Rodriguez, S. Nutt and E. Lavernia, Scripta Mater. 49 (2003)
297.

[12] T.C. Chu, W.F. Ranson, M.A. Sutton and W.H. Peters, Exp. Mech. 25 (1985) 232.

[13] M.A. Sutton, M. Cheng, W.H. Peters, Y.J. Chao and S.R. McNeill, Image Vision
Comput. 4(3) (1986) 143.

[14] H.A. Bruck, S.R. McNeill, M.A. Sutton and W.H. Peters, Exp. Mech. 29 (1989)
261.

[15] B.W. Smith, X. Li and W. Tong, Exp. Tech. 22 (1989) 19.

[16] M.R. James, W.L. Morris and B.N. Cox, Exp. Mech. 30 (1990) 60.

[17] J. Ye, B.Q. Han, Z. Lee, B. Ahn, S.R. Nutt and J.M. Schoenung, Scripta Mater. 53
(2005) 481.

6
Chapter 2. Literature Review for Nanocrystalline Materials

2.1. Definition
The grain size of polycrystalline materials plays a significant, and often a
dominant, role in controlling many critical properties including the strength and ductility.
Generally, materials having small grain sizes obtain several advantages compared with
their coarse-grained (CG) counterparts because of their excellent physical, mechanical
and chemical properties. The concept of nanocrystalline (NC) materials with their
outstanding properties was reported in detail in an early review by Gleiter [1]. A
significant number of review articles have been published to date for the mechanical
behavior of NC materials and the important review articles on NC materials [1 −17] are
listed in Table 2.1. The scientific interest in NC metals has expanded rapidly over the last
twenty years with understanding of advantageous mechanical properties, as shown in
Table 2.1.
It is first necessary to define some basic terms in order to place the subject of NC
materials to the present study. As the name suggested, NC materials are defined as single
or multi-phase polycrystalline materials having nanoscale grain sizes of ~1.0 to 100 nm
[18]. Thus, the materials having submicrometer grain size are divided into either NC
materials having nanometer grains or the one called ultrafine-grained (UFG) materials
with upper limit of grain size regime where the grains are within the range of 100 −1000
nm. The structure of NC and UFG materials is characterized by fully homogeneous and
equiaxed microstructure with a majority of the grain boundaries (GBs) having high-angle

7
Table 2.1 Leading review articles on nanocrystalline materials.
Author Year Title
Gleiter [1] 1989 Nanocrystalline materials
Birringer [2] 1989 Nanocrystalline materials
Gleiter [3] 1992 Materials with ultrafine microstructure: retrospectives and prospective
Suryanarayana [4] 1995 Nanocrystalline materials: a critical review
Lu [5] 1996
Nanocrystalline metals crystallized from amorphous solids: nanocrystallization,
structure, and properties
Weertman et al. [6] 1999 Structure and mechanical behavior of bulk nanocrystalline materials
Suryanarayana and Koch [7] 2000 Nanocrystalline materials – current research and future directions
Valiev et al. [8] 2000 Bulk nanostructured materials from severe plastic deformation
Gleiter [9] 2000 Nanostructured materials: basic concepts and microstructure
Furukawa et al. [10] 2001 Processing of metals by equal-channel angular pressing
Mohamed and Li [11] 2001
Creep and superplasticity in nanocrystalline materials: current understanding and
future prospects
Weertman [12] 2002 Mechanical behavior of nanocrystalline metals
Kumar et al. [13] 2003 Mechanical behavior of nanocrystalline metals and alloys
Veprek et al. [14] 2005 Different approaches to superhard coatings and nanocomposites
Wolf et al. [15] 2005
Deformation of nanocrystalline materials by molecular-dynamics simulation:
relationship to experiments?
Meyers et al. [16] 2006 Mechanical properties of nanocrystalline materials
Valiev and Langdon [17] 2006 Principles of equal-channel angular pressing as a processing tool for grain refinement

8
misorientations and these submicrometer materials have advanced unique mechanical
properties compared with conventional CG polycrystalline materials having grain sizes of
> 10 µm [2,19,20].
In practice, the strength of all polycrystalline materials is related to the grain size,
d, through the Hall-Petch equation which explains that the yield stress, σ
y
, is given by:
2 1
o
σ σ
−
+ = d k
y
(2.1)
where σ
0
is the friction stress needed to move individual dislocations and k is the Hall-
Petch coefficient of material constant [21,22]. From eq. (2.1), it can be seen that the
strength of a material is enhanced by a decrease of grain size. Moreover, NC materials
may exhibit improved hardness and excellent ductility, enhanced diffusivity [23], higher
specific heat, enhanced thermal expansion coefficient, and superior soft magnetic
properties in comparison with conventional polycrystalline materials, and these have led
to an increasing interest in a fabricating materials with extremely small grain sizes.

2.2. Synthesis
2.2.1. Overview
Two basic strategic approaches to the synthesis for nanostructured materials have
been developed, which are generally termed the “bottom up” and “top down” procedures
[24]. Figure 2.1 shows a schematic illustration of the synthetic strategies of NC metallic
materials.
The first is the “bottom up” approach in which the bulk NC solids are fabricated
through the assembly of individual atoms or nano-particulate solids. Examples of this

9
Fig. 2.1 Schematic illustration of strategies for synthesis of nanostructured materials.

10
procedure include inert gas condensation [1,25,26], electrodeposition [27 −31] and
chemical and physical deposition [32].
Inert gas condensation is used for producing NC metals such as Cu, Ni and Pd.
The powder particles with grain sizes of ~5 −50 nm are produced by condensation from
the vapor phase and subsequently the particles are consolidated in a die at high pressures
with sometimes additional thermal energy.
Electrodeposition has been developed to produce NC sheet materials such as Cu
[30,33], Ni [27,34,35] and Ni-, Co- and Pb-base binary alloys [29]. Moreover, the process
is used successfully to introduce NC metallic coating on complex figures [35]. The sheets
are normally produced to have a thickness of ~100 µm or more and grain sizes of 20 −40
nm [35]. Although these approaches are reported to have the capability of producing
materials with exceptionally small grain sizes, they have disadvantages because the sizes
of the finished products are normally very small and there may be some contamination
introduced during processing and there is invariably at least a low level of residual
porosity.
Physical vapor deposition (PVD) is a variety of vacuum deposition and is a
general term used to describe any of a variety of methods to deposit thin films by the
condensation of a vaporized form of the material onto various surfaces. The coating
method involves purely physical processes such as high temperature vacuum evaporation
or plasma sputter bombardment rather than involving a chemical reaction at the surface to
be coated as in chemical vapor deposition (CVD). Chemical process of CVD produces
high-purity, high-performance solid materials. The process is normally used in the
semiconductor industry to produce thin films. In a typical CVD process, the wafer

11
substrate is exposed to one or more volatile precursors, which react and/or decompose on
the substrate surface to produce the desired deposit.
The second is the “top down” approach which avoids the introduction of either
contaminants or porosity by taking an existing bulk solid with relatively CG and then
processing it to refine the grain size to submicrometer range. This approach includes
severe plastic deformation (SPD) [8] and mechanical alloying (including ball milling and
cryomilling) with subsequent consolidation [36 −40].
A formal definition of SPD processing was introduced earlier [41] and a detailed
description was also presented in a recent review [9]. In practice, SPD processing uses an
extensive hydrostatic pressure to induce a very high strain to a bulk solid without
introducing any significant changes in the overall dimensions of the sample that leads to a
high density of lattice dislocations and consequent grain refinement. Some of the SPD
processing techniques such as equal-channel angular pressing (ECAP) [10,17,41], high
pressure torsion (HPT) [42 −44] and accumulative roll-bonding (ARB) [45,46] have been
well reported to date as the methods of refining microstructure and producing UFG bulk
materials through introducing high plastic straining. These experiments demonstrated that
UFG sizes in the submicrometer or even nanometer range are achieved by the SPD
processing.
The technique of ECAP, however, produces the UFG or NC microstructures of
materials with thermally unstable at elevated temperatures and it leads to grain growth by
microstructural recovery and recrystallization [47,48]. On the other hand, mechanical
alloying, and specific cryomilling through a medium of liquid nitrogen, has been
considered to produce NC materials in reasonably large sizes [49,50] and the produced

12
powders are cold/warm/hot consolidated to very close to full density without significant
grain growth. Several studies demonstrated the advantageous thermal stability of NC
powders introduced by the processes of mechanical alloying and milling [51,52]. The
method is successful especially for aluminum alloys because of the presence of fine
oxides/nitrides resulted from the milling process that leads to maintain the NC grains
during consolidation [38,39].

2.2.2. Cryomilling
2.2.2.1. Formation of nanostructure
A variety of ball millings, such as shaker mills, planetary mills or attrition mills,
etc., have been developed for different purpose as well as difference in geometry,
arrangement, capacity, and efficiency of milling. The main difference between various
types of mills is the milling intensity and the impact energy. Figure 2.2 schematically
shows an attrition mill, one of ball mills used in the present study. In the attrition mill, the
available range of powder capacity is relatively wide, from less than 100 g to several
hundred kg for industrial scales, while high-energy mills (e.g. shaker mills) is capable of
milling ~10 g of powder at a time. Figure 2.3 represents the development of powder
structure by ball-powder-ball collision during mechanical alloying [53]. During milling,
the powder particles are repeatedly flattened, cold welded, fractured and rewelded.
Cryomilling is a variant of mechanical attrition techniques in which powders are
milled in slurry, formed with milling balls and cryogenic medium, usually liquid nitrogen
[49]. An extremely low temperature and a relatively low milling energy are distinctive
features of cryomilling, compared with conventional high-energy ball milling

13
Fig. 2.2 Schematic diagram of an attrition mill used for mechanical alloying [53].

14
Fig. 2.3 Ball-powder-ball collision of powder mixture during mechanical alloying [53].

15
techniques [54]. Earlier investigations for cryomilling reported that shorter milling time
was required to obtain the smallest grain size [51]. Also, slower diffusion at cryogenic
temperature can allow increased extent of solid solubility and reduced time for
amorphization. Therefore, the benefits of milling at cryogenic temperatures include
accelerated grain refinement, reduced oxygen contamination from the atmosphere, and
minimized heat generated during milling.
A basic microstructural evolution for formation of nanostructure during
mechanical milling was described by Fecht [55]. Three stages of grain refinement process
proposed in the study are as follows: (i) localization of deformation into shear bands with
high dislocation density, (ii) the dislocations form sub-boundaries by aligning,
annihilating, and recombining with each other at a certain level of strain, and (iii) the sub-
GB structure transforms into randomly oriented high-angle GBs [55]. Figure 2.4 shows a
schematic depiction of the process at the microscopic level [56]. During milling at
cryogenic temperature, cryomilling, the annihilation of dislocations and the processes of
recovery and recrystallization are suppressed, and the refinement of grains down to
nanoscale can be achieved in less time. However, welding of powder particles is also
inhibited during cryomilling, so addition of a process control agent (PCA), typically an
organic compound, prevents excessive agglomeration and reduced yield. For softer
metals, ~0.2 wt.% stearic acid is generally required as PCA.

2.2.2.2. Thermal stability
Although many processing techniques have been developed to synthesize
nanocrystalline powders, significant grain growth has been observed during elevated

16
Fig. 2.4 Schematic representation of grain refinement mechanism in microscopic level
during ball milling [56].

17
temperatures in many NC materials. However, recent observations have revealed that
some NC metals or alloys exhibit inherent grain size stability at high temperatures,
reasonable level for most elevated temperature processes. The mechanism of grain
growth in NC materials is generally suggested to involve GB diffusion in terms of the
activation energy [57].
Recent studies have investigated mechanisms through which the thermal stability
of NC materials might be enhanced. For example, the pinning of NC GBs by a dispersion
of second-phase particles has been reported to be effective in suppressing thermally
activated grain growth [51,58]. The studies demonstrated that a dispersion of nanometer-
scale AlN or Al-oxynitride particles could be formed in Al powders during cryomilling,
and the presence of this stable, dispersed second phase was found to effectively stabilize
the Al grains against thermally activated growth of grains [51]. The formation of AlN is
the result of milling in the liquid nitrogen environment. Examples of microstructure of
the dispersoids in cryomilled Al-10Ti-2Cu alloy are shown in Fig. 2.5 as high-resolution
TEM (HRTEM) images [59]. A single platelet of AlN dispersoid is shown in Fig. 2.5(a),
and multiple dispersoids adjacent to GB is shown in Fig. 2.5(b).
Figure 2.6 shows the grain size variation of cryomilled Al with respect to annealing
temperatures for annealing times of 300 and 1800 s. There are two distinct stages of grain
growth observed with a transition at 723 K (0.78 T
m
) [60]. Grain growth was limited until
40−50 nm even after long anneals at temperatures as high as 0.78 T
m
, indicating a high
degree of grain size stability. However, substantial grain growth occurred as the
annealing temperature was increased above 723 K. In addition, as the annealing time

18
Fig. 2.5 High-resolution TEM images of AlN dispersoids: (a) a single dispersoid and (b)
multiple disporsoids adjacent to grain boundary [59].

19
Fig. 2.6 Grain size variation as a function of annealing temperature for two annealing
times [60].

20
increased, the grain growth behavior became more significant at a temperature range
above 723 K [60].

2.2.2.3. Consolidation
Cryomilled powder needs to be handled in an inert gas after colleted from the
attritor in order to prevent contaminations, forming oxides and hydroxides by reaction
with atmosphere. Prior to consolidation, the powder often requires being stored in a
sealed can after cryomilling. Due to the addition of the PCA resulting in the presence of
hydrogen, which is very detrimental to the ductility of the final product, the milled
powder needs to be hot-vacuum degassed. For large cans of powder, longer times and
higher temperatures are required during degassing process to remove the hydrogen,
although excessive grain growth has to be avoided [61]. Conventional consolidation
processes after degassing include: uniaxial cold pressing, vacuum hot pressing, cold
isostatic pressing (CIP), and most commonly, hot isostatic pressing (HIP). Subsequently,
secondary consolidation step, such as forging or extrusion, is required for applying
additional shear stress to break up the prior particle boundaries (PPBs) and to impart
ductility. Quasi-isostatic forging is a relatively newer technique to consolidate cryomilled
powder, introducing an additional shear stress which helps to break up PPBs [62,63].

2.3. Mechanical properties
The principles of mechanical properties in NC materials are reviewed in this
section. It is important to comprehend that mechanical properties of NC materials and

21
their deformation mechanism are not only dependent on the average grain sizes but also
significantly influenced by the distribution of grain size and the features of GBs.
In the first place, it should be noted that the existence of porosity in NC materials
may distort their properties. Especially, the reduction of properties appears in the material
processed by “bottom up” methods. Porosity of defects produced during processing is
understood to lead to the detriment of the properties to the NC materials. It is reported
earlier that lower values of the Young’s modulus were recorded in NC Pd and Cu metals
processed by inert gas condensation and compaction with increasing function of porosity
in the materials and the results are shown in Fig. 2.7 [26]. The results were concluded
that processing defects such as porosity may lead to stress concentrations in NC metals
and premature failure in tension. In practice, yield strength strongly influenced by the
volume of porosity were explained in NC Pd and Cu materials after inert gas
condensation with a noticeable decrease in the strength associated with a 1 −3% drop in
density [64]. Thus, the existence of porosities in NC materials provides initiation sites for
unexpected failure in early stage of deformation.

2.3.1. Yield strength
The eq. (2.1) indicates the grain size dependence of yield stress of polycrystalline
materials. In the early years, many of NC materials had lower density so that it was
difficult to separate the strengthening effect by the smaller grains and weakening effect
by the porosities. However recently, the quality of production improved and the NC
materials tend to have better characters and properties. For example, higher strength was
observed in NC Cu with a grain size of 25 −35 nm and Pd with a grain size of ~10 nm

22
Fig. 2.7 Young’s modulus as a function of porosity for NC Pd and Cu [26].

23
produced by inert gas condensation [65], and compared with the earlier experiments
using the materials with the same grain sizes [66,67]. However, Fougere at al. [64] also
reported that Cu and Pd were initially hardened upon annealing, which can lead to grain
growth, and then softened as the grains grow larger. The hardening could be attributed to
densification or changes in internal strains [65].
Recently, NC Cu was produced by a two-step ball milling at a combination of
liquid nitrogen temperature and room temperature and the grain size was measured as
23±10 nm [68]. Further experiments on the material showed very high 0.2% offset yield
stress of 791±12 MPa and 1120±29 MPa for ultimate tensile strength with a uniform
elongation of 14% [69]. Similarly, the same process of two-step ball milling produced
NC Cu with a grain size of 54 nm demonstrating 0.2% offset yield stress of 688 MPa [70]
which was higher than any other of the materials having the same grain size prepared in
other techniques and it is shown in Fig. 2.8 [26,64,68,70 −74].

2.3.2. Inverse Hall-Petch effect
From the Hall-Petch relationship given in eq. (2.1), in most of polycrystalline
materials having grain sizes of 100 nm or larger, an increase of yield stress is predicted
with the inverse of the square root of the grain size. However, as the microstructure is
refined to submicrometer regime, this relationship invariably breaks down and tends to
depart from that at larger grain sizes. Many experimental results have suggested that for
most metals the yield stress peaks at an average grain size of ~10 nm with further grain
refinement and Fig. 2.9 shows a schematic plot representing the variation of yield stress
as a function of grain size in conventional large grain, UFG and NC materials [13]. This

24
Fig. 2.8 Summary of experimental data from references for NC Cu [70]; A: hardness
divided by 3 [26], B: hardness divided by 3 [64], C: the solid like is the Hall-Petch
relation ship for the yield strength of Cu, σ
y
(MPa) = 25.5 + 3478.5 d
-1/2
[71], D: tensile
yield strength [70], E: tensile yield strength [72], F: compressive yield strength [64], G:
tensile yield strength [64], H: tensile yield strength [73], I: yield strength by miniaturized
disk bend test [68], and J: Compressive yield strength [74].

25
Fig. 2.9 Schematic plot of the variation of yield stress as a function of grain size from
microcrystalline to NC regimes [13].

26
inverse Hall-Petch effect is also seen in a summarized plot of experimental data for NC
Cu shown in Fig. 2.8. In the material, it is clear that the trend of yield stress is predicted a
plateau or a decrease as the grain sizes fall down to below ~20 nm (d
-1/2
≈ 0.22). Thus,
further decrease in grain size causes weakening of the polycrystalline materials.
Several theories have been proposed to date to explain the Hall-Petch breakdown
at nanograin sizes. A group of researchers inclined to consider a change of deformation
mechanism for nanostructured materials. The first report for this approach was discussed
by Chokshi et al. [66] suggesting Coble creep of NC materials at room temperature due
to the small grains and the high GB diffusivity. Armstrong et al. [75] discussed the
occurrence of GB weakening when the structure is non-equilibrium at nano-grained
materials. Schiotz et al. [76] and Schiotz and Jacobsen [77] proposed GB sliding from the
results of molecular dynamics (MD) simulations. Gryaznov et al. [78] considered GB
sliding due to porosity effect. Van Vliet et al. [79] observed there is GB migration when
the grain size is < 7 nm. Kumar et al. [13], however, mentioned that there is no sufficient
experimental evidence for the inverse Hall-Petch effect by GB sliding and Coble creep.
Meyers et al. [80] recently suggested a controlling mechanism by GB sliding since the
results from Nieman et al. [81] and Sanders et al. [82] showed the creep rates of NC Cu
and Pd which were several orders of magnitude slower than the expected Coble creep.
Another group considered the NC materials as a composite of a GB phase and a
grain interior phase. Gryaznov et al. [78] explained the GB phase is a thin inclusion
which has size dependent of yield stress. Takeuchi [83] suggested that GB phase may be
a softer phase and the thickness is independent from the grain size whereas Fu et al. [84]
reported that the GB phase is a stronger phase which has a thickness increasing with the

27
square root of gain size. Recently, Jiang and Weng [85] suggested the GB phase is an
amorphous phase having a constant thickness and pressure dependent yield stress.
However, there is no clear evidence and insufficient information on the nature of Hall-
Petch effect for small grain sizes less than ~10 −15 nm.

2.3.3. Ductility
The mechanical behavior of subgrain materials including NC and UFG materials
has been studied extensively in the last decade. Enhanced properties of hardness and
strength are reported with reduction of the grain size of materials to the nanoscale.
Improved ductility has been predicted by extrapolation of grain size in conventional
polycrystalline materials to the nanometer range. However, most of the limited number of
studies of tensile ductility in NC metals have revealed poor ductility, typically low
elongations of 1 −2% for NC metals with grain sizes of < 25 nm [86] whereas the same
material with coarse grains exhibit elongations of ~40 −60%. Koch [87] defined three
major factors of limited ductility in NC materials: artifacts from processing (e.g. porosity),
force instability in tension, and crack nucleation or shear instability. Thus, it is important
to process nanostructured samples with free from artifacts which may reduce primary
properties.
While there is a difficulty to examine the inherent mechanical properties in
materials without any artifacts, MD simulation has been developed as a valuable tool to
understand the deformation mechanism of NC materials [88 −93]. The results from the
simulations proposed several different plastic deformation mechanisms as a function of
grain size [94,95]: (a) a large grain size regime of d > 1 µm in which unit dislocations and

28
work hardening control plasticity and (b) smallest grain size regime of d < 10 nm where
there is no intergranular dislocation activity and GB shear is the dominant deformation
mechanism.
The mechanism for intermediate grain size of 10 nm < d < 1 µm is not well
understood but is believed to be due to dislocation motion generated by GB sources. This
regime is divided by the grain size of NC (d < 100 nm) and ultrafine grain (100 < d < 1
µm) and yield strength of the materials in both regimes are evaluated in terms of ductility
in tension and plotted in Fig. 2.10(a) and (b), respectively [96]. It is apparent from the
plots that there was a clear decrease in ductility as strength increases in NC materials
whereas UFG materials exhibit increased yield strength with holding excellent ductility.
It is also obvious from Fig. 2.10(a) that very low ductility of elongations of less than 5%
was observed in NC materials. Most of the data presented in Fig. 2.10(a) are for the
sample processed by “two-step” approaches such as inert gas condensation or mechanical
alloying. Although, some of the data shown in Fig. 2.10(a) exhibited elongations of > 5%
in materials processed by the inert gas condensation method, the samples prepared by
two-step processes exhibited poor ductility, probably because of artifacts during
processing [86,87]. On the contrary, as shown in Fig. 2.10(b), some experimental results
demonstrated excellent balanced mechanical properties of both high strength and ductility
in UFG Cu and Ti processed by SPD processing [97 −101]. Although the UFG materials
showed small grain size in the mean value, they have a wide range of grain size
distribution in many cases, which can be an important factor for the optimization of
mechanical properties [96].

29

Fig. 2.10 Comparison of yield strength versus elongation for (a) NC materials and
(b) UFG materials [96].

30
2.3.4. Strain hardening
Many NC materials have strengths up an order of magnitude higher than their CG
counterparts. These strengths are directly affected by the large population of GBs which
are imposed during processing and delay yielding by movement of dislocations inside the
crystalline lattice. However, theses NC and UFG materials cannot generally maintain
uniform tensile elongation. In conventional materials having large grains, the dislocation
density increases as imposed strain increases, causing “strain hardening”. But in NC
materials, since the grains are very small, dislocations cannot multiply in large numbers
and the materials are expected to have an intrinsically low ability to store dislocations,
and hence insignificant strain hardening. Several experiments showed strain hardening is
often absent in NC materials after an initial stage of rapid strain hardening for a small
plastic strain regime of ~1 −3% [102,103] and it is reported using MD simulation [88].
The saturation of dislocation density in a NC material is due to its dynamic
recovery or the annihilation of dislocation into GBs, and it leads to slow and absence of
strain hardening. Dynamic recovery usually occurs during SPD processes since rise in the
temperature during the processing induces microstructural recovery into UFG having
both low-angle and high-angle GBs [104 −107]. Work hardening may be observed only
during large strain is added.
Tensile property of a NC material shows a rapid peak and subsequent softening in
the stress-strain response. The absence of strain hardening leads to localization of
deformation and necking resulted in low ductility. In the experiment on NC Fe [108] and
Ti [106], flat curves were observed in the stress-strain response during compression.

31
Necking often occurs when plastic deformation is instable and forming of shear bands
observed in the consolidated Fe [108,109].

2.3.5. Strain rate sensitivity
There were several reports showing both increased and decreased strain rate
sensitivity with decreasing grain size in metallic materials. For example, NC Fe prepared
using ball milling and consolidation exhibited a lower value of strain rate sensitivity m ≈
0.006 at d ≈ 20 nm [110] whereas conventional Fe normally displays m value of ~0.04.
On the other hand, an opposite result was observed on UFG FCC metals of Cu, Ni and
Al-4Cu-0.5Zr [111]. The strain rate dependence was examined using strain rate range
from 0.001 to 4000 s
-1
and the results are shown in Fig. 2.11(a) −(c). The strain rate
sensitivity was measured as 0.015 for Cu, 0.006 for Ni and 0.005 for Al-4Cu-0.5Zr,
which were found to be significantly higher than that typical for annealed FCC
polycrystalline metals [111]. Higher m values are also measured in NC Ni [112], Au
[113] and electrodeposited Cu [114] whereas these observations were conducted in creep
tests where GB deformation is dominant.
A clear tendency of increased strain rate sensitivity was observed by Wei et al.
[115] where a grain size is below a critical value. This improved strain rate sensitivity
was observed in pulse electrodeposited Ni through nanoindentation hardness testing [116]
showing higher m value at d = 20 nm than at d > 60 nm. Moreover, although
conventional polycrystalline Al has an m value of ~0.007, an UFG Al showed a very high
value of m = 0.027 after the processing of ECAP [117].

