Ulm, 31.10.2016
Nanostructure Evolution and Enhancement of Mechanical
Properties for Medium Carbon Steel through High Pressure
Torsion Processing - Nanosteels
DISSERTATION
zur Erlangung des akademischen Grades eines
DOKTOR-INGENIEURS
(DR.-ING.)
der Fakultät für Ingenieurwissenschaften,
Informatik und Psychologie der Universität Ulm
von
Mike Haddad
aus Amman, Jordanien
Gutachter: Prof. Dr. Hans-Jörg Fecht
Prof. Dr. Ruslan Valiev
Amtierende Dekanin: Prof. Dr. Frank Kargl
To my father, mother, brother and sister for their endless support especially during the past six
years. Without them, I would not be able to finish this research.
البحث هذا النهي اكن لم بدونهم والذين اعوام، ستة اخر خالل خصوصا الالمتناهي لدعمهم اختي و اخي امي، ابي، الى
(i)
CHAPTER 1 INTRODUCTION AND MOTIVATION 1
1.1 INTRODUCTION 3
1.2 RESEARCH MOTIVATION 5
CHAPTER 2 LITERATURE REVIEW AND GENERAL PRINCIPLES 9
2.1 GENERAL OVERVIEW 11
2.2 HIGH PRESSURE TORSION (HPT) 12
2.2.1 THE PRINCIPLE OF HIGH PRESSURE TORSION (HPT) 12
2.2.2 STRAIN IMPOSED BY HIGH PRESSURE TORSION (HPT) 14
2.2.3 IMPORTANT PARAMETERS IN HIGH PRESSURE TORSION (HPT) PROCESSES 15
2.2.4 FUTURE PROSPECTS AND INNOVATION POTENTIAL 16
2.3 EQUAL CHANNEL ANGULAR PRESSING (ECAP) 17
2.3.1 THE PRINCIPLE OF EQUAL CHANNEL ANGULAR PRESSING (ECAP) 18
2.3.2 THE STRAIN ACHIEVED BY ECAP 19
2.3.3 FACTORS INFLUENCING ECAP PROCESS 20
2.3.4 ECAP FUTURE PROSPECTS FOR INDUSTRIAL APPLICATIONS 22
2.4 ACCUMULATIVE ROLL BONDING (ARB) 22
2.5 TWIST EXTRUSION 24
2.6 CHARACTERISTICS OF MATERIALS PROCESSED BY SPD 26
2.6.1 GRAIN REFINEMENT 26
2.6.2 MECHANICAL PROPERTIES OF NANOSTRUCTURED MATERIALS 28
2.6.3 DEFORMATION MECHANISMS 28
2.6.4 STRENGTHENING MECHANISMS 29
2.6.5 THE STRENGTH-DUCTILITY PARADOX 31
CHAPTER 3 EXPERIMENTAL PROCEDURES 35
3.1 MATERIALS USED IN THIS RESEARCH 37
3.2 PROCESSING METHODS 38
3.2.1 PATENTING TREATMENT 38
3.2.2 HIGH PRESSURE TORSION (HPT) AND WARM HIGH PRESSURE TORSION (WHPT) 40
3.3 MICROSTRUCTURE CHARACTERIZATION 41
3.3.1 SCANNING ELECTRON MICROSCOPY (SEM) 41
3.3.2 ELECTRON BACKSCATTER DIFFRACTION (EBSD) 42
3.3.3 TRANSMISSION ELECTRON MICROSCOPY (TEM) 43
3.3.3.1 TEM specimen preparation using Focused ion beam (FIB) 43
3.3.3.2 Automated phase and crystal orientation mapping TEM (ACOM) 44
3.4 MECHANICAL PROPERTIES 45
3.4.1 HARDNESS 46
3.4.1.1 Vickers microhardness 46
(ii)
3.4.1.2 Nano-indentation 47
3.4.2 IN-SITU TENSILE TEST 49
CHAPTER 4 EXPERIMENTAL RESULTS 53
4.1 AS-DELIVERED SAMPLES PROCESSED BY RT HPT 55
4.1.1 MICROHARDNESS