32

Fig. 2.11 Stress-strain response of UFG materials at different strain rates: (a) Cu, (b) Ni
and (c) Al-4Cu-0.5Zr [111].

33
There were experiments for a HCP metal of Zn having UFG and NC processed by
ball milling [118]. The m values are ~0.15 for ball-milled Zn tested at 20 and 40 ºC and
~0.17 for Zn tested at 60 ºC, which are significantly higher compared with the m values
for normal FCC materials.

2.4. Deformation mechanisms
This section describes the fundamental deformation mechanisms leading to the
specific behaviors. From the previous section, it was summarized that NC materials
exhibit a high strength whereas there may be an inverse Hall-Petch phenomenon when
the grain size is around or less than 10 nm. Moreover, the ductility of the materials with
nanoscale grains tend to be reduced due to a low work hardening rate leading to strain
localization in early stage of deformation and to a reduced ability of the materials to
accommodate the progression of clacks by extensive plastic deformation.

2.4.1. Dislocation pile-up
The dislocation pile-up is the basic concept of the traditional explanation for the
Hall-Petch relationship. When the grain size is reduced, the number of dislocation pile-up
decreases with decrease of the number of GB in a grain where it is under a fixed stress
level. In other words, higher imposed stress is necessary to generate the same number of
dislocations for pile-up when the grain size becomes smaller and it is demonstrated in Fig.
2.12(a) [16]. The dislocation sources are assumed at the center of the grain and the
dislocation pile-ups with positive and negative directions are activated by a Frank-Read
source. However, this concept is no longer applicable to explain the plastic flow when the

34
Fig. 2.12 Braking up of dislocation pile-up in the (a) microcrystalline regime and
(b) NC regime [16].

35
grain size is reduced to a critical size. As it is well displayed in Fig. 2.12(b), the number
of dislocations at the pile-up is significantly reduced compared with that for a grain size
in the micrometer regime, as the grain size is reduced down to a nanometer regime of
critical grain size. Thus, the stress multiplication by pile-up is vanished. This mechanism
was first proposed by Pande et al. [119] and it is developed further later [120].

2.4.2. Grain boundary sliding
Grain boundary sliding is recognized as an important flow mechanism to achieve
superplasticity, especially during high temperature deformation [121,122]. Many reports
have been published to evaluate the significance of GB sliding theoretically [123,124]
and experimentally using superplastic Pb-62% Sn eutectic alloy [125] and superplastic
Zn-22% Al eutectic alloy [126 −128] where the grain size of these materials is normally in
the micrometer regime. Figure 2.13 is a schematic illustration of GB sliding model which
one layer of grains slides with respect to the other leading to a shear strain in the process
[16]. For NC materials, this process is proposed to be the dominant deformation
mechanism at grain sizes of < 50 nm.
Hahn et al. [129] proposed the hardness relationships in the dislocation-
dominated regime and in the GB sliding regime. Subsequently, employing the concept of
thermally-activated shear, Conras and Narayan [130] suggested that the macroscopic
shear rate is produced by the atomic shear events at the GB. Recently, Conrad analyzed
the data of flow stress in Cu on the effect of grain size at the range from millimeter to
nanometer [95] and the results identified three different regimes: Regime I (d > 10
-6
m)
denotes the dominant deformation mechanism controlled by dislocation pile-up, Regime

36
Fig. 2.13 Schematic illustration of grain boundary sliding model (a) initial position of
grains and (b) position after top layer has slid to right [16].

37
II (d ≈ 10
-8
−10
-6
m) denotes the main deformation mechanism of GB shear promoted by
the pile-up of dislocations, and Region III (d < 10
-8
m) denotes the dominant deformation
mechanism controlled by GB shear.

2.4.3. Grain boundary rotation and grain coalescence
It was pointed out earlier by Jia et al. [108] that there is a possibility that the
grains in nanometer regime rotate during plastic deformation and create a dynamic shear
zone by grain coalescence leading to large paths for dislocation movement. The
schematic illustration of this process is shown in Fig. 2.14, where the orientations of the
slip systems with highest Schmid factors are represented by a short line in each grain [16].
Neighboring grains may rotate in a way that brings their orientation close together during
plastic deformation, as shown in Fig. 2.14(a). This process eliminates a GB between
these grains (Fig. 2.14(b)) and provides a path for more extended dislocation motion (Fig.
2.14(c)). Since this mechanism allows localized softening in the material, it can explain
the inverse Hall-Petch relationship in NC materials.
Murayama et al. [131] observed a grain rotation motion generated by partial
disclination defects in a mechanically milled Fe. The observation using HRTEM showed
two wedge-shaped regions that form a partial disclination dipole. A disclination is a line
defect characterized by a rotation of the crystalline lattice around its line [132]. A
disclination dipole which consists of two disclinations causes rotation of crystal lattice
between them and these dipoles are energetically permitted only for disclinations that are
close to each other [133]. It is concluded in the experiment on a mechanically milled Fe
that the large stress files associated with partial wedge dislocations makes it difficult for

38
Fig. 2.14 Grain rotation and grain coalescence and during plastic deformation leading to
creation of elongated grains by annihilation of grain boundary [16].

39
other deformation defects to move through the metal leading to its higher strength [131].
However, it has been argued that the reorientation associated with the generation and
interaction of partial wedge disclinations assists in the grain break-down mechanism, thus
disclinations contribute both deformation and strengthening. It is suggested as an
alternative deformation mechanism of GB sliding.

2.4.4. Shear band formation
For the materials having submicron grains, it is known that the deformation mode
changes as the grain size decreases. The development of shear band was normally
observed immediately after the onset of plastic deformation for materials having grain
size of d < 300 nm. It is correlated to transition of behavior in strain hardening at those
grain sizes, since the work hardening is missing with increase of dislocation density. Wei
et al. [134] observed in polycrystalline iron under dynamic loading condition that the
deformation characteristics in an iron with d = 268 nm is relatively non-uniform whereas
deformation was significantly uniform without localized deformation in an iron having d
= 980 nm [134]. These results were consistent in both lower and higher strain rate tests.
The development of shear bands was discussed by Jia et al. in a mechanically
milled NC iron during quasi-static deformation [135]. The results demonstrated increase
of the number of shear bands with increasing strains from ~3 to 8%. In the material under
compression, it was observed that the process of shear banding involves: (i) the
nucleation of new bands, (ii) propagation along the shear plane, (iii) increase in strain
(flow) within the band, and (iv) increase in shear band widths which are at least 50 times
of the grain size at small strains.

40
2.4.5. Grain boundary dislocation creation and annihilation
Both experiments and MD simulations have shown to date that NC materials
cannot be characterized by grain size alone, and that many other structural parameters
exhibit an important role in the mechanical behavior. Computational atomistic modeling
has been very helpful in providing better insight into the structural and mechanical
properties of the NC metals, especially since the method provides direct access to many
of the hidden parameters in experiments, such as the structure of GBs, their stored excess
energy, the free volume in GBs and triple junctions, and the internal stress distribution.
Van Swygenhoven and coworkers observed using MD simulations that a lower
dislocation density after appreciable plastic deformation and development of a combined
GB source-sink model in NC materials [136 −140]. The mean free path of dislocations
generated at GB sources is significantly limited when the grain size is reduced to the NC
regime. These dislocations can move freely until they meet the opposing GB, which is
acting as a sink, rather than cross slipping and generating work hardening. It makes the
dislocation density lower in the material throughout the plastic deformation and thus
work hardening is not significant. The GB source-sink model is shown schematically in
Fig. 2.15 [16]. Dislocations generated at a GB run without any impediments until they
meet the opposing GB. The GBs will become virtually free of ledges when the grain size
decreases below 20 nm whereas the GB ledges are responsible for generating plastic flow
in the conventional polycrystalline regime. Intrinsic and extrinsic GB dislocations have to
be pushed out into the nanoscale grains. Moreover, since the mean free path of
dislocations is limited by the small grain size, dislocation cross slip and multiplication are
significantly prohibited when the grain size is in the NC regime.

41

Fig. 2.15 Grain boundary source-sink model [16].

42
References

[1] H. Gleiter, Prog. Mater. Sci. 33 (1989) 223.

[2] R. Birringer, Mater. Sci. Eng. A 117 (1989) 33.

[3] H. Gleiter, Nanostruct. Mater. 1 (1992) 1.

[4] C. Suryanarayana, Intl. Mat. R. 40 (1995) 41.

[5] K. Lu, Mater. Sci. Eng. R 16 (1996) 161.

[6] J.R. Weertman, D. Farkas, K. Hemker, H. Kung, M. Mayo, R. Mitra and H. Van
Swygenhoven, MRS Bull. 24 (1999) 44.

[7] C. Suryanarayana and C.C. Koch, Hyperfine Interact. 130 (2000) 5.

[8] R.Z. Valiev, R.K. Islamgaliev and I.V. Alexandrov, Prog. Mater. Sci. 45 (2000)
103.

[9] H. Gleiter, Acta Mater. 48 (2000) 1.

[10] M. Furukawa, Z. Horita, M. Nemoto and T.G. Langdon, J. Mater. Sci. 36 (2001)
2835.

[11] F.A. Mohamed and Y. Li, Mater. Sci. Eng. A 298 (2001) 1.

[12] J.R. Weertman, In: C. Koch, editor, Nanostructured materials: Processing,
Properties, and Applications, 2
nd
Edition, Norwich, NY: Noyes Publications (2007)
pp. 537−564.

[13] K.S. Kumar, H. Van Swygenhoven and S. Suresh, Acta Mater. 51 (2003) 5743.

[14] S. Veprek, M.G.J. Veprek-Heijiman, P. Karvankoa and J. Prochazka, Thin Solid
Films 476 (2005) 1.

[15] D. Wolf, V. Yamakov, S.R. Philipot, A. Mukherjee and H. Gleiter, Acta Mater. 53
(2005) 1.

[16] M.A. Meyers, A. Mishra and D.J. Benson, Prog. Mater. Sci. 51 (2006) 427.

[17] R.Z. Valiev and T.G. Langdon, Prog. Mater. Sci. 51 (2006) 881.

[18] R.W. Siegel, In: D. Fiorani and G. Sberveglieri, editors. Fundamental properties of
nanostructured materials, Singapore: World Scientific (1993) p. 3.

43
[19] E. Artz, Acta Mater. 46 (1998) 5611.

[20] R.Z. Valiev, Nature Mater. 3 (2004) 511.

[21] E.O. Hall, Proc. Phys. Soc. B64 (1951) 747.

[22] N.J. Petch, J. Iron Steel Inst. 174 (1953) 25.

[23] R. Wurschum, S. Herth and U. Brossmann, Adv. Eng. Mater. 5 (2003) 365.

[24] C. Suryanarayana and B. Prabhu, In: C.C. Koch, editor, Nanostructured materials:
Processing, Properties, and Applications, 2
nd
Edition, Norwich, NY: Noyes
Publications (2007) pp. 47 −90.

[25] P.G. Sanders, G.E. Fougere, L.J. Thompson, J.A. Eastman and J.R. Weertman,
Nanostruct. Mater. 8 (1997) 243.

[26] P.G. Sanders, J.A. Eastman and J.R. Weertman, Acta Mater. 45 (1997) 4019.

[27] A.M. El-Sharik, U. Erb, G. Palumbo and K.T. Aust, Scripta Metall. Mater. 27
(1992) 1185.

[28] U. Erb, A.M. El-Sherik, G. Palumbo and K.T. Aust, Nanostruct. Mater. 2 (1993)
383.

[29] U. Erb, Nanostruct. Mater. 6 (1995) 533.

[30] F. Ebrahimi, Q. Zhai and D. Kong, Scripta Mater. 39 (1998) 315.

[31] N.Y.C. Yang, T.J. Headley, J.J. Kelly and J.M. Hruby, Scripta Mater. 51 (2004)
761.

[32] C. Suryanarayana and C.C. Koch, In: C. Suryanarayana, editor, Non-equilibrium
processing of materials, Amsterdam, Netherlands: Pergamon (1999).

[33] L. Lu, S.X. Li and K. Lu, Scripta Mater. 45 (2001) 1163.

[34] C. Schuh, T.G. Nieh and T. Yamasaki, Scripta Mater. 46 (2002) 735.

[35] C. Schuh, T.G. Nieh and H. Iwasaki, Acta Mater. 51 (2003) 431.

[36] C.C. Koch and Y.S. Cho, Nanostruct. Mater. 1 (1992) 207.

[37] C.C. Koch, Nanostruct. Mater. 9 (1997) 13.

44
[38] V.L. Tellkamp, A. Melmed and E.J. Lavernia, Metall. Mater. Trans. 32A (2001)
2335.

[39] F. Zhou, X.Z. Liao, Y.T. Zhu, S. Dallek and E.J. Lavernia, Acta Mater. 51 (2003)
2777.

[40] X. Zhang, H. Wang, R.O. Scattergood, J. Narayan and C.C. Koch, Mater. Sci. Eng.
A 344 (2003) 175.

[41] R.Z. Valiev, Y. Estrin, Z. Horita, T.G. Langdon, M.J. Zehetbauer and Y.T. Zhu,
JOM 58 (2006) 33.

[42] A.P. Zhilyaev, B.-K. Kim, G.V. Nurislamova, M.D. Baró, J.A. Szpunar and T.G.
Langdon, Scripta Mater. 46 (2002) 575.

[43] G. Sakai, Z. Horita and T.G. Langdon, Mater. Sci .Eng. A 393 (2005) 344.

[44] A.P. Zhilyaev and T.G. Langdon, Prog. Mater. Sci. (2008) in press.

[45] S.V. Dobatkin, E.N. Bastarache, G. Sakai, T. Fujita, Z. Horita and T.G. Langdon,
Mater. Sci. Eng. A 408 (2005) 141.

[46] Y. Saito, N. Tsuji, H. Utsunomiya, T. Sakai and R.G. Hong, Scripta Mater. 39
(1998) 1221.

[47] J. Wang, Y. Iwahashi, Z. Horita, M. Furukawa, M. Nemoto, R.Z. Valiev and T.G.
Langdon, Acta Mater. 44 (1996) 2973.

[48] F.J. Humphreys, P.B. Prangnell, J.R. Bowen, A. Gholinia and C. Harris, Phil. Trans.
R Soc. Lond. A 357 (1999) 1663.

[49] D.B. Witkin and E.J. Lavernia, Prog. Mater. Sci. 51 (2006) 1.

[50] A.P. Newbery, S.R. Nutt and E.J. Lavernia, JOM 58 (2006) 56.

[51] M.J. Luton, C.S. Jayanth, M.M. Disko, S. Matras and J. Vallone, In: L.E.
McCandlish et al., editors, Multicomponent Ultrafine Microstructures, Pittsburgh,
PA: Mater. Res. Soc. (MRS) Symp. Proc. 132 (1989) pp. 79 −86.

[52] R.J. Perez, B. Huang and E.J. Lavernia, Nanostruct. Mater. 7 (1996) 565.

[53] P.S. Gilman and J.S. Benjamin, Ann. Rev. Mater. Sci. 13 (1983) 279.

[54] F. Zhou, D. Witkin, S.R. Nutt and E.J. Lavernia, Mater. Sci. Eng. A 375 −377
(2004) 917.

45
[55] H.J. Fecht, Nanostruct. Mater. 6 (1995) 33.

[56] Y. Xun, E.J. Lavernia and F.A. Mohamed, Metall. Mater. Trans. 35A (2004) 573.

[57] T.R. Malow and C.C. Koch, In: D.L. Bourell, editor, Synthesis and Processing of
Nanocrystalline Powder, Warrendale, PA: The Minerals, Metals & Materials
Society (TMS) (1996) p. 33.

[58] R.J. Perez, H.G. Jiang, C.P. Dogan and E.J. Lavernia, Metall. Mater. Trans. 29A
(1998) 2469.

[59] R.W. Hayes, P.B. Berbon and R.S. Mishra, Metall. Mater. Trans. 35A (1998) 3855.

[60] F. Zhou, J. Lee, S. Dallek and E.J. Lavernia, J. Mater. Res. 16 (2001) 3451.

[61] B. Ahn, A.P. Newbery, E.J. Lavernia and S.R. Nutt, Mater. Sci. Eng. A 463 (2007)
61.

[62] H.W. Chan, Mater. Design 9 (1988) 355.

[63] A.P. Newbery, B. Ahn, P. Pao, S.R. Nutt and E.J. Lavernia, Adv. Mater. Res.
29−30 (2007) 21.

[64] C.J. Youngdahl, P.G. Sanders, J.A. Eastman and J.R. Weertman, Scripta Mater. 37
(1997) 809.

[65] G.E. Fougere, J.R. Weertman, R.W. Siegel and S. Kim, Scripta Metall. Mater. 26
(1992) 1879.

[66] A.H. Chokshi, A. Rosen, J. Karsh and H. Gleiter, Scripta Metall. 23 (1989) 1679.

[67] G.W. Nieman, J.R. Weertman and R.W. Siegel, Scripta Metall. 23 (1989) 2913.

[68] K.M. Youssef, R.O. Scattergood, K.L. Murty and C.C. Koch, Appl. Phys. Lett. 85
(2004) 929.

[69] K.M. Youssef, R.O. Scattergood, K.L. Murty, J.A. Horton and C.C. Koch, Appl.
Phys. Lett. 87 (2005) 091904.

[70] S. Cheng, E. Ma, Y.M. Wang, L.J. Kecskes, K.M. Youssef, C.C. Koch, U.P.
Trociewitz and K. Han, Acta Mater. 53 (2005) 1521.

[71] M.A. Meyers and K.K. Chawia, Mechanical Metallurgy: Principles and
Applications, Englewood Cliffs, NJ: Prentice-Hall (1984) p. 496.

46
[72] Y.M. Wang, K. Wang, D. Pan, K. Lu, K.J. Hemker and E. Ma, Scripta Mater. 48
(2003) 1581.

[73] M. Legros, B.R. Elliott, M.N. Rittner, J.R. Weertman and K.J. Hemker, Philos.
Mag. A80 (2000) 1017.

[74] Y. Champion, S. Gue’rin-Mailly, J.L. Bonnentien and P. Langlois, Scripta Mater.
44 (2001) 1609.

[75] R.W. Armstrong, F.R.N. Nabarro and H. Conrad, In: I. Ovid’ko et al., editors,
Mechanical Properties of Nanostructured Materials and Nanocomposites,
Pittsburgh, PA: Mater. Res. Soc. (MRS) Symp. Proc. 791 (2003) 69.

[76] J. Schiøtz, F.D. Di Tolla and K.W. Jacobsen, Nature 391 (1998) 561.

[77] J. Schiøtz and K.W. Jacobsen, Science 301 (2004) 1357.

[78] V.G. Gryaznov, M.Y. Gutkin, A.E. Romanov and L.I. Trusov, J. Mater. Sci. 28
(1993) 4359.

[79] K.J. Van Vliet, S. Tsikata and S. Zurres, Appl. Phys. Lett. 83 (2003) 1441.

[80] M.A. Meyers, A. Mishra and D. Benson, JOM 58 (2006) 41.

[81] G.W. Nieman, J.R. Weertman and R.W. Siegel, Scripta Metall. Mater. 24 (1990)
145.

[82] P.G. Sanders, M. Rittner, E. Kiedaisch, J.R. Weertman, H. Kung and Y.C. Lu,
Nanostruct. Mater. 9 (1997) 433.

[83] S. Takeuchi, Scripta Mater. 44 (2001) 1483.

[84] H.H. Fu, D.J. Benson and M.A. Meyers, Acta Mater. 49 (2001) 2567.

[85] B. Jiang and G.J. Weng, J. Mech. Phys. Solids 52 (2003) 1125.

[86] C.C. Koch, D.G. Morris, K. Lu and A. Inoue, MRS Bull. 24 (1999) 54.

[87] C.C. Koch, J. Metastable. Nanocryst. Mater. 18 (2003) 9.

[88] H. Van Swygenhoven and A. Caro, Nanostruct. Mater. 9 (1997) 669.

[89] S.L. Frederiksen, K.W. Jacobsen and J. Schiøtz, Acta Mater, 52 (2004) 5019.

[90] S. Yip, Nature 391 (1998) 532.

47
[91] H. Van Swygenhoven, A. Caro and M. Spaczer, Acta Mater. 47 (1999) 3117.

[92] V. Yamakov, D. Wolf, M. Salazar, S.R. Phillpot and H. Gleiter, Acta Mater. 49
(2001) 2713.

[93] V. Yamakov, D. Wolf, S.R. Phillpot and A.K. Mukherjee, Nature Mater.1 (2001)
45.

[94] H. Conrad and J. Narayan, Acta Mater. 50 (2002) 5067.

[95] H. Conrad, Metall. Mater. Trans. 35A (2004) 2541.

[96] C.C. Koch, Scripta Mater. 49 (2003) 657.

[97] R.Z. Valiev, I.V. Alexandrov, Y.T. Zhu and T.C. Lowe, J. Mater. Res. 17 (2002) 5.

[98] Y.T. Zhu and T.G. Langdon, JOM 56 (2004) 58.

[99] Y.T. Zhu, Y.R. Kolobov, G.P. Grabovetskaya, V.V. Stolyarov, N.V. Girsova and
R.Z. Valiev, J. Mater. Res. 18 (2003) 1011.

[100] R.Z. Valiev, A.V. Sergueeva and A.K. Mukherjee, Scripta Mater. 49 (2003) 669.

[101] Y.T. Zhu and X. Liao, Nature Mater. 3 (2004) 351.

[102] R.S. Iyer, C.A. Frey, S.M.L. Sastry, B.E. Aller and W.E. Buhro, Mater. Sci. Eng. A
264 (1999) 210.

[103] J.E. Carsley, A. Fisher, W.W. Milligan and E.C. Aifantis, Metall. Mater. Trans.
29A (1998) 2261.

[104] T.R. Malow and C.C. Koch, Acta Mater. 46 (1998) 6459.

[105] R.Z. Valiev and I.V. Alexandrov, Nanostruct. Mater. 12 (1996) 35.

[106] D. Jia, Y.M. Wang, K.T. Ramesh, E. Ma, Y.T. Zhu and R.Z. Valiev, Appl. Phys.
Lett. 79 (2001) 611.

[107] Y.M. Wang, E. Ma and M.W. Chen, Appl. Phys. Lett. 80 (2002) 2395.

[108] D. Jia, K.T. Ramesh and E. Ma, Scripta Mater. 42 (2000) 73.

[109] Q. Wei, D. Jia, K.T. Ramesh and E. Ma, Appl. Phys. Lett. 81 (2002) 1240.

48
[110] T.R. Malow, C.C. Koch, P.Q. Miraglia and K.L. Murty, Mater. Sci. Eng. A 252
(1998) 36.

[111] G.T. Gray, T.C. Lowe, C.M. Cady, R.Z. Valiev and I.V. Aleksandrov, Nanostruct.
Mater. 9 (1997) 477.

[112] N. Wang, Z.R. Wang, K.T. Aust and U. Erb, Mater. Sci. Eng. A 237 (1997) 150.

[113] H. Tanimoto, S. Sakai and H. Mizubayashi, Nanostruct. Mater. 12 (1999) 751.

[114] Q.P. Kong, B. Cai, L. Lu and K. Lu, Def. Diff. Met. 188 (2001) 45.

[115] Q. Wei, S. Cheng, K.T. Ramesh and E. Ma, Mater. Sci. Eng. A 381 (2004) 71.

[116] J. May, H.W. Höppel and M. Göken, Mater. Sci. Forum 503 −504 (2005) 781.

[117] J. May, H.W. Höppel and M. Göken, Scripta Mater. 53 (2005) 189.

[118] X. Zhang, H. Wang, R.O. Scattergood, J. Narayan, C.C. Koch, A.V. Sergueeva and
A.K. Mukherjee, Acta Mater. 50 (2002) 4823.

[119] C.S. Pande, R.A. Masumura and R.W. Armstrong, Nanostruct. Mater. 2 (1993) 323.

[120] T.R. Smith, R.W. Armstrong, P.M. Hazzledine, R.A. Masumura and C.S. Pande,
In: M.A. Otooni et al., editors, Grain Size and Mechanical Properties:
Fundamentals and Applications, Pittsburgh, PA: Mater. Res. Soc. (MRS) Proc. 362
(1995) p. 31.

[121] R.C. Gifkins, Metall. Trans. A (1976) 1225.

[122] J. Pilling and N. Ridley, Superplasticity in Crystalline Solids, London, UK: The
Institute of Metals (1989).

[123] T.G. Langdon, J. Mater. Sci. 16 (1981) 2613.

[124] T.G. Langdon, Acta Metall. Mater. 42 (1994) 2437.

[125] R.B. Vastava and T.G. Langdon, Acta Metall. 27 (1979) 251.

[126] I.I. Novikov, V.K. Portnoy and T.E. Terentieva, Acta Metall. 25 (1977) 1139.

[127] P. Shariat, R.B. Vastava and T.G. Langdon, Acta Metall. 30 (1982) 285.

[128] Z.-R. Lin, A.H. Chokshi and T.G. Langdon, J. Mater. Sci. 23 (1998) 2712.

49
[129] H. Hahn, P. Mondal and K.A. Padmanabhan, Nanostruct. Mater. 9 (1997) 603.

[130] H. Conrad and J. Narayan, Scripta Mater. 42 (2000) 1025.

[131] M. Murayama, J.M. Howe, H. Hidaka and S. Takaki, Science 295 (2002) 2433.

[132] P. Klimanek, V. Klemm, A.E. Romanov and M. Seefeldt, Adv. Eng. Mater. 3
(2001) 877.

[133] I.A. Ovid'ko, Science 295 (2002) 2386.

[134] Q. Wei, L. Kecskes, T. Jiao, K.T. Hartwig, K.T. Ramesh and E. Ma, Acta Mater.
52 (2004) 1859.

[135] D. Jia, K.T. Ramesh and E. Ma, Acta Mater. 51 (2003) 3495.

[136] A. Hasnaoui, H. Van Swygenhoven and P.M. Derlet, Acta Mater. 50 (2002) 3927.

[137] P.M. Derlet and H. Van Swygenhoven, Scripta Mater. 47 (2002) 719.

[138] P.M. Derlet, A. Hasnaoui and H. Van Swygenhoven, Scripta Mater. 49 (2003) 629.

[139] A. Hasnaoui, P.M. Derlet and H. Van Swygenhoven, Acta Mater. 52 (2004) 2251.

[140] A. Froseth, H. Van Swygenhoven and P.M. Derlet, Acta Mater. 52 (2004) 2259.

50
Chapter 3. Novel Experimental Techniques

3.1. Nanoindentation
3.1.1. Motivation
Most of investigations about deformation mechanisms come from studies on bulk
materials characterized under a relatively simple state of stress such as uniaxial tension or
compression. Indentation tests are perhaps the most commonly applied means of testing
the mechanical properties of materials. However, small specimens are difficult to work
with conventional experiments used to investigate bulk materials whereas the growing
interest in observations of the mechanical properties in microscopic specimens such as
thin films and nano-wires. Nanoindentation testing offers the possibility to measure the
mechanical properties on a very small scale. Therefore, nanoindentation technique has
been increasingly relied to describe the mechanical properties of small specimens and the
technique might be a useful as a screening technique for characterizing many samples.

3.1.2. Equipment and basic method
The nanoindentation technique has been developed in the last two decades from a
simple method [1,2] to the one with high-resolution testing equipment to observe the
materials parameters, such as hardness, yield strength, strain hardening exponent and
Young’s modulus at the micrometer and nanometer scale. The main objective of
nanoindentation tests is to extract these mechanical properties of the specimen from load-
displacement measurement. Nanoindentation technique leads to the size of the residual
51
impression of only a few microns and it makes difficulties to measure the indentation size
using optical techniques. Therefore, the depth of penetration beneath of the indenter is
measured as the load is applied to the indenter so that it allows the size of the area of
contact to be determined.
Nanoindentation hardness testing is generally conducted using either spherical or
pyramidal indenters. Most common indenter tip, Berkovich indenter, was also used in the
present study and an image of Berkovich indenter tip and schematic parameters of the
indenter are shown in Fig. 3.1(a) and (b), where A is the projected area of contact, h
p
is
depth of penetration measured from the edge of the area of contact, and θ is a semi-angle
at the faces [3], respectively. The Berkovich indenter shown in Fig. 3.1(a) is generally
used in small scale indentation studies and has the advantage of easy construction of the
pyramidal edges to meet at a single point.
The method was developed to measure the mechanical parameters of hardness
and elastic modulus from indentation load-displacement data obtained during one cycle
of loading and unloading. A schematic plot of a typical load-displacement data set
obtained with Berkovich indentation is displayed in Fig. 3.2, where P designates the load
and h is the displacement relative to the initial undeformed surface [4]. From the load-
displacement data, there are three important quantities must be measured: the maximum
load, P
max
, the maximum displacement, h
max
, and the elastic unloading stiffness, S =
dP/dh, defined as the slope of the upper portion of the unload curve during the initial
stages of unloading (the constant stiffness). Additional important parameter is the final
depth, h
f
, which is the permanent depth of penetration after the indenter is fully unloaded.
52

Fig. 3.1 (a) SEM image of an indentation made by Berkovich indenter tip and (b)
indentation parameters of the Berkovich tip [3].
53

Fig. 3.2 Schematic plot of an indentation load-displacement curve showing important
parameters [2].
54
3.1.3 Indentation hardness and modulus
The exact processes to measure the hardness, H, and elastic modulus, E, are based
on the unloading procedures shown schematically in Fig. 3.3, where h
c
is the vertical
distance along which contact is made (the contact depth) and h
s
is the displacement of the
surface at the perimeter of the contact [1]. The basic assumption is that the periphery of
the indenter sinks in a manner that is described by models for indentation of a flat elastic
half-space by rigid punches of simple geometry. Then, supposing pile-up of material at
the contact periphery is negligible, the elastic models show that the amount of sink-in, h
s
,
is given as follows:
S
P
h
max
s
ε = (3.1)
where ε is a constant that depends on the geometry of the indenter. Using eq. (3.1) and
following the geometry of Fig. 3.3, the vertical displacement of the contact periphery, h
c
,
is given by:

S
P
h h
max
max c
ε − = (3.2)
When the contacting area, A, is described as a function of h
f
, the hardness of
specimen is estimated by:

A
P
H
max
= (3.3)
It should be noted that the estimated hardness value may deviate from the traditional
hardness value if there is significant elastic recovery during unloading whereas it is
important to consider only when the material has a very small value of E/H. The
measured unloading stiffness is then estimated by:
55
Fig. 3.3 Schematic representation of a section through an indentation showing various
quantities used in the analysis [1].
56
A E
dh
dP
S
eff
π
β
2
= = (3.4)
where β is dimensionless parameter, and E
eff
is the effective modulus defined by:

i
i
eff
E E E
2
2
1 1 1 ν ν −
+
−
= (3.5)
where E and ν are Young’s modulus and Poisson’s ratio, respectively, for the specimen
and E
i
and ν
i
are the same parameter for the indenter.

3.1.4. Load-displacement curves
There are possible variations on the basic load-unload cycle including partial
unloading during each loading increment, superimposing an oscillatory motion on the
loading, and holding the load steady at a maximum load and recording changes in depth.
These different types of testing give opportunities to measure the viscoelastic properties
of the material.
Practically, nanoindentation testing is performed using a variety of substances
from soft polymers to diamond-like carbon thin films. Therefore, the shape of the load-
displacement curve is displayed to be a rich source of information, not only for
measuring the mechanical parameters of the material, but also for the indentation of non-
linear events such as phase transformations, cracking and delamination of films.
Schematic example of commonly observed load-displacement curves for different
materials are shown in Fig. 3.4 [3]. It should be noted that in many cases the permanent
deformation or residual impression is not the result of plastic flow but may involve non-
linear event of cracking or phase changes within the specimen.
57

Fig. 3.4 Schematic examples of load-displacement curves for different material responses
and properties: (a) elastic solid, (b) brittle solid, (c) ductile solid, (d) crystalline solid, (e)
brittle solid with cracking during loading, and (f) polymer exhibiting creep [3].
58
3.1.5. Factors affecting nanoindentation data
Along with the miniaturization and micromation of materials, and with the
development of advanced materials, the mechanical properties of materials are measured
by nanoindentation techniques. But hardness measurements based on the conventional
method presented by Oliver and Pharr [4] showed various errors at measuring the small
indentation depth and some relate them to indentation size effects. There have been
several reports recording some of the factors causing errors in indentation test data
including adhesion due to the presence of adsorbed liquid on the surface of the specimen,
creep of the material being tested, surface roughness, tip radius of the indenter and the
methodology itself of the nanoindentation experiments [3]. Moreover, errors in testing
data are associated with the pile-up or sink-in of the material on the edges of the indent
during the indentation process.
When nanoindentation testing is employed, care must be exercised at small
indentation depth. As the indentation depth approaches zero, the method may yield
erroneous hardness and Young’s modulus mainly because the contact area approaches
zero and the testing may not provide accurate values for the contact depth or area. This
error may become a contributing factor to indentation size effect in nanometer scales.

3.2. Digital image correlation
The accurate measurements of displacement and strain during deformation of
advanced materials and devices have been an important issue in the field of experimental
mechanics. As one of the advanced experimental stress/strain analysis technologies,
image correlation technique is a newly developed technique, applied to many different
59
fields in mechanics recently [5 −11]. The image correlation technique was first used a few
decades ago focusing crack-opening displacements and fracture mechanics [5], and had
evolved into full-field displacement and strain measurement tool [8].
Digital image correlation (DIC) a non-contacting full-field strain measuring
technique using digital images for obtaining surface displacements and strains of objects
during deformation. The theory of correlation has been reported by many researchers
[6 −9]. The principle of this technique involves mathematical tracking of numerical
patterns between two digital images. Points on the un-deformed original surface can be
tracked to new positions on a deformed image using an error minimization technique.
The surface of materials must have unique patterns consisting of random gray level
patterns which can make a small area surrounding the points able to be tracked. For each
point of interest, a small subset (neighborhoods in digital pixels) of surrounding the point
is selected in an un-deformed reference image and the subset in a deformed image that is
the closest in gray level patterns to the original subset is sought based on linear elastic
displacement theory in conjunction with grey level intensity of points in the image.
In DIC technique, a set of neighboring points in the un-deformed state is assumed
to remain neighboring after deformation. Two-dimensional displacements of an object
are schematically illustrated in Fig. 3.5. The subset S (quadrangle in dot lines) represents
a selected region in the reference image, and the subset S
1
(quadrangle in solid lines) is
showing the region in the deformed image. For the point M in Fig. 3.5, the deformed
region (S
1
) can be correlated with the corresponding reference region (S) in order to
calculate in-plane displacements, u
m
and v
m
between points M and M
1
. If the selected
60
Fig. 3.5 Schematic concept of image correlation technique showing planar deformation
from the original state (S) to the deformed state (S
1
) [12].
61
region is small enough for the correlation, the coordinates of points in S
1
region (shown
in Fig. 3.5) can be approximated by first-order Taylor expansion as follows:
y
y
u
x
x
u
u x x
M
M
m m n
Δ
∂
∂
+ Δ
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
∂
∂
+ + + = 1
1
(3.6)
y
y
x
x
y y
M
M
m m n
Δ
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
∂
∂
+ + Δ
∂
∂
+ + =
υ υ
υ 1
1
(3.7)
Let f(x,y) and f
d
(x,y) be the gray level distributions of the reference and the
deformed images, respectively. For the subset S, a correlation coefficient C is defined as:
( ) ( ) []
()
∑
∑
∈
∈
−
=
S N
n n
n n d n n S N
y x f
y x f y x f
C
2
2
1 1
,
, ,
(3.8)
where (x
n
, y
n
) is a point in the subset S, and (x
n1
, y
n1
) is a corresponding point, defined by
eq. (3.6) and (3.7), in the subset S
1
. The correlation coefficient C will be zero when
parameters, u
m
and v
m
, are real displacements and when ∂u/ ∂x|
M
, ∂v/ ∂x|
M
, ∂u/ ∂y|
M
, and
∂v/ ∂y|
M
are the displacement derivatives of point M. Therefore, the best estimation of the
parameters can be made by minimization of the coefficient C [12].
Figure 3.6 shows a visualized example of actual strain distributions correlated by
DIC technique [13]. Two images were recorded during deformation at the beginning of
the deformation (Fig. 3.6(a)) and at the end of the deformation (Fig. 3.6(b)). The lower
contour images show the horizontal strain measure by two-dimensional DIC for the upper
images, respectively.
To conduct the DIC analysis in the present study, pairs of micrographs were
selected consisting of a reference image and a deformed state image recorded within a
range of both large strain and micro-strain. The resolution of digital images was
62
Fig. 3.6 DIC analysis showing horizontal strain of Al specimen during deformation
recorded at (a) the beginning and (b) the end of the deformation. A single camera was
used to measure the deformation. [13].
63
1392×1040 pixels, and the corresponding field of view was 1727×1290 µm at low
magnification for large strain and 434×324 µm at higher magnification for micro-strain
measurements. Two-dimensional DIC was performed using software (VIC-2D,
Correlated Solutions, Inc). For suitable spatial resolution and reliable results of DIC,
proper subset size and step size were selected. The pixel subset size should be selected
large enough to encompass a unique pattern for the area of interest, but also small enough
to distinguish strain differences in certain small regions. A large subset size would
increase overall accuracy of strain mapping, but could cause an averaging effect on the
strain field. The step size controls the density of the analyzed data. A larger step size
yields faster results but coarser data.

3.3. Micro-tensile test
A number of testing methods have been developed to understand mechanical
properties of materials. One of the most widespread methods is tensile test. Most of
conventional tensile tests provide information about only macroscopic properties.
However, demands for comprehensive knowledge of deformation behavior down to
microscopic scale have significantly increased in recent research paradigm. Meantime,
microscopy, both electron and light microcopy, has been a powerful tool to investigate
microstructure of materials. Micro-tensile test module in the present study allows
combination of two experimental approaches above, namely, microscopic analysis during
mechanical tests. The module is shown in Fig. 3.7 with descriptions for major
components. Experiments using this module can overcome uncertainties in interpreting
traditional stress/strain data performed ex-situ, and can reveal much more about the
64
Fig. 3.7 Micro-tensile test module.
65
deformation behaviors as it takes place, as opposed to a traditional post failure
examination. The module can perform most of principal mechanical tests, such as tensile,
compression and three/four-points bending tests, and can be attached with scanning
electron microscope (SEM) or optical microscope. It is not simply a scaled down version
of existing materials testing systems, but has been specifically designed to facilitate low
drift observation of the high stress region of samples.
The micro-testing module in the present study, as shown in Fig. 3.7, was designed
for low displacement rates (down to 300 nm/s), permitting micro-strain range
deformation for most material testing and covering cover a wide range of loads (2 −5,000
N). It was attached to a light microscope (Vanox AHMT3, Olympus), as shown in Fig.
3.8(a), and the dynamic deformation during the tensile test was sequentially captured
with time intervals of half a second using a digital camera (QICam, Qimaging) installed
with the microscope, as shown in Fig. 3.8(b). The dynamic observation system consisting
of the micro-test module and microscope in the present study was used with the DIC
technique mentioned previously to conduct strain field mapping during deformation.

3.4. Sample preparations for electron microscopy
3.4.1. Tripod polishing for TEM
Transmission electron microscope (TEM) has been well developed as an essential
tool for both fundamental and applied research and is one of the most powerful analytical
instrument providing microstructural details down to the atomic level. The preparation of
TEM samples, however, is the most time-consuming and labor intensive part of the
whole TEM analysis procedures. The requirements for TEM specimens are numerous,
66
Fig. 3.8 Experimental apparatus for dynamic observation of deformation, showing (a)
computer-controlled micro-tensile module attached to a light microscope and (b) close
view of the specimen on the module with an objective lens focusing on the surface of
specimen.
67
with the “ideal” specimen being uniformly thin, representative of the bulk specimen,
clean and free of contamination, easily handled, stable under the electron beam, self
supporting, and nonmagnetic.
Tripod polisher is a new mechanical preparation technique to generate relatively
wide area for electron transparency of specimens. This technique is based on similar
mechanical grinding and polishing mechanisms used with other techniques. The tripod
polisher is a small hand held fixture with three adjustable micrometers which are
independently controlled, as shown in Fig. 3.9. A specimen is attached to a glass insert
using an L-shaped bracket which contacts the specimen grinding surface. The specimen
is thinned in a wedge shape, shown in Fig. 3.10, using progressively finer diamond
lapping films until a smooth, highly polished specimen is produced with electron
transparent area at the front. The polishing process generally requires two steps. The first
step (first-side polishing) is to polish the specimen to the point of interest (POI) using a
progressive series of diamond lapping films. The POI is aligned to the rear two feet of the
tripod and determined where the polishing plane will be completed, as shown in Fig.
3.10(a). Once the POI is reached, the polishing plane should be parallel with the back feet.
The second step (second-side polishing) is defined as the “wedge technique” in which the
specimen is thinned at a pre-defined angle, creating a thicker wedge of material at the
rear of the specimen, providing mechanical support and stability, as shown in Fig. 3.10(b).
A TEM micrograph of Al-Ti-Cu alloy specimen prepared by the tripod polishing is
shown in Fig. 3.11(a) [14]. In Fig. 3.11(a), the field of view with electron transparency
obtained by the tripod polishing at a magnification of 3,000× was about 30 µm × 20 µm,
68
Fig. 3.9 Tripod polisher.
69
Fig. 3.10 Schematic of tripod polisher alignment for wedge shape production: (a) first-
side polishing and (b) second-side polishing. The point of interest is represented by star.
70
Fig. 3.11 (a) TEM and (b) SEM micrographs of bulk nanocrystalline Al-Ti-Cu alloy [14].
71
which is relatively wide in conventional TEM microstructures. For comparison, an SEM
micrograph recorded at similar magnification is shown in Fig. 3.11(b) [14].
The tripod polisher offers distinct advantages over traditional methods such as
electro-polishing and dimpling. These include precision cross sections of sub-micron
devices, repeatability, production of large electron transparent area, reduced ion milling
time, specimen stability, and combined plane view and cross section ability. The
technique has been widely used for a variety of materials including ceramics, composites,
semiconductors, and metals.

3.4.2. Cross section polishing for SEM
Scanning electron microscope (SEM) has been used for various applications in
many different fields of research and development. In SEM, there are many instances in
which observation of cross section of specimens is important as well as observation of
surface of specimens. Observation of cross sections provide plenty of materials
information, such as the size of crystal grains, structures and thickness of layers, or
existence of voids, etc., which can’t be revealed easily by observation of surfaces.
Mechanical abrasive polishing, one of the most conventional methods, has been
generally used to prepare specimens for cross-sectional specimens. However, it is
difficult to achieve good cross sections by the mechanical polishing method for the
following cases: (i) multi-component materials with different hardness, since the
polishing surface becomes uneven as the softer parts are cut faster and easier, (ii) soft
materials with voids, since the material around voids stretches and deforms, or debris get
72
stuck in the voids, and (iii) metals, due to distortion caused by mechanical polishing on
the polished surface, information about the crystal becomes difficult to obtain.
Focused ion beam (FIB) method is a relatively recent method to resolve the
problems above. FIB system has been widely used as a micro-machining tool, and the
most prominent feature in the FIB technique is sample preparations for electron
microscopy at any user-defined specific site of materials. However, it isn’t effective for
preparations of a wide surface area of specimens.
The cross section polishing (CP) method using Argon ion beam in the present
study does not incur the problem from mechanical polishing described above, providing a
good cross section of complex materials containing both soft and hard materials and
minimizing distortion of the polished surface and enabling to observe crystalline
channeling contrast clearly. In addition, it allows specimen preparation of much wider
area of specimens than the FIB technique.
Figure 3.12(a) shows the cross section SEM micrographs of a card edge connector
prepared by the CP method. For comparison, the cross section prepared by mechanical
polishing is shown in Fig. 3.12(b). The specimen consists of Au-NiP-Cu2 (second layer)-
Cu 1 (first layer) from the top. The layers are composed of soft metals (Au and Cu) and
hard metal (NiP). The specimen prepared by the CP method showed clear crystalline
contrast, as shown in Fig. 3.12(a). In contrast, the specimen produced by conventional
mechanical abrasive polishing in Fig. 3.12(b) fails to show good results due to polish
damage on the surface [15].
In the present study, cryomilled Al 5083 was used for microstructural analysis of
the cross section. For the sample preparation, the cryomilled powder was mixed with
73
Fig. 3.12 SEM micrographs of cross section of a multi-layered card edge connector
prepared by (a) the CP method and (b) the conventional mechanical polishing [15].
74
epoxy (G-1, Gatan Inc.), pressed, and cured to produce a pellet. Then, the pellet was
cross-sectioned using the CP method. Figure 3.13 shows the cross sections of the
cryomilled powder prepared by the CP method. The surface clearly reveals the grain
structure in cryomilled powder particles. In Fig. 3.13(a), both nanocrystalline and coarse-
grain regions are evident, along with voids and micro-voids (normally obscured in
conventionally polished sections). Figure 3.13(b) exhibits a global view showing oxides,
intermetallic phases, micro-voids, and the grain structure.

75

Fig. 3.13 SEM backscattered electron image of cryomilled Al powder cross-sectioned by
the CP method showing (a) pores between powder particles and (b) NC and CG
structures as well as submicron voids.
76
References
[1] W.C. Oliver and G.M. Pharr, J. Mater. Res. 7 (1992) 1564.

[2] G.M. Pharr, W.C. Oliver and F.R. Brotzen, J. Mater. Res. 7 (1992) 613.

[3] A.C. Ficher-Cripps, In: F.F. Ling, editor, Nanoindentation, Verlag, NY: Springer-
Verlag (2002) p.23.

[4] W.C. Oliver and G.M. Pharr, J. Mater. Res. 19 (2004) 3.

[5] W.H. Peters and W.F. Ranson, Opt. Eng. 21(3) (1982) 427.

[6] W.H. Peters, W.F. Ranson, M.A. Sutton, T.C. Chu and J. Anderson, Opt. Eng.
22(6) (1983) 738.

[7] T.C. Chu, W.F. Ranson, M.A. Sutton and W.H. Peters, Exp. Mech. 25 (1985) 232.

[8] M.A. Sutton, M. Cheng, W.H. Peters, Y.J. Chao and S.R. McNeill, Image Vision
Comput. 4(3) (1986) 143.

[9] H.A. Bruck, S.R. McNeill, M.A. Sutton and W.H. Peters, Exp. Mech. 29 (1989)
261.

[10] B.W. Smith, X. Li and W. Tong, Exp. Tech. 22 (1989) 19.

[11] M.R. James, W.L. Morris and B.N. Cox, Exp. Mech. 30 (1990) 60.

[12] Y.H. Huang, C. Quan, C.J. Tay and L.J. Chen, Opt. Eng. 44(8) (2005) 087011-1.

[13] Correlated Solutions, Inc., Columbia, SC, http://www.correlatedsolutions.com.

[14] Z. Lee, R. Rodriguez, R.W. Hayes, E.J. Lavernia and S.R. Nutt, Metall. Mater.
Trans. 34A (2003) 1473.

[15] M. Shibata, JEOL News, 39(1) (2004) 28.

77
Chapter 4. Nanocrystalline Al Alloys via Cryomilling

4.1. Degassing parameter effects
Al-Mg alloy (Al 5083) powder was ball-milled in liquid nitrogen to obtain a
nanocrystalline (NC) structure, then vacuum degassed to remove contaminants at
different temperatures. The degassed powders were consolidated by cold isostatic
pressing (CIP) and then forged to produce bulk, low-porosity material. The material
microstructure was analyzed at different stages using optical microscope, transmission
electron microscopy (TEM), and density measurements. The impurity concentration of
the final product was also measured. The forged material exhibited a bimodal grain size
distribution, consisting of both ultrafine grains (UFG) and coarse grains (CG). The
bimodal distribution was attributed to the presence of residual CG in the as-milled
powder. Higher degassing temperatures resulted in higher density values and lower
hydrogen content in the consolidated materials, although these materials also exhibited
more extensive grain growth. The microstructural evolution of NC Al 5083 was
investigated to determine the effects of degassing temperature.

4.1.1. Motivation
During cryomilling, the powder particles are repeatedly cold-welded and fractured.
The process is effective only when the welding and fracturing of particles are balanced.
Addition of a process control agent (PCA), typically an organic compound, mediates the
cold welding, thereby preventing excessive agglomeration and reduced yield [1].
78
However, the PCA can be a source of impurities, particularly hydrogen [2]. Therefore,
prior to consolidation, the cryomilled powder must be degassed at elevated temperatures
under vacuum in order to remove the PCA and other volatile contaminants. Dissolved
and trapped gaseous phases can result in porosity and degrade mechanical properties.
Thus, the degassing process is critical to achieve a high-quality final product.
A common process for primary consolidation of cryomilled powder is hot
isostatic pressing (HIP), which provides essentially full density. However, because of the
high temperatures involved, HIP can promote unwanted grain growth, thereby sacrificing
the desired nano-sized grains [3]. In this study, CIP was employed as an alternative
method of primary consolidation. CIPping typically results in a compact (green body)
with relatively low density compared to HIP, although CIPping benefits include cost and
time savings. In the present case, CIPped compacts were subsequently forged to complete
the densification, to break up prior particle boundaries, and to homogenize the material.
The purpose of this investigation was to assess the effect of powder degassing
temperature on the microstructure and impurity content of cryomilled, CIPped and forged
Al 5083.

4.1.2. Degassing details and characterization procedures
Gas-atomized -325 mesh Al 5083 (Al-4.4Mg-0.7Mn-0.15Cr wt.%) powder,
(produced by Valimet, Inc.) was cryomilled in a 20 kg batch (by DWA Aluminum
Composites) for 8 hours in a modified Szegvari attritor. Stainless steel balls were used
with a ball-to-powder weight ratio of 32 : 1. Prior to cryomilling, ~0.2 wt.% of stearic
acid was introduced into the milling chamber as a PCA. The cryomilled powder was
79
transferred to a glove box under a dry nitrogen atmosphere and subsequently loaded in
welded aluminum cans for storage.
For degassing, 1.2 kg portions of powder were decanted into a steel can with a
detachable lid. Degassing was performed at nominal temperatures of 400 ºC (low), 450
ºC (medium) and 500 ºC (high). Initial plans called for a 10-hour ramp time followed by
a 12-hour heat soak at the prescribed temperature. However, because of problems with
process control, slightly different hold temperatures and times were obtained. Throughout
degassing, the vacuum remained below 10
-4
Torr and, after cooling down to room
temperature, was typically ~10
-5
Torr.
The degassed powder was transferred to a rubber CIP bag inside a glovebox filled
with dry nitrogen, then consolidated by CIP at 310 MPa pressure. The resulting CIPped
billet was dry-machined to a cylindrical shape, 6.4 cm (diameter) × 14.7 cm (height),
exposed to atmosphere. A second degassing was then performed at 400 ºC for 1 hour on
the CIP billet in a different steel can. Inside a glove box, the CIP billet was removed and
fitted into a stainless steel forging can, which was then sealed by welding-on an end cap.
The canned CIP billet was heated at 510 ºC for 75 minutes, then forged at high-strain-rate
using a Dynapak press. The estimated initial strain rate was ~70 s
-1
. The forging process
was performed without a die, although die stops were employed to obtain 3.4 cm thick
disks. All degassing, CIPping and forging operations were performed by Pittsburgh
Materials Technology, Inc.
Microstructural characterization was performed on samples sectioned from the
central part of the consolidated billets. Cryomilled powders were prepared by mixing
with an epoxy (G-1, Gatan, Inc.) and curing at room temperature, then ground, polished,
80
dimpled and ion milled to perforation for TEM examination (with a Philips EM420). The
powder pellet was also cross-sectioned using a JEOL cross-section polisher (CP)
described in Chapter 3. The section was examined in an SEM (JEOL 7000F FEG) to
determine grain structure. TEM samples from bulk material were prepared by electrolytic
jet polishing. Grain size was measured directly from the TEM images using the linear
intercept method, measuring a total of 350 grains per sample. Grain size measurements
were performed on 6 −7 powder particles from each sample. Density measurements were
performed by immersing the solid sample in water and using the Archimedes’ method.
Hydrogen and oxygen concentrations of the forged materials were measured by inert gas
fusion methods (by LECO Corp. using their RH404 and TCH600 analyzers).

4.1.3. Experimental results and discussion
The cross-sectional microstructure clearly reveals the grain structure in
cryomilled powder particles, as shown in Fig. 4.1. Both NC and CG regions are evident,
along with nano-size voids, which normally obscured in conventional polished cross
sections. Figure 4.2 shows a typical TEM image of a cryomilled powder, and histograms
showing the grain size distributions of the cryomilled powders. The microstructure of the
as-cryomilled powders consisted mostly of NC grains, but CG regions were also present,
as shown in Fig. 4.2(a). The NC grains were generally randomly oriented, as revealed by
the uniform rings in the selected area diffraction (SAD) pattern shown in Fig. 4.2(b).
Some CG, as shown for example in Fig. 4.2(a), showed evidence of dislocation
activity. Although more than 70% of the grains were in the size range of 20 −60 nm, some
grains in the coarse regions were as large as 200 nm, as shown by the histogram in Fig.
81
Fig. 4.1 SEM backscattered electron image of cross section of as-cryomilled Al
powder showing both NC and relatively CG structures with nano-size voids.
82
Fig. 4.2 TEM microstructure of as-cryomilled powder: (a) a bright-field image
showing mostly nanocrystalline grains with arrows indicating the boundary with a
relatively coarse-grain region, (b) corresponding SAD pattern; (c) histogram of the
grain size distribution; (d) histogram of the grain aspect ratio distribution.
83
4.2(c). Overall, the number-based mean grain size of the as-cryomilled powder was 50.4
nm. The grains typically were not equiaxed, as shown by the histogram in Fig. 4.2(d),
giving an overall mean aspect ratio of 1.46. The presence of large grains and a non-
equiaxed structure suggests that, despite the use of cryogenic temperatures and a milling
time of 8 hours, the powder was incompletely milled. However, the observed bimodal
microstructure will probably enhance ductility of the final consolidated material, since
reports have shown that enhanced ductility and retention of high strength can be achieved
by the presence of larger grains within a NC structure [4 −7]. A bimodal microstructure
can be achieved by blending unmilled powder before consolidation, HIPping, or, as in
this case, arise from incompletely milled grains.
Powder samples degassed at three different temperatures were examined using
TEM. The grain size and aspect ratio were measured for each powder sample, and these
are summarized in Table 4.1. A doubling of grain size was observed in the cryomilled
powder after it had been degassed, as shown in Fig. 4.3, for the powder degassed at the
highest temperature (max. T = 500 ºC). In addition, increasing the maximum degassing
temperature, from 415 to 500 ºC, increased the mean grain size from 84 to 118 nm. There
was a minimal increase in the grain aspect ratio after the degassing and no significant
variation as a function of degassing temperature.
In the powder degassed at the highest temperature, CG regions were frequently
observed, being generally much larger than the CG regions observed in the cryomilled
powders. Figure 4.3 shows a bright field TEM image (with SADP) and a dark field TEM
image of a CG band (adjacent to a fine grain region), in which dislocation-free grains
appear to have formed by recrystallization. This observation is consistent with earlier
84

Table 4.1 Number-based mean grain size and aspect ratio of grains of the as-cryomilled
and degassed powders.

Powder
Degas condition
(ºC , hr)
Mean grain size
(nm)
Mean grain
aspect ratio
As-cryomilled – 50.4 ± 32.9 1.46 ± 0.27
Degassed (low temp.) 415, 5.5 84.0 ± 58.6 1.52 ± 0.39
Degassed (medium temp.) 440, 10 107.7 ± 74.3 1.54 ± 0.34
Degassed (high temp.) 500, 2 118.1 ± 80.2 1.54 ± 0.37

85
Fig. 4.3 TEM microstructures of the powder degassed at the highest temperature (max.
T = 500 ºC) showing neighboring CG and UFG regions; (a) bright field image with
corresponding SAD pattern and (b) dark field image.
86
work [8], in which elevated temperatures resulted in a restructuring of the grain
boundaries and grain growth processes.
Compared to the as-cryomilled powder above, degassing increased the mean grain
size, as illustrated by Fig. 4.4, which shows TEM bright field images and histograms of
the grain size distribution for the three powders degassed at 415, 440 and 500 ºC,
respectively. As the degassing temperature was increased, the mean grain size obtained
on a number basis increased from 84 to 118 nm. The regions of CG in the as-cryomilled
powder were again observed, with an increasing population of grains above 200 nm in
size as the degassing temperature was increased.
The structure of the CIPped powder degassed at the lowest temperature is shown
in Fig. 4.5(a). The powder exhibited low green density and contained large pores. As the
degassing temperature was increased, the inter-particle distance generally decreased after
CIPping, resulting in a higher density and smaller pores, as shown in Fig. 4.5(b) and (c).
Despite this, even in the billet CIPped with powder degassed at the highest temperature,
prior particle boundaries were observed, and the pores tended to be interconnected. Thus,
the as-CIPped material was still susceptible to atmospheric contamination and reactions
with moisture and oxygen, underscoring the need for a second degassing step.
Increasing the powder degassing temperature resulted in an increase in the density
of the CIPped billets. This is evident in Fig. 4.5 and in the density measurements given in
Table 4.2. However, using Archimedes’ method to measure density with water may
underestimate the presence of porosity, especially for materials containing interconnected
pores, since water tends to penetrate the material. Therefore, the true density of the as-
87
Fig. 4.4 TEM bright field images and graphs of the measured grain size distribution of
cryomilled powder degassed at different temperatures: (a) 415, (b) 440 and (c) 500 ºC.
88
Fig. 4.5 Optical micrographs of as-CIPped powder degassed previously at (a) low (415
ºC), (b) medium (440 ºC) and (c) high temperatures (500 ºC).
89

Table 4.2 Summary of material characteristics after forging.

Degas temperature
(ºC)
ρ
(%)
Mean grain size
(nm)
Mean grain
aspect ratio
415 (low) 98.4 (84.7) 186.5 ± 117.4 1.78 ± 0.50
440 (medium) 98.8 (86.0) 198.0 ± 134.2 1.76 ± 0.55
500 (high) 99.1 (87.9) 213.9 ± 141.5 1.80 ± 0.50

Note: ρ is the relative density determined using the standard density of Al 5083 (2.66 g/cm
3
),
and the values in parentheses indicate that of as-CIPped materials before forging.
90
CIPped material is expected to be lower than the value given in Table 4.2. Indeed, image
analysis of as-CIPped materials yielded a relative density of ~80%.
Forging of the CIPped material resulted in a further increase in density, to ~99%
theoretical density (Table 4.2). The forged material produced from the powder degassed
at the highest temperature yielded the highest density. Microstructural examination
showed that all of the material was now essentially pore-free and the prior-particle
structure, to a large extent, had been eliminated. Etching of polished sections produced
optically lighter-toned streaks, indicating the presence of relatively CG regions arranged
perpendicular to the forging direction, as shown in Fig. 4.6(a). These features were also
observed by TEM, which revealed elongated CG, up to about 1 µm in length, extending
perpendicular to the forging direction, as shown in Fig. 4.6(b). As stated previously, the
presence of micron-size grains within the NC cryomilled powder is expected to impart
greater ductility to the material [4−7].
The grain size distributions were determined from TEM images for the as-forged
materials made from powders degassed at different temperatures (415, 440 and 500 ºC),
as shown in Fig. 4.7 and summarized in Table 4.2 as well as grain aspect ratio. The mean
grain size of the forged material was typically ~200 nm, which constitutes a substantial
increase compared to the degassed powders shown in Table 4.1. However, there was a
relatively small apparent increase in grain size (from 186 to 213 nm) as the degassing
temperature was increased. In addition, the variation in grain size also increased as the
degassing temperature increased, as shown in Fig. 4.7 and Table 4.2. Both of these trends
are consistent with observations of the degassed powders. Forging results in a mean grain
91
Fig. 4.6 Microstructure of forged material made from CIPped powder degassed at the
highest temperature (max. T = 500 ºC): (a) optical (CG regions lighter-toned), and (b)
bright field TEM (showing a CG band perpendicular to the forging axis).
92
Fig. 4.7 Histograms of the measured grain size distribution of the as-forged materials
previously degassed at different temperatures: (a) 415, (b) 440 and (c) 500 ºC.
93
aspect ratio of ~1.8, which was greater than that of the as-degassed powders. However,
degassing temperature had no significant effect on the grain size aspect ratio.
Analysis of the impurity content revealed some important trends. The hydrogen
content of the forged material decreased from 31.3 to 24.0 ppm as the degassing
temperature was increased, as shown in Table 4.3. However, the hydrogen content was
still much higher than desirable, even when using the highest degassing temperature, and
this could diminish property levels, since hydrogen is known to cause deleterious effects
that include embrittlement or ductility reduction in high-strength materials [9].
Preliminary mechanical testing has shown that the forged material degassed at the highest
temperature had the greatest tensile elongation. However, this was accompanied by a
drop in strength, a phenomenon attributed to the increased grain size.
In contrast, the oxygen content of the forged material did not change as the
degassing temperature was increased. The content was steady at the relatively high level
of ~0.5 wt.%, as shown in Table 4.3. The high content of both hydrogen and oxygen in
the material examined in this study suggests that contamination occurred, either during
handling of the cryomilled powder in the glove box with inadequate atmosphere control,
or during machining of the CIPped billet. The constant oxygen levels indicate that it is
not removed by degassing. This is because it is present as an oxide or a hydroxide (which
also introduces hydrogen), neither of which is broken down by the degassing conditions
employed. Although a small amount of oxygen is introduced into the cryomilled powder
by the use of stearic acid, most of the oxygen results from the oxide films on the pre-
milled gas-atomized powder and reaction of the cryomilled powder with atmospheric
moisture and/or oxygen. The reactions that occur after cryomilling are likely to be more
94

Table 4.3 Impurity contents of the forged materials.

Degas temperature
(ºC)
Hydrogen (H) content
(ppm)
Oxygen (O) content
(wt.%)
415 (low) 31.3 0.50
440 (medium) 27.2 0.48
500 (high) 24.0 0.51

95
detrimental, because the reaction products will exist on the powder particle surface. In
contrast, the oxide formed during gas-atomization gets extensively broken up and
incorporated into the matrix of the cryomilled particles.

4.1.4. Conclusions
Cryomilled Al 5083 powder, with a mean grain size of 50 nm, was hot vacuum
degassed at different elevated temperatures (0.78 −0.89 T
m
) prior to consolidation.
Significant grain growth occurred during degassing, and the mean grain size increased as
the degassing temperature was increased. Extensive grain growth also occurred after the
combined effects of CIPping, a short second degassing and forging, leading to a mean
grain size of ~200 nm. The grain size of the forged material also increased as the
degassing temperature was increased. The final consolidated material had a non-uniform
microstructure with a bimodal grain size distribution. This bimodal gain distribution is
attributed largely to incompletely cryomilled powder particles. The hydrogen content of
the final material decreased as the degassing temperature was increased, while the
oxygen content remained constant.
The CIP/forge route is being compared to alternative consolidation methods.
Although CIPping affords saving in cost and time, the presence of interconnected
porosity in the as-CIPped condition renders the compacted powder susceptible to
atmospheric contamination. From the results presented in this study, a higher degassing
temperature will help reduce CIP porosity. Alternative CIPping conditions (e.g. higher
pressures) are being investigated to increase the density of the CIPped material.

96
4.2. Consolidation and forming method effects
Al 5083 powder was cryogenically ball-milled, to obtain a nanocrystalline (NC)
structure, hot vacuum degassed, and then consolidated by one of two routes: (1) hot
isostatic pressing (HIP) and extrusion, and (2) two-step quasi-isostatic (QI) forging. The
consolidated billet in both cases was hot rolled to make plate approximately 19 mm thick.
Despite grain growth during consolidation and deformation processing, a similar ultra-
fine grain structure was obtained in both plates. Both rolled materials had a similar tensile
strength in the plane of the plate, much greater than conventionally processed ‘armor
grade’ Al 5083-H131. However, the QI forged plate had significantly higher fracture
toughness, the toughness for the HIPed and extruded plate being particularly low for
crack surfaces in the plane of the plate. This was attributed to differences in the prior
powder particle boundary structure arising from the two consolidation methods. During
QI forging, the uniaxial pressure was transformed into a QI pressure field by the
application of a granular pressure transmitting medium (PTM), and this introduced an
additional shear stress which helped to break up prior particle boundaries (PPBs) in the
bimodal alloys.

4.2.1. Motivation
In order to retain acceptable ductility in the consolidated alloys, the cryomilled
powder must undergo shear deformation during the consolidation process to break up and
disperse PPBs. A combination of HIP and extrusion has often been used to produce
cryomilled rod or bar of high strength with greater ductility than as-HIPped [10]. Using
an extrusion die of the appropriate rectangular shape, HIP/extrusion reportedly can be
97
used to produce plate stock from cryomilled Al alloys if the extrusion is subsequently
rolled. However, to produce a more convenient pre-form for rolling, forging is more
suitable. Recently, QI forging, formerly known as Ceracon forging, has been used to
produce cryomilled Al 5083 plate with enhanced mechanical properties [11]. During QI
forging, the material is surrounded by a particulate PTM within a closed die. The forging
process can also be used to consolidate the canned powder, replacing the need for
HIPping as the primary step [12]. The objective of this study is to compare and evaluate
two powder consolidation processes – (1) HIP and extrusion and (2) two-step QI forging
– for producing large plates ( ∼19 mm × 450 mm × 250 mm) from cryomilled Al 5083,
with respect to microstructure and mechanical properties, especially fracture toughness.

4.2.2. Materials and processing details
Two 20 kg batches of gas-atomized -325 mesh (< 44 µm) Al 5083 (Al-4.4Mg-
0.7Mn-0.15Cr wt.%) powder produced by Valimet, Inc. (Stockton, CA) was cryomilled
in liquid N
2
by DWA Aluminum Composites (Canoga Park, CA) for 8 hours in a
modified Szegvari attritor. Stainless steel milling balls were used with a ball-to-powder
weight ratio of 32 : 1, and 40 g (0.2 wt.%) stearic acid was added to improve yield. After
milling, the powder was transferred to a glove box under liquid N
2
, ensuring that
atmospheric contamination of the cryomilled powder was minimized. The powder was
vacuum (< 10
-6
Torr) degassed by heating each can to 450 ºC.
A 295 mm diameter, 165 mm high can containing 14.0 kg cyromilled powder was
HIPped by Kittyhawk, Inc. (Garden Grove, CA) at a maximum pressure and temperature
of 103±1 MPa and 396±4 ºC for 4 hours. The can was machined to leave a cylindrical
98
billet of consolidated material, 223 mm diameter and 105 mm height. Two cylindrical Al
6061 sections of the same diameter, the header and the follower, were fitted along with
the cryomilled billet into a thin-walled Cu jacket, 231 mm (outer diameter), to give a 45
kg slug of total length ~360 mm. The assembled slug was extruded through a 127 mm ×
51 mm rectangular cross section die by HC Starck (Coldwater, MI), using a billet pre-
heat temperature of 204 ºC. After removing the Cu and Al 6061, a 430 mm length of
extruded cryomilled Al 5083, weighing 6.8 kg, 48% of the powder used, was obtained.
A 150 mm diameter, 255 mm high can containing 5.6 kg of cryomilled powder
was QI forged by Advance Materials & Manufacturing Technologies, LLC (Roseville,
CA) employing a 330 mm diameter die and a die pressure of 260 MPa. The can, pre-
heated to 454 ºC, was placed in the die within a bed of graphitic particulate, pressure
transmitting medium (PTM), pre-heated to 500 ºC. After a first forging, the can was
machined to obtain a cylindrical billet in dimensions of 127 mm diameter and 133 mm
height. This billet was then reheated to a temperature of 407 ºC and forged for a second
time using the same die and load as given above. After the second forging, and milling of
the top and bottom surfaces in preparation for rolling, the material formed a disc, ~223
mm diameter and thickness of 46 mm. The weight of the billet was 4.3 kg, corresponding
to 78% of the cryomilled powder used.
Uniaxial rolling, of both HIPped and extruded and QI billets, pre-heated to 450 ºC,
was carried out by Niagara Specialty Metals (Akron, NY) achieving a final transverse (T)
thickness of 19.3±0.5 mm. The HIPped and extruded plate, rolled normal to the extrusion
axis, was approximately rectangular, with dimensions of ~450×295 mm, in the extrusion
(E) and rolling (R) directions, respectively. The QI forged plate was oval, with
99
maximum dimensions of ~454×229 mm, the shorter dimension being normal (N) to the
rolling direction (R).
For comparisons with the consolidated plates after cryomilling, a rectangular
section of conventionally processed ‘armor grade’ Al 5083 plate, 19.0 mm transverse (T)
thickness, was used. The exact processing history of this plate was unknown, although it
was assumed that it was made by standard ingot metallurgy (IM) route, followed by hot
rolling, and then cold working to achieve a H131 condition. It was also assumed that this
plate had primary (R1) and secondary/normal (R2) rolling directions parallel to its length
(L) and width (W), respectively.

4.2.3. Mechanical tests and characterization procedures
Tensile properties in two normal directions within the plane of the cryomilled
plates were obtained using flat dog-bone specimens approximating to sub-size ASTM
E8M (with gauge section of 40×6×3 mm). The specimens were strained at a rate of 10
-3
s
-
1
using an Instron 8801 universal testing machine, with an attached dual-camera video
extensometer. Fracture toughness was measured using a mini-compact tension (CT)
specimen of 5.1 mm thick and 12.7 mm width. Side grooves with depths equal to 5% of
the thickness were introduced on both sides of each CT specimen to enhance crack
constraint. CT specimens were oriented to give four combinations of crack plane and
propagation direction with respect to the extrusion, forging or rolling directions for each
plate made from cryomilled powder. For the standard plate, CT specimens were also
made to give two orientations of crack plane and propagation. The fracture toughness
tests were performed using an ATS 904 Universal Test Machine and conducted in accord
100
with ASTM E399 in ambient air (20 °C and 42% relative humidity). In practice, because
of the high fracture toughness, the plain strain conditions required for a valid K
Ic

measurement were not met, and so the values obtained were designated to be K
q
, where
K
q
is greater than K
Ic
. After testing, fracture surfaces were examined using SEM (Zeiss
Leo 1550), the fractographs presented with the crack growth direction horizontal on the
page.
The concentration of metallic alloying elements was measured using DC plasma
emission spectroscopy according to ASTM E 1097-03 by Luvak, Inc. (Boylston, MA).
The concentration of non-metallic elements were measured by LECO, Inc. (St. Joseph,
MI) using inert gas fusion with their RH404 and TCH600 analyzers for H and for O & N,
respectively, and combustion combined with IR detection with their CS600 analyzer for
C. Density measurements were performed using a Micromeritics AccuPyc-1330 Ar gas
displacement pycnometer with a 3.5 cm
3
chamber, at least 80% of which was filled by
the samples. Four sets of 20 measurements were made for each material, achieving an
accuracy of close to ±0.001 g/cm
3
.
Standard metallographic techniques were performed on the sections for
examination in an optical microscope (Olympus Vanox AHMT3). Polished samples were
etched with detergent to emphasize the CG regions, which appeared a lighter tone to the
darker UFG matrix. Slices of the materials were thinned for TEM examination (Philips
EM420) by jet polishing in an ethanol containing 8% perchloric acid + 10% 2-
butoxyethanol. The maximum dimensions of 400 individual grains were measured from
TEM images, generating the mean grain size, number based grain size histogram, and
aspect ratio.
101
4.2.4. Experimental results and discussion
The concentration of metallic elements for both the plates made from cryomilled
powder were within the specification for Al 5083, as shown in Table 4.4. However, as
expected, the levels of the non-metallic elements, H, O, N and C, were all substantially
higher than plate produced by standard processing methods, as also shown by Table 4.4.
Although the two cryomilled plates were produced from different batches of powder, the
compositions were similar. The QI-forged plate was slightly lower in Mn and C.
The central part of the HIPped and extruded plate had a measured density of
2.642 g/cm
3
, 99.3% of the reference value for Al 5083 (2.66 g/cm
3
), indicating that it was
not fully dense. In addition, the density varied with respect to position within the plate,
between 2.639 and 2.650 g/cm
3
. This was confirmed by optical microscopy, which
revealed the presence of pores as large as 10 µm, as shown in Fig. 4.8(a). Many of these
pores were created by the rolling, since the as-extruded billet had a higher density,
measured to be 2.659 g/cm
3
. In contrast, the QI forged plate had a density of 2.670 g/cm
3
,
higher than the reference value for Al 5083 and 0.5% greater than the standard Al 5083-
H131 plate used for comparison in the present study (2.656 g/cm
3
). This high density is
attributed to the incorporation of oxides and nitrides into the structure due to gas
atomization and cryomilling. The QI forged plate also exhibited very little variation in
density with respect to position, all samples measuring between 2.670 and 2.671 g/cm
3
.
Microscopic examination revealed a distribution of small pores, mostly < 1 µm, as shown
in Fig. 4.8(b).
For the HIP/extruded plate, TEM confirmed that the grains, also elongated in the
extrusion axis, had a wide range of sizes, as shown in Fig. 4.9. Although most grains

102

Table 4.4 Chemical analysis of rolled Al 5083 plates. Values in wt.%, except H (ppm).

Al 5083 Processing Al Mg Mn Fe Cr H O N C
HIP & Extrusion 93.9 4.63 0.68 0.26 0.08 11.2 0.47 0.43 0.18
Cryomilled
QI Forging 94.2 4.60 0.47 0.21 0.06 10.2 0.48 0.44 0.14
Standard H131 92.4-95.6*4.0-4.9* 0.4-1.0* 0.4 max.* 0.05-0.25*1.3 0.003 <0.0050.001

* Specification for Al 5083: concentration of alloying elements in standard plate was not measured.

103
Fig. 4.8 Optical micrographs (unetched) of cryomilled Al 5083 plate showing pores
(black) viewed normal to the rolling direction: (a) HIP/extruded and (b) QI forged.

104
Fig. 4.9 TEM micrographs, with grain size distributions, of HIP/extruded cryomilled
Al 5083 plate viewed in three directions with respect to processing: (a) rolling, (b)
extrusion, and (c) transverse axes.

105
were less than 400 nm, a significant population of grains larger than 1 µm w ∼ as present,
as also shown in Fig. 4.9. The mean grain size was approximately 440 nm × 350 nm ×
220 nm, in the extrusion, rolling and transverse directions, respectively, a ratio of
approximately 2 : 1.6 : 1, as shown in Table 4.5. However, for the QI forged plate, TEM
images revealed that the proportion of grains larger than 1 µm was slightly less than for
the HIP/extruded plate, as also shown by Fig. 4.10. In general, the grain structure was
similar to the HIP/extruded plate, although the mean grain size was larger, as shown in
Table 4.5. The mean grain dimensions were approximately 490 nm × 430 nm × 240 nm
in the rolling, normal and forging (transverse) directions, respectively, a ratio of
approximately 2 : 1.8 : 1. The grains in the standard Al 5083 plate were larger by several
orders of magnitude than the cryomilled plates, and they were more elongated, as shown
in Fig. 4.11 and Table 4.5. The mean grain length parallel to the rolling axis was 220 ∼
µm, almost 500 times that of the cryomilled plates, and the ratio of the mean grain
dimension in the primary rolling direction (R1) to that of the transverse direction (T) was
6.5 : 1, over three times than that of both the cryomilled plates. The presence of coarser,
micron-sized grains within the UFG structure originating from cryomilling can be
expected to enhance ductility and fracture toughness [13,14]. For the HIPped material,
generation of CG arises from filling of the interstices between the cryomilled particles by
diffusion [15]. The QI forged plate still contained a number of grains above 500 nm and
these probably originated from variations in grain size within the as-milled powder,
which have been observed to be as large as 200 nm [12]. Although the HIP/extruded plate
had a larger population of micron-sized grains, they did not impart greater ductility,

106

Table 4.5 Mean grain size and aspect ratio of rolled Al 5083 plates.

Al 5083 Processing Viewpoint Measurement method
Length
(nm)
Width
(nm)
Aspect
ratio
Rolling (R) 440 (E) 239 (T) 1.84
Extrusion (E) 362 (R) 201 (T) 1.80 HIP & Extrusion
Transverse (T) 447 (E) 346 (R) 1.3
Normal (N) 480 (R) 248 (F) 1.9
Rolling (R) 438 (N) 240 (F) 1.8
Cryomilled
QI Forging
Forging (F)
Individual grains
from TEM images
496 (R) 428 (N) 1.2
Width (R2) 228 µm (R1) 35 µm (T) 6.5
Standard H131
Transverse (T)
Linear intercept
from optical images
211 µm (R1) 72  µm (R2) 2.9

107
Fig. 4.10 TEM micrographs, with grain size distributions, of QI-forged cryomilled Al
5083 plate viewed in three directions with respect to processing: (a) normal to forging
and rolling, (b) rolling, and (c) forging (transverse) axes.

108
Fig. 4.11 Optical micrographs (etched) of standard Al 5083 H131 plate: viewed
normal to the primary rolling direction (R1): (a) normal to the transverse axis (R2) and
(b) parallel to the transverse axis.

109
which has been observed previously for HIP/forged cryomilled Al 5083 [12]. Both plates
produced from cryomilled powder exhibited high strength, as great as 60% (yield stress)
and 35% (UTS) more than the standard Al 5083 H131 plate, as shown in Table 4.6. The
QI forged plate exhibited a yield stress and UTS close to 420 and 475 MPa, respectively,
in both directions with respect to rolling. These strength levels were nearly identical to
the HIP/extruded plate in the extrusion direction. However, the strength of the
HIP/extruded plate was lower in the rolling direction, at 394 and 455 MPa, respectively.
The QI forged plate was more ductile in both directions tested, particularly normal to the
rolling direction, elongating over 13%. For the HIP/extruded plate, the ductility was less
than the standard H131 plate, and the ductility was higher in the extrusion direction than
in the rolling direction.
The fracture toughness of the QI forged plate was markedly greater than the
fracture toughness of the HIP/extruded plate, and also greater than the standard plate, as
shown in Table 4.7. In particular, the fracture toughness for cracks in the plane of the
HIP/extruded plate was low ( ∼7 MPa √m). The toughness in the plane of the plate was
low enough that cracks initiated on a plane normal to either the rolling or extrusion
directions changed onto this plane during fracture. In contrast, the QI forged plate was
more isotropic: the fracture toughness for cracks on planes parallel to forging was 26–27
MPa √m, only ∼4 MPa √m greater than for planes normal to forging. Even for the weaker
orientations, the QI-forged plate had greater fracture toughness than the standard plate,
which was isotropic, despite its elongated grain structure. The fracture surfaces for cracks
in the plane of the HIP/extruded plate provided clear evidence that inter-particle de-

110

Table 4.6 Tensile properties of rolled Al 5083 plates.

Al 5083 Processing Direction
0.2% YS
(MPa)
UTS
(MPa)
Elongation
(%)
Rolling 387 455 8.9
HIP & Extrusion
Extrusion 428 469 10.1
Normal to Rolling 413 476 13.1
Cryomilled
QI Forging
Rolling 427 476 11.5
Length (R1) 275 352 10.5
Standard H131
Width (R2) 266 354 10.2

111

Table 4.7 Fracture toughness of rolled Al 5083 plates.

Al 5083 Processing
Crack
Designation
Crack Plane
Normal
Crack
Propagation
K
q

(MPa √m)
T
R
Transverse Rolling 7.2
T
E
Transverse Extrusion 6.8
E
T
Extrusion Transverse 13.8

HIP &
Extrusion
R
T
Rolling Transverse 16.5
F
R
Forging/Transverse Rolling 22.2
F
N
Forging/Transverse 90º to Rolling 22.7
N
F
90º to Rolling Forging/Transverse 26.3
Cryomilled

QI Forging
R
F
Rolling Forging/Transverse 26.5
F
L
Transverse Plate length (R1) 18.6
Standard H131
W
L
Plate width (R2) Plate length (R1) 18.7

112
cohesion had occurred during fracture, as shown in Fig. 4.12(a). Well-defined outlines of
prior particles, elongated in the extrusion direction and flattened in the transverse
direction, were observed. Closer inspection revealed a dimpled morphology, as also
shown in Fig. 4.12(a), indicating some local ductility to the fracture process. The dimples,
however, were shallow. In comparison, the fracture surface of the QI-forged plate with
the equivalent crack plane orientation showed a less defined prior particle structure, as
shown in Fig. 4.12(b). Evidence of the prior particle structure was still apparent, the
surface undulating over a scale of 10–50 µm, but overall the features were more rounded.
The surface showed ductile dimples of from 200 nm to 2 µm, on a scale similar to the
grain size observed, and they often contained small particles. These dimples were
generally deeper and smoother than those of the HIP/extruded plate. The fracture surface
of standard Al 5083 plate indicated extensive inter-granular failure, as shown in Fig. 4.13.
The fracture morphology consisted of a series of flat plateaus terminating in steep ledges
on the scale of 20–200 µm, matching the microstructure shown previously in Fig. 4.11(b).
Surprisingly, despite reasonable fracture toughness, there was little evidence of ductile
failure, and most of the surface was smooth and/or cracked.
The parameters for the HIP/extrusion described here did not result in complete
densification of the cryomilled powder. Although the effect of porosity on ductility and
fracture toughness are yet to be fully determined, the fracture surfaces gave little
indication that the pores had any significance. Instead, the fracture surfaces clearly
revealed that the primary differences between the plates were the nature of the PPBs.
This is a common issue for powder metallurgical Al alloys where the surfaces of the

113
Fig. 4.12 SEM fractographs for crack surfaces in the plane of rolled cryomilled Al
5083 plates: (a) HIP/extruded (T
R
) and (b) QI forged (F
R
).

114
Fig. 4.13 SEM fractograph for a crack in the plane (F
L
) of standard Al 5083 H131
plate.

115
particles are typically covered with a thin layer of oxide and adsorbed water/hydroxides.
These surface films weaken the structure and provide pathways for easy crack
propagation. PPBs are likely to be especially influential for consolidated cryomilled
powder because of the high strength of the matrix.
Because of the repeated fracturing of the particles in liquid N
2
during cryomilling,
the surface of the as-milled powder is relatively free from oxide phases. However, during
handling of the powder, even in the low O
2
/moisture environment of a glove box, oxide
formation will occur because the partial pressure of O
2
required to oxidize aluminum is
extremely low, < 10
-10
Torr [16]. In addition, the hot vacuum degassing of the powder,
performed to remove the stearic acid, may also contribute to oxidation and contamination
of the particle surfaces.

4.2.5. Conclusions
Two consolidation and deformation processing methods, HIP/extrusion and two-
step QI forging, successfully produced, from cryomilled Al 5083 powder, plates of a
larger size than has been achieved before. Both of the plates possessed a similar UFG
microstructure, with grains mostly in the range 100–500 nm elongated in the deformation
directions. The QI forged plate had grains 40–80 nm greater than the HIP/extruded plate,
depending on orientation. The tensile strengths in the plane of the plates were similar and
up to 60% greater than conventionally processed Al 5083, with acceptable ductility.
However, the fracture toughness of the QI forged plate was clearly superior. Despite
highly deforming the powder particles, extrusion did not sufficiently increase the strength
of the PPBs and these were very weak, leading to delamination and low toughness for

116
crack surfaces in the plane of the plate. The QI forging more efficiently broke up the
PPBs, leading to relatively isotropic fracture toughness, greater than standard Al 5083
plate. Although there is some uncertainty involved in making an absolute comparison
between HIP/extrusion and QI forging routes because of different processing parameters,
the QI forging is a simple route, negating the need for HIP, and obtaining fairly high
material yield. In preference to extrusion, the forging route is geometrically suitable for
making large rolling pre-forms.

4.3. Strain rate sensitivity
The hardness of as-cryomilled NC powder and UFG material was measured by
nanoindentation using loading rates in the range of 50−50,000 µN/s, and results were
compared with the conventional grain size alloy. Negative strain rate sensitivity (SRS)
was observed in the cryomilled NC powder and the forged UFG plate, while the
conventional CG alloy was relatively strain rate insensitive. The cryomilled NC showed
greater negative SRS than the forged UFG, indicative of the effect of grain size on SRS.
Compression tests for bulk materials confirmed the negative SRS of UFG material, and it
was compared with the conventional alloy which showed negligibly positive SRS
exponent.

4.3.1. Motivation
Deformation of bulk NC/UFG materials at various strain rates has been used to
investigate mechanisms for improving the strength and ductility of the materials [17 −21].
Usually an increase in the flow stress is observed at higher strain rates, indicative of

117
positive SRS. According to some studies, the effect of positive SRS becomes significant
with decreasing grain size in face-centered cubic (FCC) metals [17 −19], while body-
centered cubic (BCC) metals exhibit decreased SRS with decreasing grain size [21,22].
Recently, negative SRS found in the UFG Al 5083 alloy along with serrated plastic flow
[19] has been attributed to dynamic strain aging (DSA), while that found in BCC
tantalum has been attributed to twinning at higher strain rates.
Recent studies to estimate the SRS have involved either standard bulk specimen
testing [19] or nanoindentation experiments [20,23]. Nanoindentation measurements have
been developed to measure mechanical properties [24,25] locally in small regions, such
as the cross section of powder particles. In the present study, the Al 5083 alloy specimens
with grain size scale of NC, UFG and conventional CG have been subjected to
nanoindentation at different loading rates and compression testing to examine the effect
of grain size on the SRS.

4.3.2. Materials and processing details
A 20 kg batch of gas-atomized, -325 mesh (< 44 µm) Al 5083 powder produced
by Valimet, Inc. (Stockton, CA) was cryomilled in liquid N
2
by DWA Aluminum
Composites (Chatsworth, CA) for 8 hours in a modified Szegvari attritor. Stainless steel
(440C) milling balls were used with a ball-to-powder weight ratio of 32 : 1, and 40 g (0.2
wt.%) ‘stearic acid’ (~50/50 palmitic/stearic acids) was added to the liquid N
2
/powder
slurry to improve process yield. After milling, the powder was transferred to a glove box
under liquid N
2
, which was subsequently allowed to boil off, ensuring that atmospheric
contamination of the cryomilled powder was minimized. The cryomilled powder was

118
degassed by heating it in a tube furnace to 450 ºC under a vacuum (~10
-5
Torr) drawn
through a stem in the lid of container and cooled to room temperature, and the stem was
permanently welded shut.
The degassed powder was QI forged by Advanced Materials & Manufacturing
Technologies, LLC (Roseville, CA) by placing in a bed of graphitic powder, pre-heated
to 500 ºC, contained within a 200 mm diameter cylindrical die. The maximum pressure
applied to the die was 343 MPa. The forged disk was machined to remove the canning
material, then QI forged once again. Final dimensions of the disk were 128 mm
(diameter) × 16 mm (thickness).

4.3.3. Mechanical tests and characterization procedures
The nanoindentation measurements were performed using a TriboIndenter
(Hysitron, Minneapolis, MN) with a 100 nm Berkovich tip. Tests were carried out on
resin-mounted Al 5083 alloy samples of the cryomilled NC powder, the forged UFG
plate, and the conventional CG alloy. The loading rates were varied between 50 and
50,000 µN/s, and the maximum load used for each test was 5,000 µN. About 25–50
indents were performed at each indentation rate for each specimen. During the tests, the
standard nanoindentation drift correction software was applied. For uniaxial compression
tests of bulk materials, cubic samples with an edge of 5 mm in length were prepared, and
the tests were performed using an Instron universal testing machine.
Samples of the materials, both cryomilled powder and forged plate, were prepared
for TEM examination (Philips EM420). Cryomilled powders were prepared by mixing
with an epoxy (G-1, Gatan Inc) and curing at ambient temperature, then ground, polished,

119
dimpled and ion milled to perforation for electron transparency. TEM samples from the
forged plate were prepared by electrolytic jet polishing using an ethanol solution
containing 8% perchloric acid + 10% 2-butoxyethanol. Grain size was measured directly
from the TEM images using two different methods, a linear intercept method and
individual grain measurements for ~350 grains generating grain size histograms. In case
of the cryomilled powder, the grain size measurements were performed on 6–7 powder
particles. The mean grain size and aspect ratio of the standard Al 5083 plate was obtained
from optical micrographs after chemical etching using a linear intercept method. For
chemical analysis of metallic and non-metallic alloying elements, the detailed processes
are described in Chapter 4.2.3.

4.3.4. Experimental results and discussion
Composition of the consolidated cryomilled Al 5083 powder is shown in Table
4.4, along with the specified composition of standard Al 5083. Examination of the
compositions shown in Table 4.4 indicates that the concentrations of non-metallic
elements, N, H, O and C, are much higher in the forged UFG plate made using
cryomilled powder than in the specified standard Al 5083 alloy composition. TEM
studies and chemical analyses have not shown much evidence of the presence of nitrides,
carbides and oxides, which indicate that a large of the N, H, O and C are expected to be
present as interstitial atoms in the Al 5083 alloy. Again, thermodynamics dictate that the
high angle grain boundaries will be enriched in solute atoms.
The microstructures showing the grain structure of the cryomilled NC powder and
the forged UFG plate are shown in Fig. 4.14(a) and (b), respectively. The selected area

120
Fig. 4.14 TEM bright field micrographs of (a) cryomilled NC powder and (b) forged
UFG plate with corresponding SAD patterns, respectively

121
diffraction (SAD) patterns from the areas displayed in the bright field TEM images are
also shown as inserts in Fig. 4.14(a) and (b), respectively. The SAD pattern in Fig.
4.14(a) from the cryomilled NC powder shows only rings, which are characteristic of
randomly distributed nano-size grains. On the other hand, individual spots arranged in
rings are distinguishable in the SAD pattern shown in Fig. 4.14(b) from the consolidated
UFG material. The spots in the SAD pattern arise from the CG structure. Figure 4.15
shows histograms depicting the grain size distributions in the cryomilled CG powder and
in the forged UFG plate, respectively. The average grain size of all three materials along
with the aspect ratio of the grains, the cryomilled NC powder, the forged UFG plate and
the standard CG plate, is shown in Table 4.8. The average grain size of the cryomilled
powder (50 nm) was less than a fifth of that in the forged plate (~250 nm). In addition,
the grain size distribution of the forged plate was much wider than that of the cryomilled
powder, and the sizes of the conventional CG material are an order of magnitude larger.
The larger average grain size and the wider grain size distribution in the forged UFG
material (compared to the cryomilled NC powder) is attributed to grain coarsening during
consolidation at high temperature. The average grain size for the conventional CG
material is three orders of magnitude greater than that in the forged UFG material, as
shown in Table 4.8.
Figure 4.16 shows load vs. depth indentation curves performed at a constant
loading rate (1,000 µN/s) for all three materials, NC powder, UFG plate and CG plate,
respectively. Indentation depth achieved at the end of the loading segment varies with
respect to the grain size from ~325 nm (CG plate) to ~200 nm (NC powder), indicative of

122
Fig. 4.15 Histograms of the grain size distribution of (a) cryomilled NC powder and
(b) forged UFG plate when viewed normal to the forging direction.

123

Table 4.8 Mean grain size and aspect ratio of the as-cryomilled NC powder, forged
UFG plate, and standard Al 5083 plate (measured from TEM micrographs and optical
micrographs).

Individual grains Linear Intercept method
Materials
Mean size Aspect ratio Mean Size Aspect ratio
Cryomilled NC powder 50 nm 1.5 − −
Forged UFG plate
a
252 nm 1.8 254 nm 1.7
Standard Al 5083 plate
b
− − 132 µm 6.5

a
viewed normal to the forging direction;
b
viewed normal to the rolling direction.

124
Fig. 4.16 Load vs. indentation depth curves performed at a loading rate of 1,000 µN/s
for: (a) conventional CG plate, (b) UFG plate and (c) NC powder.

125
grain size effect on the hardness. When grain size decreases, more deviation of unloading
curves was observed indicating more deviation of the hardness measured. Hardness
variation for all materials with respect to the loading rates (50 −50,000 µN/s) measured by
nanoindentation is summarized and plotted in Fig 4.17. The standard deviations are
shown as error bars for each test result. From a careful study of the plot, the following
may be inferred: (i) the standard deviations of the hardness values are greater for the
cryomilled NC powder and the forged UFG material than for the conventional CG
material, as also shown in Fig. 4.16; (ii) the hardness decreases with increasing loading
rate for the NC powder and the UFG material, while it remains unchanged for the CG
material; (iii) the drop in hardness with loading rate for the NC powder is much sharper
than for the UFG material; (iv) the drop in hardness with loading rate is more significant
at the lower loading rates; (v) the average hardness of the UFG material is 40 −60%
greater than the CG alloy, depending on the loading rate; and (vi) the hardness of the NC
powder is ~2 and 2.6 times greater than the CG material at the lowest and highest loading
rates, respectively. The greater standard deviations for hardness values obtained from the
cryomilled NC powder and the forged UFG sample are attributed to microstructural
heterogeneities, particularly due to the grain size distributions. The higher hardness
values in samples with the NC structure (50 nm) follow expectations based on the Hall-
Petch relationship. The decrease in the hardness with loading rate indicates negative SRS
in these alloys, particularly when the average grain size is less than 100 nm.
Figure 4.18(a) shows plots of yield stress against strain rate by compression tests
for bulk samples of the forged UFG and conventional CG plates. Comparison of the plots

126
Fig. 4.17 Hardness variations of the cryomilled NC powder, forged UFG and
conventional CG plate with respect to the loading rates measured by nanoindentation.

127
Fig. 4.18 Compression behavior of the UFG and CG plates: (a) strain rate sensitivity
exponent and (b) true stress-strain behavior with respect to different strain rates.

128
shows: (i) the yield stress of the forged UFG alloy was higher; and (ii) the SRS exponent
(m) of the UFG alloy was negative (m = –0.0108), while that of CG alloy is smaller and
positive (m = 0.0025). The higher yield stress and negative SRS of the UFG alloy are
consistent with the results of hardness measurements depicted in Fig. 4.17. Figure 4.18(b)
shows typical true stress-true strain curves of the forged UFG and CG alloys recorded
during compression tests at the strain rates of 10
-1
, 10
-3
and 10
-5
s
-1
. The stress-strain
behaviors indicate that the strain hardening rates are independent of the strain rates and
also of the grain size. However, transition between elastic and plastic deformation was
sharper for the UFG material, compared with smooth transition in the CG material.
Furthermore, the slope of the elastic regime of the plots representing the CG alloy is
steeper than that of the UFG alloy, suggesting that the Young’s modulus of the former
material is higher. The lower Young’s modulus of the UFG than the CG alloy is
attributed to defects, such as porosity, left after consolidation of the powder.
The negative SRS observed during the nanoindentation experiments and
compression tests for the NC and UFG samples is contrary to the behavior reported for
the FCC metals and alloys in literature [17,18,21]. Negative SRS is usually observed in
materials exhibiting dynamic strain aging [19,26]. The dynamic strain aging is caused by
the attractive interaction of dislocations with interstitial solute atoms including those of
nitrogen, carbon, hydrogen and oxygen. The pinning of dislocations by the interstitial
solute atoms causes jerky glide of the dislocations. At high loading rates or strain rates,
there is a rapid increase in the density of mobile dislocations, particularly when the
dislocations are locked-in and are not free to move otherwise. The larger plastic

129
deformation accompanying the generation of increased density of mobile dislocations
manifests as larger indentation size and lower hardness during nanoindentation and lower
yield strength in compression tests. Both the density of dislocations and their mobility
decrease as the grain size decreases, and this is particularly pronounced when the average
grain size is less than 100 nm. Hence, the effect of higher loading rates on triggering of
dislocation motion is more pronounced when the grain size is in the NC range. In the NC
and UFG alloys, the grain boundaries act not only as principal sources of dislocations, but
also as sinks of dislocations. As the grain boundaries are expected to be enriched in solute
atoms, the solute-dislocation interaction is expected to increase with decrease in grain
size. The fact that negative SRS is more pronounced in the indentation tests with slower
loading rates is intriguing and suggests that thermally activated grain boundary diffusion
is involved in deformation at lower strain rates, in addition to dislocation glide. The
higher hardness or strength observed at the lower strain rate implies that the solute atoms
are capable of diffusing to dislocation cores and form a Cottrell type atmosphere
restricting their motion. Hence, the drag of solute atoms on mobile dislocations is
significantly large at lower strain rates.

4.3.5. Conclusions
The strain rate sensitivity of Al 5083 alloy in the form of cryomilled NC powder,
forged UFG plate and conventional CG material with the average grain size of 50 nm,
253 nm and 132 µm, respectively, has been examined using nanoindentation at different
loading rates and compression tests with strain rates between 3×10
-6
and 3×10
-1
s
-1
. The
negative SRS was relatively less prominent in the UFG material than in the NC powder,

130
while it was slightly positive in the conventional CG alloy. The negative SRS in the NC
and UFG materials is attributed to the pinning of dislocations by interstitial solute atoms
as well as the small grain size, both of which restrict dislocation motion. Indentation at
high loading rates is expected to trigger dislocation motion by unlocking from solute
atoms, increasing the density of mobile dislocations, causing higher plastic deformation
at the same load. The negative SRS was more pronounced at slower loading rates during
nanoindentation, suggesting that grain boundary diffusion of solute atoms have an
important role in pinning dislocations.

131
References
[1] C. Suryanarayana, Prog. Mater. Sci. 46 (2001) 1.

[2] P.S. Gilman and J.S. Benjamin, Ann. Rev. Mater. Sci. 13 (1983) 279.

[3] J.R. Groza, In: C.C. Koch, editor, Nanostructured Materials: Processing,
Properties, and Applications, 2nd Edition, Norwich, NY: William Andrew
Publishing (2007) pp. 173 −233.

[4] V.L. Tellkamp, A. Melmed and E.J. Lavernia, Metall. Mater. Trans. 32A (2001)
2335.

[5] Z. Lee, R. Rodriguez, R.W. Hayes, E.J. Lavernia and S.R. Nutt, Metall. Mater.
Trans. 34A (2003) 1473.

[6] D. Witkin, Z. Lee, R. Rodriquez, S. Nutt and E. Lavernia, Scripta Mater. 49 (2003)
297.

[7] Z. Lee, D.B. Witkin, V. Radmilovic, E.J. Lavernia and S.R. Nutt, Mater. Sci. Eng.
A 410 −411 (2005) 462.

[8] V.L. Tellkamp, S. Dallek, D. Cheng and E.J. Lavernia, J. Mater. Res. 16 (2001)
938.

[9] D.E.J. Talbot, The Effects of Hydrogen in Aluminum and Its Alloys, Cambridge,
MA: Maney (2004).

[10] D. Witkin, B.Q. Han and E.J. Lavernia, Metall. Mater. Trans. 37A (2006) 185.

[11] A.P. Newbery, B. Ahn, P. Pao, S.R. Nutt and E.J. Lavernia, Adv. Mater. Res.
29−30 (2007) 21.

[12] A.P. Newbery, B. Ahn, R.W. Hayes, P. Pao, S.R. Nutt and E.J. Lavernia, Metall.
Mater. Trans. (2008) in press.

[13] B.Q. Han, Z. Lee, D. Witkin, S. Nutt and E.J. Lavernia, Metall. Mater. Trans. 36A
(2005) 957.

[14] P.S. Pao, H.N. Jones, C.R. Feng, D.B. Witkin and E.J. Lavernia, In: Y.T. Zhu et al.,
editors, Ultrafine Grained Materials IV, Warrendale, PA: The Minerals, Metals and
Materials Society (TMS) (2006) pp. 331 −336.

132
[15] D.B. Witkin and E.J. Lavernia, In: L.L. Shaw et al., editors, Processing and
Properties of Structural Nanomaterials, Warrendale, PA: The Minerals, Metals and
Materials Society (TMS) (2003) pp. 117 −124

[16] B. Rufino, F. Boulc’h, M.-V. Coulet, G. Lacroix and R. Denoyel, Acta Mater. 55
(2007) 2815.

[17] R. Schwaiger, B. Moser, M. Dao, N. Chollacoop and S. Suresh, Acta Mater. 51
(2003) 5159.

[18] B.Q. Han, J.Y. Huang, Y.T. Zhu and E.J. Lavernia, Scripta Mater. 54 (2006) 1175.

[19] B.Q. Han, J. Huang, Y.T. Zhu and E.J. Lavernia, Adv. Eng. Mater. 8 (2006) 945.

[20] Y.M. Wang, A.M. Hodge, P.M. Bythrow, T.W. Barbee, Jr. and A.V. Hamza, Appl.
Phys. Lett. 89 (2006) 081903.

[21] Q. Wei, S. Cheng, K.T. Ramesh and E. Ma, Mater. Sci. Eng. A 381 (2004) 71.

[22] Q. Wei, J. Mater. Sci. 42 (2007) 1709.

[23] P. Cavaliere, Physica B 403 (2008) 569.

[24] W.C. Oliver and G.M. Pharr, J. Mater. Res. 7 (1992) 1564

[25] G.M. Pharr, W.C. Oliver and F.R. Brotzen, J. Mater. Res. 7 (1992) 613.

[26] P. Rodriguez, Bull. Mater. Sci. 6 (1984) 653.

133
Chapter 5. Al Alloys with Bimodal Microstructure

5.1. Multi-scale concept
5.1.1. Microstructure
Bulk nanostructured materials (BNMs) are defined as bulk solids with nano-scale
or partially nano-scale microstructures. They have received increasing attention because
of the enhancement of mechanical properties and the potential for structural applications.
For example, BNMs exhibit enhanced strength compared with conventional materials [1–
3], and the phenomenon is attributed to the fine grain size on the basis of the Hall-Petch
mechanism [4,5]. Different techniques have been developed to produce nanocrystalline
(NC) or ultrafine-grained (UFG) structures, including mechanical alloying [6,7], severe
plastic deformation [8], gas-phase condensation [9] and electrodeposition [10]. However,
BNMs generally exhibit unacceptably low ductility and toughness, greatly limiting their
widespread practical utility. Multiple processing approaches have been employed to
improve these properties [11–14]. In one such approach, the microstructural modification
led to a multi-scale microstructure characterized by a bimodal grain structure [15–17].
The microstructure was comprised of micron-sized CG regions embedded in a NC matrix.
Figure 5.1 shows a schematic of multi-scale synthesis. The multi-scale microstructure
was achieved by blending unmilled coarse-grained (CG) powder with cryomilled NC
powder and subsequently consolidating. Two different grain size scales, bimodal
structure, fabricated by cryomilling, HIP and extrusion processes were attempted with
various CG fractions [15–17]. Figure 5.2 shows optical micrographs of etched as-
134
Fig. 5.1 Schematic of synthesis of a multi-scale grain structured material, incorporating
CG material into cryomilled NC material to improve ductility and toughness.
135
Fig. 5.2 Optical micrographs of as-extruded bimodal alloys containing 15, 30 and 50%
CG content, (a), (b) and (c), respectively, when viewed from normal and transverse to the
extrusion direction. (brighter tone indicates CG regions in darker NC matrix).
136
extruded bimodal materials with three different CG fractions (15, 30 and 50 %), viewed
from two different orientations (the normal and transverse to the extrusion direction). CG
regions (brighter tone in the micrographs) formed elongated band structures along the
extrusion direction in the continuous NC matrix (darker tone in the micrographs). TEM
micrographs of a bimodal material containing 50% CG are shown in Fig. 5.3, when
viewed normal to the extrusion direction. Multi-scale grain structures are apparent in both
bright field image (Fig. 5.3(a)) and dark field image (Fig. 5.3(b)). The mean grain sizes
were ~200 nm for NC regions and ~3 µm for CG regions.

5.1.2. Tensile behavior
The bimodal alloys reportedly showed improved ductility and toughness with
only a modest decrease in strength when compared to the purely NC control alloy [15–
18]. The tensile properties of bimodal materials (15 and 30% CG) are shown in Fig. 5.4,
compared with a 100% NC material and a conventional Al 5083 alloy with 100% CG
structure [15,17]. In case of a bimodal material with 50% CG, an estimated stress-strain
behavior calculated by finite element method (FEM) simulation was also shown in Fig.
5.4 [18]. The 100% NC cryomilled material exhibited yield strength more than four times
that of conventional one. The bimodal material with 15% CG exhibited a slight
improvement in ductility and brief work softening. The ductility of the bimodal material
with 30% CG increased more than three times compared to 100% NC material, with a
modest decrease in the yield strength of 14% [15]. According to the FEM simulation, the
material with 50% CG expected to show a smooth transition from elastic to plastic
deformation with extended range of strain hardening [18]. The resulting properties have
137
Fig. 5.3 TEM micrographs showing global microstructure of as-extruded bimodal alloy
containing 50% CG: (a) a bright field image and (b) a dark field image.
138
Fig. 5.4 Comparison of tensile stress-strain curves between experimental behavior [15]
and calculated behavior by FEM simulation [18] of bimodal materials as well as 100%
NC and conventional CG materials.
139
generated strong interest, although the fundamental mechanisms of plastic deformation of
the bimodal alloys are not well understood.

5.2. Strain localization
The deformation response of a UFG Al-Mg alloy with bimodal grain structure
was investigated using a micro-straining unit and a strain mapping technique. Atomized
Al 5083 powder was ball-milled in liquid N
2
to obtain a NC structure, then blended with
50 wt.% unmilled CG powder and consolidated to produce a bimodal grain structure. The
blended powder was hot vacuum degassed to remove residual contaminants, consolidated
by cold isostatic pressing (CIP), and then quasi-isostatic (QI) forged twice. The resultant
material consisted of a UFG matrix and CG regions. The dynamic response during tensile
deformation was observed using a light microscope, and the surface displacements were
mapped and visualized using a digital image correlation (DIC) technique. The DIC
results showed inhomogeneous strain between the UFG and CG regions after yielding,
and the strain was localized primarily in the CG regions. Strain hardening in the CG
regions accompanied the localization, and was confirmed by variations in Vickers
hardness.

5.2.1. Motivation
Cryomilling, which refers to ball milling in a cryogenic fluid (typically liquid
nitrogen), is a particularly effective process for producing NC powders, both in small
batches (100’s of grams) and commercial quantities (30–40 kg), while minimizing heat
generation and oxygen contamination [7]. In this process, the powder particles are
140
repeatedly sheared, fractured, and cold-welded over a period of several hours. The
cryomilled powder is subsequently hot-vacuum degassed to remove process control
agents and other contaminants. For primary consolidation, hot isostatic pressing (HIP), is
often employed to achieve high compaction density. However, the HIP process often
causes undesired grain growth, and CIP is sometimes employed to avoid this. Although
CIPped materials exhibit lower green-body density (~80%), the CIP process effectively
prevents grain growth, and greatly reduces the cost and time of processing. In this study,
CIP consolidation was followed by the QI forging [19,20] to break up prior particle
boundaries (PPBs) remaining after CIP. During QI forging, the uniaxial pressure was
transformed into a QI pressure field by the application of a granular pressure transmitting
medium, and this introduced an additional shear stress which helped to break up PPBs in
the bimodal alloys.
However, BNMs often possess insufficient ductility and a reduced toughness
compared to CG conventional materials. In the present study, a novel approach analogous
to ductile phase toughening was used to achieve a balance of high strength and acceptable
toughness. Multi-scale grain structures were produced consisting of cryomilled NC
regions and conventional CG regions. The unusual bimodal grain structure of the
resulting alloys raises some interesting questions concerning the distribution of strain
during deformation, since the deformation mechanisms in such multi-scale materials are
not fully understood.
The accurate measurements of local displacements and strains during deformation
have been an important issue in the field of experimental mechanics. However, research
trends toward reducing size scales to the micro/nanometer range give rise to difficulties
141
when attempting to measure strains using conventional extensometry. In the case of non-
uniform complex microstructures, such as the multi-scale structured material in the
present study, conventional methods measure strains through the material without regard
for the multi-scale microstructure, thus resulting in averaged strain values.
DIC is a non-contacting strain field measuring technique that relies on
mathematical correlation of digital images recorded during deformation under an applied
load. The DIC technique was first conceived a few decades ago and has been well
developed recently to analyze and to visualize the actual strain distributions within
materials [21 −25]. The basis of the DIC technique involves the mathematical tracking of
numerical patterns consisting of different gray-scales of pixels between two digital
images. In the present case, two-dimensional (2-D) DIC was performed using a high
resolution digital camera coupled with a light microscope, and it was used to measure
surface displacements of planar specimens. In the 2-D DIC process, a computer-aided
correlation is carried out using a pair of sequential images to determine surface
displacements, which are recognized by movement of speckle patterns in digital images.

5.2.2. Experimental procedures
Gas-atomized -325 mesh (< 44 µm) Al 5083 (Al-4.4Mg-0.7Mn-0.15Cr wt.%)
powder (produced by Valimet, Inc.) was cryomilled in a 20 kg batch by DWA Aluminum
Composites for 8 hours in a modified Szegvari attritor. Stainless steel balls with a
diameter of 6.4 mm were used with a ball-to-powder weight ratio of 32 : 1. To improve
milled powder yield, 0.2 wt.% of stearic acid was added. The cryomilled powders were
combined and mechanically blended with 50 wt.% non-cryomilled gas-atomized Al 5083
142
powder in an inert atmosphere to achieve a uniform distribution of bimodal
microstructure. The blended powder was degassed at 450 ºC under a vacuum of 10
-6
Torr,
and the degassed powder was transferred to a rubber CIP bag inside a glove box filled
with dry nitrogen, then consolidated by CIP at 310 MPa pressure by Pittsburgh Materials
Technology, Inc. The resulting CIPped billet was dry-machined to a cylindrical shape, 5
cm (diameter) × 12.7 cm (height). The CIPped billet was sealed by Al 6061 can, pre-
heated for 2 hours at 454 ºC, and then quasi-isostatically forged (Advance Materials &
Manufacturing Technologies, LLC). After the first forging, the billet was re-heated to
454 ºC for half an hour, then secondarily forged at 180º to the original direction. Final
dimensions of the material were 6 cm (diameter) × 5 cm (height).
Two tensile specimens were prepared from the as-forged material to achieve a flat
“dog-bone” shape. The gauge section was perpendicular to the forging direction with the
dimensions of 4 mm (length) × 2.5 mm (width) × 1.5 mm (thickness). To confine the area
of stress concentration for dynamic observation, two notches were introduced in the
middle of the gauge section on opposite sides, as shown in Fig. 5.5(a). The first specimen
was used to acquire the nominal stress-strain curve, and the second one was used for
hardness measurements during multiple intermissions during a tensile test. The
specimens were mechanically polished to obtain mirror-like surfaces and then lightly
etched for subsequent observation of the bimodal grain structure.
A micro-tensile stage module (Deben, UK) was utilized for the tests. The module
was specifically designed to facilitate observation of micro-strain in microscopes. All
tensile tests were performed at room temperature with a displacement rate of 1.67 µm/s,
which, for the custom designed specimens in this study, was approximately equivalent to
143
Fig. 5.5 (a) Geometry of the double-notched tensile specimen showing experimental
scheme and (b) stress-strain curve of the notched specimen showing the points
interrupted for hardness measurements.
144
a strain rate of 4.18×10
-4
s
-1
. The stage module was attached to a light microscope
(Olympus Vanox AHMT3), and the dynamic deformation during the tensile test was
sequentially recorded at time intervals of half a second using a digital camera (QICam
12-bit, QImaging Corp.) installed with the light microscope.
For hardness measurements, a tensile test was interrupted before fracture at
different tensile stresses (0, 182, 364, 455, and 515 MPa) as shown in Fig. 5.5(b). The
Vickers hardness of both UFG regions and CG regions was obtained at each intermission
using a microhardness tester (Future-Tech Corp). The indentations were made on two
different locations on the gauge section - adjacent to the notches and in the central part
between the notches - using a 10 gf load and a dwell time of 15 seconds.
To measure the displacements that developed during tensile deformation, pairs of
micrographs were recorded and selected. Each pair consisted of a reference and a
deformed image within a range of both large strain and micro-strain, and the resolution of
digital images was 1392×1040 pixels. Two-dimensional digital image correlation was
performed using software (VIC-2D, Correlated Solutions, Inc.) with a subset size of
41×41 pixels and a step spacing of 5 pixels.
Details of the microstructures were characterized by transmission electron
microscopy (TEM) and scanning electron microscopy (SEM). The TEM was used to
investigate the bimodal grain structure, and the specimens were prepared by electrolytic
twin jet polishing in a solution containing ethanol + 8% perchloric acid + 10% 2-
butoxyethanol at a temperature of −40 ºC. The number-based mean grain size was
obtained directly from TEM images by measuring ~400 individual grains. The fracture
surface of the tensile specimen was examined by SEM (Hitachi S-4800 FE-SEM).
145
5.2.3. Experimental results and discussion
Microstructures of as-forged material are shown in Fig. 5.6. The bimodal grain
structure is apparent in optical micrographs (Fig. 5.6(a) and (b)), and consists of a UFG
matrix and CG regions (light contrast). In general, the CG areas were uniformly
distributed in the UFG matrix, as shown in Fig. 5.6(a) and (b). The grains were elongated
perpendicular to the forging direction, as shown in Fig. 5.6(c) and (d). The mean grain
size measured directly from TEM images was 247 nm for UFG regions and ~4 µm for
CG regions in the bimodal microstructure. As reported previously, the as-cryomilled
powder exhibited a mean grain size of 50 nm with a narrow size distribution, although
some grains were as large as 200 nm [26]. However, most of the nanosized grains
produced by cryomilling underwent limited grain growth during consolidation.
(Although CIP was employed for primary consolidation in lieu of HIP, post-CIP
processing included hot degassing and forging, which involved moderately high
temperatures.)
The tensile behavior of a notched specimen of the as-forged bimodal alloy is
shown in the nominal stress-nominal strain curve shown in Fig. 5.7. For comparison,
engineering stress-strain curves for a 100% NC Al-7.5Mg alloy [15] and the conventional
Al 5083 alloy [15] are also plotted in Fig. 5.7. The nominal stress was calculated by the
applied load divided by the net cross-sectional area at the notches, and the nominal strain
was obtained using the displacements divided by the gauge length regardless of the
notches. The yield strength of a notched specimen in the direction normal to the forging
direction was 265 MPa, almost twice the strength of Al 5083 alloy with conventional
grain size (145 MPa) [15]. The tensile elongation (5.5%) of this bimodal grain structured
146
Fig. 5.6 Microstructures of the as-forged material showing a multi-scale grain structure;
(a) and (b) optical micrographs (etched), (c) a bright field TEM image, and (d) a dark
field TEM image.
147
Fig. 5.7 Nominal stress-strain behavior of the notched bimodal specimen with 50% CG,
compared with engineering stress-strain curves of a 100% NC [15] and a conventional Al
5083 alloy [15].
148
materials (50% NC + 50% CG) was also four times greater than that of cryomilled 100%
NC material (1.4%) [15].
A selected sequence of images of deformation and fracture around a notch during
a tensile test is shown in Fig. 5.8. Figure 5.8(a) is an initial image before the deformation
and serves as a reference image. The images of deformed material, Fig. 5.8(b) −(d), were
selected from different stages of deformation, 60, 100 and 110 seconds after the reference,
which are equivalent to total strains of 2.51×10
-2
, 4.18×10
-2
and 4.59×10
-2
, respectively.
Figure 5.8(b) shows that micro-voids and micro-cracks started to nucleate at the root of a
notch short time after the material yielded. In Fig. 5.8(c), the micro-cracks grew and
coalesced, forming a network. Finally, the network expanded rapidly toward the center of
the specimen, developing macro-cracks and resulting in material rupture, as shown in Fig.
5.8(d). Prior to failure, no significant area reduction was observed in the gage section.
Figure 5.9 shows the strain field near a notch tip, as measured by DIC using a
reference image and subsequent images after deformation. The images were selected after
yield of the material to correlate plastic strain only. The strain field measured by DIC for
three different time intervals is shown in Fig. 5.9. Strain evolution during the deformation
is readily apparent. In Fig. 5.9(a), DIC was employed for a 30-second interval (1.25×10
-2

strain) between the reference and deformed images. The correlation shows that the
maximum strain occurred in front of the notch tip as a result of the stress concentration.
Also, strain localization is apparent at orientations of approximately 45º to the tension
axis, where the resolved shear stress is a maximum, generating a gradient toward the
center of the specimen. When DIC was performed for a 6-second interval (2.51×10
-3

strain) as shown in Fig 5.9(b), the maximum strain was significantly decreased compared
149
Fig. 5.8 Successive optical micrographs recorded around a notch during tensile
deformation: (a) initial state before deformation started (t
1
= 0 s); (b) voids nucleated
short time after yielding (t
2
= 80 s); (c) fracture started (t
3
= 100 s); and (d) immediately
before rupture of the specimen (t
4
= 110 s).
150
Fig. 5.9 Strain fields around a notch measured by DIC at different strains: (a) large strain
(1.25×10
-2
, 30 second interval), (b) intermediate strain (2.51×10
-3
, 6 second interval), and
(c) micro-strain (4.18×10
-4
, 1 second interval).
151
with Fig 5.9(a). No distinct strain variation was observed between UFG and CG regions
in Fig. 5.9(b). However, when two images in a micro-strain range of 1 second interval
(4.18×10
-4
strain) were correlated, slight strain localization was observed, manifest as
scattered spots in Fig. 5.9(c). The strain localization was attributed to different
deformation behavior of UFG and CG regions. However, the resolution of the DIC
results in Fig. 5.9 was insufficient to provide detailed evidence of the local strain
distribution with respect to the multi-scale grain structure.
In order to investigate the strain localization in detail, another DIC was performed
using two images taken at half-second intervals after yielding, approximately equivalent
to a micro-strain of 2.09 × 10
-4
, near the notches and at the central region between the
two notches. Figure 5.10 shows light micrographs and the corresponding micro-strain
contours for the bimodal structure at the two locations. The DIC results in Fig. 5.10
indicate that the strain is inhomogeneous between UFG and CG regions in both locations.
The strain near the notches shown in Fig. 5.10(a) is overall greater than that between the
notches in Fig. 5.10(b), as one would expect. Also, UFG regions adjacent to CG bands
were more highly strained near the notches compared with distant regions, a consequence
of the higher stress concentration near the notches. For both locations, the strain is
localized primarily within CG regions, indicating that the softer CG regions were more
deformed than the harder UFG matrix. Therefore, more dislocations are generated within
the CG regions and undergo slip during the deformation. Dislocations glide more easily
in CG regions on account of the lower density of obstacles (principally grain boundaries)
in the larger grains. Conversely, the mobility of dislocations is limited within UFG
regions because of the smaller grains. However, any strain differences between UFG and
152
Fig. 5.10 Optical micrographs and corresponding DIC results showing inhomogeneous
strain measured from two locations: (a) near the notches and (b) central part between the
notches.
153
CG regions in micro-strain must be accommodated ultimately, since the total strain
should be equal in both UFG and CG regions. (This was also shown in Fig. 5.9(a) by the
absence of contour variations between UFG and CG at large strains). Thus, different
deformation mechanisms are involved in the UFG regions that ensure compatibility
(equal strains) at the UFG-CG boundaries.
As discussed above, dislocation plasticity is limited in UFG regions compared to
CG regions. However, grain boundary mediated plasticity, such as grain boundary sliding
and grain rotation, is also less likely in the grain size of ~240 nm especially at room
temperature. The initiation and evolution of micro-cracks in UFG regions or at interfaces
between UFG and CG regions may occur to compensate the plastic strain in CG regions
and accommodate strain inhomogeneity. Concurrently, the initiation and fast propagation
of cracks through UFG regions can be delayed by plastic deformation and crack blunting
in CG regions. In general, voids/cracks are expected to initiate and grow from existing
defects in the materials, such as porosity or poor interparticle boundaries.
DIC result when porosity exists at the boundary between UFG and CG regions is
shown in Fig. 5.11. Two images were selected for the DIC analysis, also taken at half-
second intervals (2.09×10
-4
strain) after yielding. The porosity is illustrated as blue circles
in the reference image shown in Fig. 5.11(a), and the corresponding micro-strain contour
is shown in Fig. 5.11(b). The red arrows in Fig. 5.11(b) indicate regions in which the
strain is highly concentrated, and the corresponding regions are indicated by similar red
arrows in Fig. 5.11(a). The strain is concentrated in CG regions adjacent to the pores,
indicative of greater plastic deformation and relaxation of the stress concentration at the
pores.
154

Fig. 5.11 (a) Optical micrograph and (b) corresponding strain field showing strain
concentration (arrows) in CG regions adjacent to porosity (circles).
155
Further evidence of inhomogeneous strain was manifest in Vickers hardness (H
v
)
values measured at the two different locations and within both UFG and CG regions. The
measured H
v
values at each intermission (0, 182, 364, 455, and 515 MPa) during the
tensile test are plotted in Fig. 5.12. The first comparison of H
v
variation is between UFG
and CG regions at the same location. In both locations, the H
v
of CG regions increased
more than that of the UFG regions as the applied stress increased, especially after the
yielding (265 MPa). This implies that more strain hardening occurred in CG regions
compared to the UFG regions during the plastic deformation. This finding is consistent
with the strain heterogeneity revealed by the DIC results. The difference in magnitude of
the strain hardening between UFG and CG regions is attributed to the greater increase in
dislocation density within CG regions, as well as the greater dislocation mobility. The
density of dislocations significantly increased in CG regions after yielding, and the
observed hardening arose from tangles of dislocations gliding on slip systems. The
second comparison is between the two locations for UFG and CG. In this case, the
increment in H
v
of CG near the notches was greater than at the central part. Conversely,
the variation in UFG H
v
at the two locations was negligible. The stress concentration
adjacent to the notch tips gave rise to these variations in both UFG and CG regions.
The tensile fracture surface of the as-forged material is shown in Fig. 5.13. In
general, two different fracture surface morphologies were apparent; (i) a relatively brittle
fracture surface comprised of small dimples and (ii) a ductile fracture surface with large
dimples. The former fracture surface (i) arose from the fracture of UFG regions, a
manifestation of the fact that the fracture becomes more brittle as grain size decreases.
This phenomenon is also attributed to reduced mobility of dislocations. In UFG regions,
156
Fig. 5.12 Vickers hardness variation measured from both UFG and CG regions at two
locations (near the notches and between the notches) with respect to the nominal tensile
stress.
157
Fig. 5.13 SEM micrographs of tensile fracture surfaces of the notched specimen showing
two different surface morphologies comprised of small and large dimples, when viewed
from directions (a) normal to the surface and (b) tilted from the surface.
158
dislocation glide distances were restricted by grain boundaries, limiting dislocation
plasticity. The second fracture surface (ii) arose from the fracture of CG regions. The
dimple size was typically between 1 and 4 µm, roughly equivalent to the CG size as
shown in Fig. 5.13(b). In this case, the large surface dimples indicate that the CG regions
underwent significantly more plastic strain than UFG regions. This can account for the
nature of enhanced ductility in bimodal materials compared to 100% NC/UFG materials.

5.2.4. Conclusions
A DIC technique was successfully applied to measure surface displacements
during tensile deformation experiments. Strain inhomogeneity was observed between
UFG and CG regions when the displacements were correlated in micro-strain range. The
strain localization in CG regions stemmed from the greater dislocation mobility in these
regions compared to UFG regions. This conclusion was supported by hardness
measurements, which revealed different degrees of strain hardening during plastic
deformation. The hardness variations between CG and UFG regions indicated a greater
strain hardening effect in CG regions, an observation that was attributed to the increase in
dislocation density and interactions. The strain mapping by DIC in this study,
accompanied by dynamic observations, provides insights and understanding of
deformation behavior of multi-scale materials.

159
References
[1] H. Gleiter, Prog. Mater. Sci. 33 (1989) 223.

[2] J.R. Weertman and R.S. Averback, In: A.S. Edelstein et al., editors,
Nanomaterials: Synthesis, Properties and Applications, Bristol, UK: Institute of
Physics Publishing (1996) pp. 323 −45.

[3] A.P. Newbery, B.Q. Han, E.J. Lavernia, C. Suryanarayana and J.A. Christodoulou,
In: J.R. Groza et al., editors, Materials Processing Handbook, Boca Raton, Fl:
CRC Press (2007) pp. 13-1 −13-28.

[4] E.Q. Hall, Proc. Phys. Soc. B64 (1951) 747.

[5] N.J. Petch, J. Iron Steel Inst. 174 (1953) 25.

[6] C. Suryanarayana, Prog. Mater. Sci. 46 (2001) 1.

[7] D.B. Witkin and E.J. Lavernia, Prog. Mater. Sci. 51 (2006) 1.

[8] R.Z. Valiev, R.K. Islamgaliev and I.V. Alexandrov, Prog. Mater. Sci. 45 (2000)
103.

[9] P.G. Sanders, G.E. Fougere, L.J. Thompson, J.A. Eastman and J.R. Weertman,
Nanostruct. Mater. 8 (1997) 243.

[10] U. Erb, Nanostruct. Mater. 6 (1995) 533.

[11] Y. Wang, M. Chen, F. Zhou and E. Ma, Nature 419 (2002) 912.

[12] G. He, J. Eckert, W. Loser and L. Schultz, Nature Mater. 2 (2002) 33.

[13] V.L. Tellkamp, A. Melmed and E.J. Lavernia, Metall. Mater. Trans. 32A (2001)
2335.

[14] K.M. Youssef, R.O. Scattergood, K.L. Murty and C.C. Koch, Scripta Mater. 54
(2006) 251.

[15] D. Witkin, Z. Lee, R. Rodriguez, S. Nutt and E. Lavernia, Scripta Mater. 49 (2003)
297.

[16] B.Q. Han, Z. Lee, D. Witkin, S. Nutt and E.J. Lavernia, Metall. Mater. Trans. 36A
(2005) 957.

160
[17] Z. Lee, D.B. Witkin, V. Radmilovic, E.J. Lavernia and S.R. Nutt, Mater. Sci. Eng.
A 410 −411 (2005) 462.

[18] Z. Lee, J. Lee, E.J. Lavernia and S.R. Nutt, In: P.M. Anderson et al., editors,
Nanoscale Materials and Modeling-Relations among Processing, Microstructure,
and Mechanical Properties, Pittsburgh, PA: Mater. Res. Soc. (MRS) Symp. Proc.
821 (2004) P9.11.1.

[19] H.W. Chan, Mater. Design 9 (1988) 355.

[20] A.P. Newbery, B. Ahn, P. Pao, S.R. Nutt and E.J. Lavernia, Adv. Mater. Res.
29−30 (2007) 21.

[21] T.C. Chu, W.F. Ranson, M.A. Sutton and W.H. Peters, Exp. Mech. 25 (1985) 232.

[22] M.A. Sutton, M. Cheng, W.H. Peters, Y.J. Chao and S.R. McNeill, Image Vision
Comput. 4(3) (1986) 143.

[23] H.A. Bruck, S.R. McNeill, M.A. Sutton and W.H. Peters, Exp. Mech. 29 (1989)
261.

[24] B.W. Smith, X. Li and W Tong, Exp. Tech. 22 (1989) 19.

[25] M.R. James, W.L. Morris and B.N. Cox, Exp. Mech. 30 (1990) 60.

[26] B. Ahn, A.P. Newbery, E.J. Lavernia and S.R. Nutt, Mater. Sci. Eng. A 463 (2007)
61.
161
Chapter 6. Al Nanocomposites Reinforced with B
4
C Particulates

6.1. Multi-scale concept
6.1.1. Microstructure
Metal matrix composites (MMCs) have been reported to have attractive physical
and mechanical properties, such as high specific modulus, superior strength and
improved thermal stability [1,2], for numerous applications in different fields. Particulate
reinforced MMCs have received great attention because of their ease of fabrication, lower
cost and isotropic properties. In the field of bulk nanostructured materials (BNMs),
besides grain size strengthening by Hall-Petch relationship, the introduction of stiffer and
stronger ceramic particles into the Al alloys is an alternative strengthening approach that
leads to metal matrix composites [3]. B
4
C, a good reinforcing candidate for the aluminum
based composites, ranks third in hardness, just after diamond and cubic boron nitride, and
possesses a low density of 2.51 g/cm
3
(even less than that of Al) [4]. To be a high-quality
composite, a good interface between the reinforcement and the matrix is desired to
achieve a more effective strengthening from the reinforcement [3]. Cryomilling of a
mixture of ceramics and Al has been demonstrated as a feasible method to synthesize a
composite powder with the ceramic particles uniformly distributed in the nanocrystalline
(NC) Al, with a good interface between them [5].
However, the increase in strength is always accompanied with some sacrifice in
ductility for NC metals in comparison to their microcrystalline counterparts [6,7]. The
introduction of the ceramic particles will further reduce the ductility of the NC materials,
162
although MMCs provide higher strength. It has been demonstrated by a previous work
that NC Al with B
4
C particles fractured in the elastic deformation range, regardless of its
high strength [5]. Several approaches have been evaluated to improve the ductility for the
NC materials [7–9], including the introduction of course-grained (CG) Al into NC Al
[10–12]. In the present study, a novel approach analogous to ductile phase toughening
was used to achieve improved ductility. Al alloy powders and B
4
C particles were
mechanically milled in liquid nitrogen slurry to form a composite powder with B
4
C
particles in NC Al. The cryomilled composites powders were blended with CG Al
powders, as shown in Fig. 6.1. The blended powders were consolidated using cold
isostatic pressing (CIP) and extrusion. The resultant bulk composite is a complicated
materials system consisting of three phases, 10 wt.% B
4
C, 50 wt.% CG Al and the
balanced NC Al. It is expected that B
4
C particles and cryomilled NC regions contribute
to high strength, while CG regions provide the dislocation-mediated ductility.
The microstructures of the trimodal composite along the extrusion and transverse
directions are shown in Fig. 6.2. The CG regions appear brighter tone in darker tone NC
matrix with black B
4
C particles in the optical micrographs shown in Fig. 6.2. In general,
the B
4
C particles are distributed homogeneously in the NC regions. CG regions are
elongated along the extrusion direction, forming a band structure, as shown in Fig. 6.2(a).
The CG bands have a dimension of 1 −20 µm in diameter and 10 −100 µm in length.
Figure 6.3 shows TEM bright filed images with a selected area diffraction (SAD) patterns.
The B
4
C particles are uniformly distributed in the NC matrix with a size range of
submicron to micron, as shown in Fig. 6.3(a), and the NC regions are alternately stacked
with the CG regions. The SAD pattern in Fig. 6.3(b) taken from the interface of a
163
Fig. 6.1 Schematic of synthesis of a multi-scale trimodal composite, incorporating CG
material into cryomilled NC and B
4
C to improve ductility and toughness.
164
Fig. 6.2 Optical micrographs of the as-extruded trimodal composite, when viewed from
(a) normal to the extrusion direction and (b) transverse to the extrusion direction. CG Al
regions appear brighter tone in gray NC Al matrix containing black B
4
C particles.
165
Fig. 6.3 TEM micrographs of the as-extruded trimodal composite: (a) bright field image
showing global view from normal to the extrusion direction, (b) selected area diffraction
(SAD) patterns of NC and B
4
C from their interface, and (c) and (d) bright field images
showing B
4
C particles primarily residing in NC Al matrix without direct contact with CG
regions.
166
B
4
C particle and NC region shows a series of uniform ring patterns (NC region) and a set
of spot patterns (B
4
C particle). No spots for other phases are detected from the SAD
patterns, indicating there are no interfacial reaction products between the B
4
C and the NC
region. The average grain size of the NC regions was ~165 nm, determined statistically
from the TEM images. B
4
C particles are generally surrounded by NC regions without
direct contact with CG regions, as shown in Fig. 6.3(c) and (d). This structure contributes
to the extremely high strength of the trimodal composite. Dislocations were observed in
CG regions adjacent to B
4
C particles, as shown in Fig. 6.3(c) and (d), which reveal
significant dislocation activities occurred in CG regions providing the dislocation-
mediated plasticity.

6.1.2. Compressive behavior
The yield strength and compressive strain-to-failure of the nanocomposite are
shown in Table 6.1, and compared with conventional 100% CG materials and a
cryomilled 100% NC material. The room temperature yield strength was as high as 1065
MPa, which represents more than a three-fold increase over the yield strength of the
strongest commercial available conventional equivalent (5083-H343) and more than a
two-fold increase over cryomilled 100% NC Al 5083 [13,14]. It is notable that the
introduction of 10% of B
4
C can greatly improve the yield strength. The composite
possesses very limited strain-to-failure (0.8%), as shown in Table 6.1. The composite
annealed at 450 ºC has a higher strain-to-failure (2.5%) with a small reduction in the
yield strength, as shown in Table 6.1, and the stress-stain curve for the annealed
composite is shown in Fig. 6.4. It exhibited an elastic-perfectly plastic deformation
167

Table 6.1 Room temperature mechanical properties of the trimodal nanocomposite,
compared with a conventional CG material and a cryomilled NC material.

Materials
Yield strength
(MPa)
Tensile elongation
(%)
Compressive strain
to failure
(%)
Al 5083-O [17] 124 16 −
NC Al 5083 [17] 334 8.4 −
Composite 10/50 [13] 1065 − 0.8
Composite 10/50-A [13] 1058 − 2.5

Note: Al 5083-O stands for the Al 5083 with O type of heat treatment, Composite 10/50
represents the composite with 10 wt.% B
4
C and 50 wt.% CG, and composite 10/50-A stands
for the composite 10/50 after annealing at 723 K for 2 hours.
168
Fig. 6.4 True stress-strain curve for the annealed trimodal composite tested at ambient
temperature at a strain rate of 10
-3
s
-1
.
169
behavior, as indicated by the flat stress plateau after yielding, i.e. the absence of work
hardening. The strain is limited by cavitations, and the presence of reinforcing particles
increases the extent of cavitations by constraining the grain boundary sliding and
providing stress concentrations at the interface. The pre-existing residual stress at the
interface will yield a higher stress concentration at the interface, resulting in a quick
failure and smaller elongation. The annealing released the residual stress accumulated
from the fabrication processes, relaxed the strained structures in the CG regions and
allowed for grain growth in the NC Al, resulting in moderate decrease in yield strength
and improved strain-to-failure. The cryomilled Al alloy is thermally stable, due to the
presence of nano-sized aluminum oxide and nitride dispersoids that result from the
cryomilling [15]. It was found that the annealing of the cryomilled Al-7.5Mg did not
change in the yield strength, although the elongation increased [16]. This indicates that,
in the present study, the increase in strain-to-failure after annealing is due to the release
of residual stress and structural relaxation in the CG regions.
Figure 6.5 shows a TEM bright field image and SAD patterns of the as-extruded
nanocomposite. The interfaces between B
4
C particles and NC regions, of crucial to the
mechanical properties of the composite, are very clear without other phases or voids, as
observed in the TEM micrograph in Fig. 6.5. Good interfaces allow the load to transfer
from the matrix to the reinforcements more effectively and lead to a stronger composite.
In the present trimodal composite, the applied load on the CG Al can be effectively
transferred to the NC Al, because a strong interface between the CG Al and the NC Al
can easily be developed during consolidation due to the same chemical composition in
both regions. In addition, the load applied on the NC Al can be further transferred to the
170
Fig. 6.5 TEM bright field image showing clear interface between B
4
C particles and NC
matrix as well as corresponding SAD patterns from NC regions and a B
4
C particle,
respectively.
171
stronger B
4
C particles, because of clean metallurgical interfaces between the NC regions
and the B
4
C particles formed during cryomilling. As a consequence, most of the applied
load is sustained by the NC regions and the B
4
C particles and only a small fraction of the
load remains in the CG regions. This load transfer sequence provides the justification for
observing high yield strength in the trimodal composite.
The high temperature properties of the trimodal composite were investigated
under uniaxial compression at elevated temperatures. Figure 6.6(a) shows the stress-strain
curves at elevated temperatures with a flat plateau after yielding. Homogeneous
compressive deformation was observed for each temperature shown in Fig. 6.6(a). The
yield strength decreased rapidly when the test temperature was increased to 473 K (55%
T
m
of 5083 Al), followed by a relatively slow decrease in yield strength at higher
temperatures, as shown in Fig. 6.6(b). Composite materials normally lose their strength at
elevated temperatures, because the strengthening mechanisms that operate at lower
temperatures are relaxed at higher temperatures. In the trimodal composite in the present
study, the relaxation of the strengthening that results from the presence of the B
4
C may
be retarded by its surrounding NC Al, due to the thermal stability of the NC Al. Also, the
grain size strengthening is still effective at higher temperatures, resulting in the retention
of a high yield strength at elevated temperatures, which allows this composite to be an
excellent candidate for high temperature applications.

6.2. Microstructural evolution during cryomilling
Ball milling at cryogenic temperature, cryomilling, was successfully employed to
fabricate particulate B
4
C reinforced Al matrix nanocomposite powders. The composite
172
Fig. 6.6 (a) True stress-strain curves for the trimodal composite tested at various elevated
temperatures and (b) corresponding temperature dependence of the yield strength.
173
powders were cryomilled for different milling times in order to investigate the
microstructural evolution during cryomilling. The microstructures were characterized to
reveal the formation mechanism of the nanocomposite. The morphology and size of the
cryomilled powders, the size and distribution of B
4
C particles and the dimension of the
Al grains were discussed with respect to the milling time.

6.2.1. Motivation
The presence of the B
4
C in this ductile-brittle (Al/B
4
C) material system changes
the fracturing mode during cryomilling, but has no significant influence on the ultimate
degree of grain refinement in the Al matrix. It is necessary to better understand the
microstructural evolution during cryomilling, so that composite powders with specific
structural characteristics can be achieved, and the properties of a bulk composite
consolidated from these powders can be optimized. Thus, the purpose of the current study
was to investigate the microstructural evolution for both the B
4
C reinforcement and the
Al alloy matrix during cryomilling and the mechanism by which a nanostructured
composite with uniformly distributed B
4
C was formed.

6.2.2. Experimental procedures
Commercial gas-atomized Al 5083 alloy powder (90 wt.%) and particulate B
4
C
(10 wt.%) were blended and then cryomilled under liquid nitrogen atmosphere.
Cryomilling was carried out at a temperature of –180 ºC using a Szegvari attritor with a
ball-to-powder ratio of 32 : 1 and a rotation speed of 180 rpm, from which cryomilled
powders milled for different times were collected in order to study the microstructural
174
evolution of the Al matrix and B
4
C reinforcement and the mechanism by which the B
4
C
was incorporated into the matrix. 0.2 wt.% stearic acid was added into the milling
chamber as a process control agent (PCA) to prevent the severe adhesion of the
aluminum onto the chamber and milling balls.
XRD measurements were performed using Cu K
α
radiation in a Scintag XDS 2000 X-ray
diffractometer. The grain size of the Al was calculated on the basis of X-ray peak
broadening [18,19] and measured directly from TEM micrographs. Microstructural
characterization was conducted using an optical microscope, scanning electron
microscope (Philips XL 30) and a transmission electron microscope (Philips EM420
TEM) to investigate the microstructural evolution of the milled powders. The particle
size of the composite powders and the B
4
C reinforcement were statistically calculated
from optical micrographs on the cross-sections, using image processing software (Fovea
Pro, Reindeer Graphics).

6.2.3. Experimental results and discussion
The morphology of the as-received Al 5083 powder and the composite powder
cryomilled for 8 hours is shown in Fig. 6.7. The as-received powder was gas atomized
with a spherical shape, as shown in Fig. 6.7(a). During cryomilling, the particles get
flattened, adhere together, and develop a flake-like shape. The shape gradually becomes
fine equiaxed particles, and the surface of these particles was rough, as shown in Fig.
6.7(b). Figure 6.8 shows optical micrographs demonstrating the evolution of particle
morphology with respect to the milling time. Only a few of the flake-like particles could
be found in the powder cryomilled for 6 hours. After 6 hours of cryomilling, further
175
Fig. 6.7 SEM images showing the morphology of the powders: (a) as-received gas-
atomized Al 5083 powder and (b) Al/B
4
C composite powder cryomilled for 8 hours.
176
Fig. 6.8 Optical micrographs of Al/B
4
C composite powders showing cross sectional
morphology of powders cryomilled for (a) 1 hour, (b) 2 hours, (c) 4 hours, (d) 6 hours,
(e) 8 hours, and (f) 10 hours, respectively.
177
milling resulted in very little change in the powder morphology. The size evolution of the
composite powder particle, as well as the change of the particle size distribution, is
shown in Fig. 6.9(a). In the early stage of cryomilling, the composite particle size became
larger and the size distribution became broader. During cryomilling, particles undergo
repeated cold welding and fracturing. When the rate of cold welding exceeds that of
fracturing, the particle size tends to increase; otherwise, the particles become smaller in
size. After cryomilling for more than 1 hour, the size of the particles began to decrease
and the size distribution became narrower. After 6 hours of cryomilling, the particle size
approaches a constant value and the size distribution is narrow.
Figure 6.9(b) shows the size evolution of the B
4
C reinforcements, as well as the
size deviation. The B
4
C reinforcement does not exhibit significant size reduction during
milling, compared to the composite powder particles. It decreased half the size after 10
hours of cryomilling; from ~2.5 µm down to ~1.1 µm. Because the material used for
cryomilling is a ductile-brittle (Al/B
4
C
p
) system, B
4
C particles were trapped and
embedded in the ductile Al alloys. These Al 5083 particles behave as a cushion, damping
the forces acting on the B
4
C, and, thus, the remaining energy is not sufficient to fracture
the hard B
4
C. The small change in B
4
C size during cryomilling is also attributed to the
extremely high hardness and impact resistance of B
4
C. Figure 6.10 shows optical
micrographs demonstrating the B
4
C particle distribution within the Al powder with
respect to the milling time. There were numerous B
4
C particles on the surface of the Al
powder cryomilled for 1 h, as shown in Fig. 6.10(a), and few B
4
C particles were present
inside the Al matrix for the 2 hours cryomilled powder, as shown in Fig. 6.10(b).
178
Fig. 6.9 Variation in the mean particle size during cryomilling with respect to the milling
time, determined statistically from optical micrographs: (a) the Al/B
4
C composite powder
and (b) B
4
C particles. The error bars represent the standard deviation.
179
Fig. 6.10 Optical micrographs of Al/B
4
C composite powders showing distribution of B
4
C
particles within the NC Al matrix: powders cryomilled for (a) 1 hour, (b) 2 hours, (c) 4
hours, (d) 6 hours, (e) 8 hours, and (f) 10 hours, respectively.
180
However, after 6 hours of cryomilling, most of the B
4
C is dispersed uniformly within the
Al matrix in Fig. 6.10(d), (e) and (f).
The variation of mean grain size in cryomilled NC regions measured by XRD
peak broadening is plotted in Fig. 6.11 as a function of the milling time. The average size
decreased exponentially to a size of approximately 25 nm after 10 hours of cryomilling.
The relationship between the average grain size of the Al matrix and the milling time
(derived from the results in Fig. 6.11) can be expressed as d = Ke
-t/3
, where d is the mean
grain size of the Al, t is the cryomilling time, and K is a constant. Because cryomilling is
a stochastic process, it should be noted that there is actually a distribution in the size of
the Al grains. TEM bright field images for the composite powder cryomilled for 2 hours
and 8 hours are also shown in Fig. 6.11. The microstructure of NC Al regions in the
powder cryomilled for 2 hours was not homogeneous, consisting of equiaxed grains and
elongated (or lamellar) grains, as shown in Fig. 6.11. However, the Al grains in the
powder cryomilled for 8 hours were primarily equiaxed.
The distributions in characteristic dimensions (for both the length and the width)
of the Al grain in the powder cryomilled for 8 hours were quite broad, with average
values of 114 nm in length and 47 nm in width. However, the grain size of the Al matrix
determined from XRD was 37±4 nm for the powder, which is much smaller than the
average length of the elongated grains (114 nm), but close to their average width (47 nm).
The XRD evaluates only the continuous reflecting region. The existence of dislocation
walls, low or high-angle grain boundaries (HGBs), or other defects such as nano-twins
[20,21] can disrupt the continuous reflecting regions, but might not be clearly
distinguished using traditional TEM, resulting in the apparent average grain size in the
181
Fig. 6.11 Variation in the mean grain size of Al 5083 during cryomilling with respect to
the milling time and corresponding TEM bright field images showing microstructure of
powders cryomilled for 2 hours and 8 hours.
182
composite powder milled for 2 hours appearing smaller when determined by XRD than
that observed with TEM. In contrast, the mean size of the Al grains in the 8 hours
cryomilled powder measured directly from TEM images was 28 nm, which is consistent
with the average value determined with XRD (25 nm) shown in Fig. 6.11, although some
grains were as small as 10 nm and larger than 80 nm. For the powder cryomilled for 8
hours, almost all of the dislocation cells and subgrains have evolved into individual
grains separated by HGBs. These HGBs, which act as defects to disrupt the continuous
reflecting regions in XRD, were distinguishable with conventional TEM, leading to good
agreement in average grain size between the XRD and the TEM results.

6.2.4. Conclusions
Composites powders with a ductile-brittle system (Al/B
4
C
p
) were cryomilled and
uniform distribution of B
4
C particles and good Al/B
4
C interfaces were achieved. The
particles underwent repeated cold welding and fracturing during cryomilling. As a result,
both large and small composite particles were driven to an intermediate size with a
narrower size distribution. The size of the B
4
C particles did not change much and
remained in the micron range, due to the extreme hardness and impact resistance of B
4
C.
The grains in the Al matrix evolved into the nanoscale regime through grain subdivision,
which is governed by the accumulated strain level. The CG Al subdivided into micro-
bands very quickly at the early stage of cryomilling and was gradually refined to
randomly oriented equiaxed grains with a minimum grain size of 23 nm. After
cryomilling for 6 hours, steady state had been achieved for size of B
4
C particles and
composite powders, and the B
4
C particles were uniformly distributed in the NC Al matrix.
183
References
[1] T.W. Clyne and P.J. Withers, In: An Introduction to Metal Matrix Composites,
New York, NY: Cambridge University Press (1995).

[2] S.C. Tjong and Z.Y. Ma, Mater. Sci. Eng. R 29 (2000) 49.

[3] D.J. Lloyd, Int. Mater. Rev. 39 (1994) 1.

[4] J.F. Shackelford and W. Alexander, In: Materials Science and Engineering
Handbook, 3rd Edition, Boca Raton, FL: CRC Press (2001).

[5] J.M. Schoenung, J. Ye, J. He, F. Tang and D. Witkin, Mater. Forum 29 (2005) 123.

[6] C.C. Koch, Scripta Mater. 49 (2003) 657.

[7] E. Ma, Scripta Mater. 49 (2003) 663.

[8] G. He, J. Eckert, W. Loser and L. Schultz, Nature Mater. 2 (2002) 33.

[9] Y. Wang, M. Chen, F. Zhou and E. Ma, Nature 419 (2002) 912.

[10] D. Witkin, Z. Lee, R. Rodriguez, S. Nutt and E. Lavernia, Scripta Mater. 49 (2003)
297.

[11] B.Q. Han, Z. Lee, D. Witkin, S. Nutt and E.J. Lavernia, Metall. Mater. Trans. 36A
(2005) 957.

[12] Z. Lee, D.B. Witkin, V. Radmilovic, E.J. Lavernia and S.R. Nutt, Mater. Sci. Eng.
A 410–411 (2005) 462.

[13] J. Ye, B.Q. Han, Z. Lee, B. Ahn, S.R. Nutt and J.M. Schoenung, Scripta Mater. 53
(2005) 481.

[14] V.L. Tellkamp, A. Melmed and E.J. Lavernia, Metall. Mater. Trans. 32A (2001)
2335.

[15] F. Zhou, J. Lee, S. Dallek and E.J. Lavernia, J. Mater. Res. 16 (2001) 3451.

[16] B.Q. Han, Z. Lee, S.R. Nutt, E.J. Lavernia and F.A. Mohamed, Metall. Mater.
Trans. 34A (2003) 603.

[17] J.R. Davis, ASM Specialty Handbook: Aluminum and Aluminum Alloys, Materials
Park, OH: ASM International (1993) p. 675.

184
[18] H.P. Klug and L.E. Alexander, X-ray Diffraction Procedures: For Polycrystalline
and Amorphous Materials, 2nd Edition, New York, NY: John Wiley & Sons (1974)
pp. 634–682.

[19] J. He, J. Ye, E.J. Lavernia, D. Matejczyk, C. Bampton and J.M. Schoenung, J.
Mater. Sci. 39 (2004) 6957.

[20] J. He and E.J. Lavernia, J. Mater. Res. 16 (2001) 2724.

[21] X.Z. Liao, J.Y. Huang, Y.T. Zhu, F. Zhou and E.J. Lavernia, Phil. Mag. 83 (2003)
3065.
185
Chapter 7. Suggestions for Future Works

Future works will be devoted to optimize the properties of various cryomilled
nanocrystalline materials including the bimodal and trimodal nanocomposite materials.
The objectives include: to optimize consolidation parameters, to determine the processing
parameters in the bimodal and trimodal materials regarding the fraction and distribution
of coarse grains, and to design the range of mechanical properties for practical
applications. In particular, extensive microstructural analyses using multiple analytical
techniques as well as mechanical tests are required as follows.

Investigate microstructural evolution as a function of process parameters, such as
consolidation methods, temperatures, or pressures, to optimize properties.
Perform in-situ micro-tensile/compression tests in a scanning electron microscope
(SEM) to investigate the fracture process of bimodal Al alloys.
Investigate evolution of strain distribution as a function of the fraction of coarse
grains in bimodal Al alloys.
Measure surface displacements in trimodal Al nanocomposite subjected to micro-
straining experiments. Measure local strains as functions of tensile and compressive
macro-strains and develop strain maps for specimens with different geometries
responding to tensile and compressive loads.
Analyze the deformation for different trimodal alloys and relate deformation to
reinforcement content and coarse grain fraction.
186
Perform nanoindentation experiments on trimodal Al alloys to explore strain rate
effects as a function of grain size, and hardening in the vicinity of ceramic particles.
Conduct compression experiments on nano-pillars fabricated by focused ion beam
(FIB) machining to probe local mechanical properties in trimodal microstructure.
Determine evolution of damage in deformed trimodal Al alloys as function of strain
history and strain rate using FIB cut sections.
187
Bibliography

Ahn B., Newbery A.P., Lavernia E.J. and Nutt S.R., Mater. Sci. Eng. A 463 (2007) 61.

Armstrong R.W., Nabarro F.R.N. and Conrad H., In: Ovid’ko I. et al., editors,
Mechanical Properties of Nanostructured Materials and Nanocomposites, Pittsburgh,
PA: Mater. Res. Soc. (MRS) Symp. Proc. 791 (2003) 69.

Artz E., Acta Mater. 46 (1998) 5611.

Birringer R., Mater. Sci. Eng. A 117 (1989) 33.

Bruck H.A., McNeill S.R., Sutton M.A. and Peters W.H., Exp.Mech. 29 (1989) 261.

Carsley J.E., Fisher A., Milligan W.W. and Aifantis E.C., Metall. Mater. Trans. 29A
(1998) 2261.

Cavaliere P., Physica B 403 (2008) 569.

Champion Y., Guérin-Mailly S., Bonnentien J.-L. and Langlois P., Scripta Mater. 44
(2001) 1609.

Chan H.W., Mater. Design 9 (1988) 355.

Cheng S., Ma E., Yang Y.M., Kecskes L.J., Youssef K.M., Koch C.C., Trociewitz U.P.
and Han K., Acta Mater. 53 (2005) 1521.

Chokshi A.H., Rosen A., Karsh J. and Gleiter H., Scripta Metall. 23 (1989) 1679.

Chu T.C., Ranson W.F., Sutton M.A. and Peters W.H., Exp. Mech. 25 (1985) 232.

Clyne T.W. and Withers P.J., An Introduction to Metal Matrix Composites, New York,
NY: Cambridge University Press (1995).

Conrad H. and Narayan J., Acta Mater. 50 (2002) 5067.

Conrad H. and Narayan J., Scripta Mater. 42 (2000) 1025.

Conrad H., Metall. Mater. Trans. 35A (2004) 2541.

Correlated Solutions, Inc., Columbia, SC, http://www.correlatedsolutions.com.

188
Davis J.R., ASM Specialty Handbook: Aluminum and Aluminum Alloys, Materials Park,
OH: ASM International (1993) p. 675.

Derlet P.M. and Van Swygenhoven H., Scripta Mater. 47 (2002) 719.

Derlet P.M., Hasnaoui A. and Van Swygenhoven H., Scripta Mater. 49 (2003) 629.

Dobatkin S.V., Bastarache E.N., Sakai G., Fujita T., Horita Z. and Langdon T.G., Mater.
Sci. Eng. A 408 (2005) 141.

Ebrahimi F., Zhai Q. and Kong D., Scripta Mater. 39 (1998) 315.

El-Sharik A.M., Erb U., Palumbo G. and Aust K.T., Scripta Metall. Mater. 27 (1992)
1185.

Erb U., El-Sherik A.M., Palumbo G. and Aust K.T., Nanostruct. Mater. 2 (1993) 383.

Erb U., Nanostruct. Mater. 6 (1995) 533.

Fecht H.J., Nanostruct. Mater. 6 (1995) 33.

Ficher-Cripps A.C., In: Ling F.F., editor, Nanoindentation, Verlag, NY: Springer-Verlag
(2002) p. 23.

Fougere G.E., Weertman J.R., Siegel R.W. and Kim S., Scripta Metall. Mater. 26 (1992)
1879.

Frederiksen S.L., Jacobsen K.W. and Schiøtz J., Acta Mater. 52 (2004) 5019.

Froseth A., Van Swygenhoven H. and Derlet P.M., Acta Mater. 52 (2004) 2259.

Fu H.H., Benson D.J. and Meyers M.A., Acta Mater. 49 (2001) 2567.

Furukawa M., Horita Z., Nemoto M. and Langdon T.G., J. Mater. Sci. 36 (2001) 2835.

Gifkins R.C., Metall. Trans. A (1976) 1225.

Gilman P.S. and Benjamin J.S., Ann. Rev. Mater. Sci. 13 (1983) 279.

Gleiter H., Acta Mater. 48 (2000) 1.

Gleiter H., Nanostruct. Mater. 1 (1992) 1.

Gleiter H., Prog. Mater. Sci. 33 (1989) 223.

189
Gray G.T., Lowe T.C., Cady C.M., Valiev R.Z. and Aleksandrov I.V., Nanostruct. Mater.
9 (1997) 477.

Greer A.L., Nature 368 (1994) 688.

Groza J.R., In: Koch C.C., editor, Nanostructured Materials: Processing, Properties, and
Applications, 2
nd
Edition, Norwich, NY: William Andrew Publishing (2007) pp. 173 −233.

Gryaznov V.G., Gutkin M.Y., Romanov A.E. and Trusov L.I., J. Mater. Sci. 28 (1993)
4359.

Hahn H., Mondal P. and Padmanabhan K.A., Nanostruct. Mater. 9 (1997) 603.

Hall E.O., Proc. Phys. Soc. B64 (1951) 747.

Han B.Q., Huang J., Zhu Y.T. and Lavernia E.J., Adv. Eng. Mater. 8 (2006) 945.

Han B.Q., Huang J.Y., Zhu Y.T. and Lavernia E.J., Scripta Mater. 54 (2006) 1175.

Han B.Q., Lee Z., Nutt S.R., Lavernia E.J. and Mohamed F.A., Metall. Mater. Trans. 34A
(2003) 603.

Han B.Q., Lee Z., Witkin D., Nutt S. and Lavernia E.J., Metall. Mater. Trans. 36A (2005)
957.

Hasnaoui A., Derlet P.M. and Van Swygenhoven H., Acta Mater. 52 (2004) 2251.

Hasnaoui A., Van Swygenhoven H. and Derlet P.M., Acta Mater. 50 (2002) 3927.

Hayes R.W., Berbon P.B. and Mishra R.S., Metall. Mater. Trans. 35A (1998) 3855.

He G., Eckert J., Loser W. and Schultz L., Nature Mater. 2 (2002) 33.

He J. and Lavernia E.J., J.Mater. Res. 16 (2001) 2724.

He J., Ye J., Lavernia E.J., Matejczyk D., Bampton C. and Schoenung J.M., J. Mater. Sci.
39 (2004) 6957.

Huang Y.H., Quan C., Tay C.J. and Chen L.J., Opt. Eng. 44(8) (2005) 087011-1.

Humphreys F.J., Prangnell P.B., Bowen J.R., Gholinia A. and Harris C., Phil. Trans. R.
Soc. Lond. A 357 (1999) 1663.

Iyer R.S., Frey C.A., Sastry S.M.L., Aller B.E. and Buhro W.E., Mater. Sci. Eng. A 264
(1999) 210.
190
James M.R., Morris W.L. and Cox B.N., Exp. Mech. 30 (1990) 60.

Jia D., Ramesh K.T. and Ma E., Acta Mater. 51 (2003) 3495.

Jia D., Ramesh K.T. and Ma E., Scripta Mater. 42 (2000) 73.

Jia D., Wang Y.M., Ramesh K.T., Ma E., Zhu Y.T. and Valiev R.Z., Appl. Phys. Lett. 79
(2001) 611.

Jiang B. and Weng G.J., J. Mech. Phys. Solids 52 (2003) 1125.

Klimanek P., Klemm V., Romanov A.E. and Seefeldt M., Adv. Eng. Mater. 3 (2001) 877.

Klug H.P. and Alexander L.E., X-ray Diffraction Procedures: For Polycrystalline and
Amorphous Materials, 2
nd
Edition, New York, NY: John Wiley & Sons (1974) pp.
634−682.

Koch C.C. and Cho Y.S., Nanostruct. Mater. 1 (1992) 207.

Koch C.C., J. Metastable. Nanocryst. Mater. 18 (2003) 9.

Koch C.C., Morris D.G., Lu K. and Inoue A., MRS Bull. 24 (1999) 54.

Koch C.C., Nanostruct. Mater. 9 (1997) 13.

Koch C.C., Scripta Mater. 49 (2003) 657.

Kong Q.P., Cai B., Lu L. and Lu K., Def. Diff. Met. 188 (2001) 45.

Kumar K.S., Van Swygenhoven H. and Suresh S., Acta Mater. 51 (2003) 5743.

Langdon T.G., Acta Metall. Mater. 42 (1994) 2437.

Langdon T.G., J. Mater. Sci. 16 (1981) 2613.

Lee Z., Lee J., Lavernia E.J. and Nutt S.R., In: Anderson P.M. et al., editors, Nanoscale
Materials and Modeling-Relations Among Processing, Microstructure, and Mechanical
Properties, Pittsburgh, PA: Mater. Res. Soc. (MRS) Symp. Proc. 821 (2004) P9.11.1.

Lee Z., Rodriguez R., Hayes R.W., Lavernia E.J. and Nutt S.R., Metall. Mater. Trans.
34A (2003) 1473.

Lee Z., Witkin D.B., Radmilovic V., Lavernia E.J. and Nutt S.R., Mater. Sci. Eng. A
410−411 (2005) 462.

191
Legros M., Elliott B.R., Rittner M.N., Weertman J.R. and Hemker K.J., Philos. Mag. A80
(2000) 1017.

Liao X.Z., Huang J.Y., Zhu Y.T., Zhou F. and Lavernia E.J., Phil. Mag. 83 (2003) 3065.

Lin Z.-R., Chokshi A.H. and Langdon T.G., J. Mater. Sci. 23 (1998) 2712.

Lloyd D.J, Int. Mater. Rev. 39 (1994) 1.

Lu K., Mater. Sci. Eng. R 16 (1996) 161.

Lu L., Li S.X. and Lu K., Scripta Mater. 45 (2001) 1163.

Luton M.J., Jayanth C.S., Disko M.M., Matras S. and Vallone J. In: McCandlish L.E. et
al., editors, Multicomponent Ultrafine Microstructures, Pittsburgh, PA: Mater. Res. Soc.
Symp. Proc. (MRS) 132 (1989) pp. 79 −86.

Ma E., Scripta Mater. 49 (2003) 663.

Malow T.R. and Koch C.C., Acta Mater. 46 (1998) 6459.

Malow T.R. and Koch C.C., In: Bourell D.L., editor, Synthesis and Processing of
Nanocrystalline Powder, Warrendale, PA: The Minerals, Metals & Materials Society
(TMS) (1996) p. 33.

Malow T.R., Koch C.C., Miraglia P.Q. and Murty K.L., Mater. Sci. Eng. A 252 (1998)
36.

May J., Höppel H.W. and Göken M., Mater. Sci. Forum 503 −504 (2005) 781.

May J., Höppel H.W. and Göken M., Scripta Mater. 53 (2005) 189.

Meyers M.A. and Chawia K.K., Mechanical Metallurgy: Principles and Applications,
Englewood Cliffs, NJ: Prentice-Hall (1984) p. 496.

Meyers M.A., Mishra A. and Benson D., JOM 58 (2006) 41.

Meyers M.A., Mishra A. and Benson D.J., Prog. Mater. Sci. 51 (2006) 427.

Mohamed F.A. and Li Y., Mater. Sci. Eng. A 298 (2001) 1.

Murayama M., Howe J.M., Hidaka H. and Takaki S., Science 295 (2002) 2433.

Newbery A.P., Ahn B., Pao P., Nutt S.R. and Lavernia E.J., Adv. Mater. Res. 29 −30
(2007) 21.
192
Newbery A.P., Ahn B., Hayes R.W., Pao P., Nutt S.R. and Lavernia E.J., Metall. Mater.
Trans. (2008) in press.

Newbery A.P., Han B.Q., Lavernia E.J., Suryanarayana C. and Christodoulou J.A., In:
Groza J.R. et al., editors, Materials Processing Handbook, Boca Raton, Fl: CRC Press
(2007) pp. 13-1 −13-28.

Newbery A.P., Nutt S.R. and Lavernia E.J., JOM 58 (2006) 56.

Nieman G.W., Weertman J.R. and Siegel R.W., Scripta Metall. 23 (1989) 2913.

Nieman G.W., Weertman J.R. and Siegel R.W., Scripta Metall. Mater. 24 (1990) 145.

Novikov I.I., Portnoy V.K. and Terentieva T.E., Acta Metall. 25 (1977) 1139.

Oliver W.C. and Pharr G.M., J. Mater. Res. 7 (1992) 1564.

Oliver W.C. and Pharr G.M., J. Mater. Res. 19 (2004) 3.

Ovid'ko I.A., Science 295 (2002) 2386.

Pande C.S., Masumura R.A. and Armstrong R.W., Nanostruct. Mater. 2 (1993) 323.

Pao P.S., Jones H.N., Feng C.R., Witkin D.B. and Lavernia E.J., In: Zhu Y.T. et al.,
editors, Ultrafine Grained Materials IV, Warrendale, PA: The Minerals, Metals and
Materials Society (TMS) (2006) pp. 331 −336.

Perez R.J., Huang B. and Lavernia E.J., Nanostruct. Mater. 7 (1996) 565.

Perez R.J., Jiang H.G., Dogan C.P. and Lavernia E.J., Metall. Mater. Trans. 29A (1998)
2469.

Petch N.J., J. Iron Steel Inst. 174 (1953) 25.

Peters W.H. and Ranson W.F., Opt. Eng. 21(3) (1982) 427.

Peters W.H., Ranson W.F., Sutton M.A., Chu T.C. and Anderson J., Opt. Eng. 22(6)
(1983) 738.

Pharr G.M., Oliver W.C. and Brotzen F.R., J. Mater. Res. 7 (1992) 613.

Pilling J. and Ridley N., Superplasticity in Crystalline Solids, London, UK: The Institute
of Metals (1989).

Rodriguez P., Bull. Mater. Sci. 6 (1984) 653.
193
Rufino B., Boulc’h F., Coulet M.-V., Lacroix G. and Denoyel R., Acta Mater. 55 (2007)
2815.

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. A 393 (2005) 344.

Sanders P.G., Eastman J.A. and Weertman J.R., Acta Mater. 45 (1997) 4019.

Sanders P.G., Fougere G.E., Thompson L.J., Eastman J.A. and Weertman J.R.,
Nanostruct. Mater. 8 (1997) 243.

Sanders P.G., Rittner M., Kiedaisch E., Weertman J.R., Kung H. And Lu Y.C.,
Nanostruct. Mater. 9 (1997) 433.

Schiøtz J. and Jacobsen K.W., Science 301 (2004) 1357.

Schiøtz J., Di Tolla F.D. and Jacobsen K.W., Nature 391 (1998) 561.

Schoenung J.M., Ye J., He J., Tang F. and Witkin D., Mater. Forum 29 (2005) 123.

Schuh C., Nieh T.G. and Iwasaki H., Acta Mater. 51 (2003) 431.

Schuh C., Nieh T.G. and Yamasaki T., Scripta Mater. 46 (2002) 735.

Schwaiger R., Moser B., Dao M., Chollacoop N. and Suresh S., Acta Mater. 51 (2003)
5159.

Shackelford J.F. and Alexander W., Materials Science and Engineering Handbook, 3
rd

Edition, Boca Raton, FL: CRC Press (2001).

Shariat P., Vastava R.B. and Langdon T.G., Acta Metall. 30 (1982) 285.

Shibata M., JEOL News 39(1) (2004) 28.

Siegel R.W., In: Fiorani D. and Sberveglieri G., editors, Fundamental Properties of
Nanostructured Materials, Singapore: World Scientific (1993) p. 3.

Smith B.W., Li X. and Tong W., Exp. Tech. 22 (1989) 19.

Smith T.R., Armstrong R.W., Hazzledine P.M., Masumura R.A. and Pande C.S., In:
Otooni M.A. et al., editors, Grain Size and Mechanical Properties: Fundamentals and
Applications, Pittsburgh, PA: Mater. Res. Soc. (MRS) Symp. Proc. 362 (1995) p. 31.

194
Suryanarayana C. and Koch C.C., Hyperfine Interact. 130 (2000) 5.

Suryanarayana C. and Koch C.C., In: Suryanarayana C., editor, Non-Equilibrium
Processing of Materials, Amsterdam, Netherlands: Pergamon (1999).

Suryanarayana C. and Prabhu B., In: Koch C.C., editor, Nanostructured materials:
Processing, Properties, and Applications, 2
nd
Edition, Norwich, NY: Noyes Publications
(2007) pp. 47 −90

Suryanarayana C., Intl. Mat. R. 40 (1995) 41.

Suryanarayana C., Prog. Mater. Sci. 46 (2001) 1.

Sutton M.A., Cheng M., Peters W.H., Chao Y.J. and McNeill S.R., Image Vision Comput.
4(3) (1986) 143.

Takeuchi S., Scripta Mater. 44 (2001) 1483.

Talbot D.E.J., The Effects of Hydrogen in Aluminum and Its Alloys, Cambridge, MA:
Maney (2004).

Tanimoto H., Sakai S. and Mizubayashi H., Nanostruct. Mater. 12 (1999) 751.

Tellkamp V.L., Dallek S., Cheng D. and Lavernia E.J., J. Mater. Res. 16 (2001) 938.

Tellkamp V.L., Melmed A. and Lavernia E.J., Metall. Mater. Trans. 32A (2001) 2335.

Tjong S.C. and Ma Z.Y., Mater. Sci. Eng. R 29 (2000) 49.

Valiev R.Z. and Alexandrov I.V., Nanostruct. Mater. 12 (1996) 35.

Valiev R.Z. and Langdon T.G., Prog. Mater. Sci. 51 (2006) 881.

Valiev R.Z., Alexandrov I.V., Zhu Y.T. and Lowe T.C., J. Mater. Res. 17 (2002) 5.

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., Nature Mater. 3 (2004) 511.

Valiev R.Z., Sergueeva A.V. and Mukherjee A.K., Scripta Mater. 49 (2003) 669.

Van Swygenhoven H. and Caro A., Nanostruct. Mater. 9 (1997) 669.
195
Van Swygenhoven H., Caro A. and Spaczer M., Acta Mater. 47 (1999) 3117.

Van Vliet K.J., Tsikata S. and Zurres S., Appl. Phys. Lett. 83 (2003) 1441.

Vastava R.B. and Langdon T.G., Acta Metall. 27 (1979) 251.

Veprek S., Veprek-Heijiman M.G.J., Karvankoa P. and Prochazka J., Thin Solid Films
476 (2005) 1.

Wang J., Iwahashi Y., Horita Z., Furukawa M., Nemoto M., Valiev R.Z. and Langdon
T.G., Acta Mater. 44 (1996) 2973.

Wang N., Wang Z.R., Aust K.T. and Erb U., Mater. Sci. Eng. A 237 (1997) 150.

Wang Y., Chen M., Zhou F. and Ma E., Nature 419 (2002) 912.

Wang Y.M., Hodge A.M., Bythrow P.M., Barbee, Jr. T.W. and Hamza A.V., Appl. Phys.
Lett. 89 (2006) 81903.

Wang Y.M., Ma E. and Chen M.W., Appl. Phys. Lett. 80 (2002) 2395.

Wang Y.M., Wang K., Pan D., Lu K., Hemker K.J. and Ma E., Scripta Mater. 48 (2003)
1581.

Weertman J.R. and Averback R.S., In: Edelstein A.S. et al., editors, Nanomaterials:
Synthesis, Properties and Applications, Bristol, UK: Institute of Physics Publishing
(1996) pp. 323 −345.

Weertman J.R., Farkas D., Hemker K., Kung H., Mayo M., Mitra R. and Van
Swygenhoven H., MRS Bull. 24 (1999) 44.

Weertman J.R., In: Koch C.C., editor, Nanostructured materials: Processing, Properties,
and Applications, 2
nd
Edition, Norwich, NY: Noyes Publications (2007) pp. 537 −564.

Wei Q., Cheng S., Ramesh K.T. and Ma E., Mater. Sci. Eng. A 381 (2004) 71.

Wei Q., J. Mater. Sci. 42 (2007) 1709.

Wei Q., Jia D., Ramesh K.T. and Ma E., Appl. Phys. Lett. 81 (2002) 1240.

Wei Q., Kecskes L., Jiao T., Hartwig K.T., Ramesh K.T. and Ma E., Acta Mater. 52
(2004) 1859.

Witkin D., Han B.Q. and Lavernia E.J., Metall. Mater. Trans. 37A (2006) 185.

196
Witkin D., Lee Z., Rodriquez R., Nutt S. and Lavernia E., Scripta Mater. 49 (2003) 297.

Witkin D.B. and Lavernia E.J., In: Shaw L.L. et al., editors, Processing and Properties of
Structural Nanomaterials, Warrendale, PA: The Minerals, Metals and Materials Society
(TMS) (2003) pp. 117 −124.

Witkin D.B. and Lavernia E.J., Prog. Mater. Sci. 51 (2006) 1.

Wolf D., Yamakov V., Philipot S.R., Mukherjee A. and Gleiter H., Acta Mater. 53 (2005)
1.

Wurschum R., Herth S. and Brossmann U., Adv. Eng. Mater. 5 (2003) 365.

Xun Y., Lavernia E.J. and Mohamed F.A., Metall. Mater. Trans. 35A (2004) 573.

Yamakov V., Wolf D., Phillpot S.R. and Mukherjee A.K., Nature Mater. 1 (2001) 45.

Yamakov V., Wolf D., Salazar M., Phillpot S.R. and Gleiter H., Acta Mater. 49 (2001)
2713.

Yang N.Y.C., Headley T.J., Kelly J.J. and Hruby J.M., Scripta Mater. 51 (2004) 761.

Ye J., Han B.Q., Lee Z., Ahn B., Nutt S.R. and Schoenung J.M., Scripta Mater. 53 (2005)
481.

Yip S., Nature 391 (1998) 532.

Youngdahl C.J., Sanders P.G., Eastman J.A. and Weertman J.R., Scripta Mater. 37
(1997) 809.

Youssef K.M., Scattergood R.O., Murty K.L. and Koch C.C., Appl. Phys. Lett. 85 (2004)
929.

Youssef K.M., Scattergood R.O., Murty K.L. and Koch C.C., Scripta Mater. 54 (2006)
251.

Youssef K.M., Scattergood R.O., Murty K.L., Horton J.A. and Koch C.C., Appl. Phys.
Lett. 87 (2005) 91904.

Zhang X., Wang H., Scattergood R.O., Narayan J. and Koch C.C., Mater. Sci. Eng. A
344 (2003) 175.

Zhang X., Wang H., Scattergood R.O., Narayan J., Koch C.C., Sergueeva A.V. and
Mukherjee A.K., Acta Mater. 50 (2002) 4823.

197
Zhilyaev A.P. and Langdon T.G., Prog. Mater. Sci. (2008) in press.

Zhilyaev A.P., Kim B.-K., Nurislamova G.V., Baró M.D., Szpunar J.A. and Langdon
T.G., Scripta Mater. 46 (2002) 575.

Zhou F., Lee J., Dallek S. and Lavernia E.J., J. Mater. Res. 16 (2001) 3451.

Zhou F., Liao X.Z., Zhu Y.T., Dallek S. and Lavernia E.J., Acta Mater. 51 (2003) 2777.

Zhou F., Witkin D., Nutt S.R. and Lavernia E.J., Mater. Sci. Eng. A 375 −377 (2004) 917.

Zhu Y.T. and Langdon T.G., JOM 56 (2004) 58.

Zhu Y.T. and Liao X., Nature Mater. 3 (2004) 351.

Zhu Y.T. Kolobov Y.R., Grabovetskaya G.P., Stolyarov V.V., Girsova N.V. and Valiev

Abstract (if available)

Abstract Nanocrystalline or ultrafine-grained Al alloys are often produced by severe plastic deformation methods and exhibit remarkably enhanced strength and hardness compared to conventional coarse-grained materials, resulting in great potential for structural applications. To achieve nanocrystalline structure, grains were refined by cryomilling (mechanical milling at cryogenic temperature) pre-alloyed powders. Cryomilling provides capability for rapid grain refinement and synthesis of commercial quantities (30−40 kg). The cryomilled powder was primarily consolidated by hot or cold isostatic pressing in general. Secondary consolidation was achieved by extrusion or forging. Alternatively, quasi-isostatic forging was applied either as an initial consolidation or as a further deformation step.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Mechanical behavior and microstructure optimization of ultrafine-grained aluminum alloys and nanocomposites

PDF

Microstructural evolution and mechanical properties of an aluminum alloy processed by equal-channel angular pressing and high-pressure torsion

PDF

Microstructural evolution and mechanical properties of a copper-zirconium alloy processed by severe plastic deformation

PDF

Processing, mechanical behavior and biocompatibility of ultrafine grained zirconium fabricated by accumulative roll bonding

PDF

Investigation of the influence of severe plastic deformation on the microstructure and mechanical properties of Al-7136 alloy

PDF

Properties of a ZK60 magnesium alloy processed by severe plastic deformation

PDF

Electrochemical evaluation of advanced anodized aluminum and chromate-free conversion coatings

PDF

Vacuum-bag-only processing of composites

PDF

Deformation patterns and subduction behavior of continental lithosphere: effects of rheology, surface processes, and material parameters

PDF

Long range internal stresses in cyclically deformed copper single crystals; and, Understanding large-strain softening of aluminum in shear at elevated temperatures

PDF

Creep in pure single and polycrystalline aluminum at very low stresses and high temperatures: an evaluation of Harper-Dorn creep

PDF

Direct observation of fault zone structure and mechanics in three dimensions: a study of the SEMP fault system, Austria

PDF

High temperature creep behaviors of additively manufactured IN625

PDF

Atomistic modeling of the mechanical properties of metallic glasses, nanoglasses, and their composites

Asset Metadata

Creator Ahn, Byungmin (author)

Core Title Deformation behavior and microstructural evolution of nanocrystalline aluminum alloys and composites

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Materials Science

Publication Date 07/02/2010

Defense Date 06/17/2008

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag aluminum alloys,deformation behavior,microstructural evolution,nanocomposites,nanocrystalline,OAI-PMH Harvest

Language English

Advisor Nutt, Steven R. (committee chair), Goo, Edward K. (committee member), Sammis, Charles G. (committee member)

Creator Email byungahn@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1292

Unique identifier UC1342385

Identifier etd-Ahn-20080702 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-74600 (legacy record id),usctheses-m1292 (legacy record id)

Legacy Identifier etd-Ahn-20080702.pdf

Dmrecord 74600

Document Type Dissertation

Rights Ahn, Byungmin

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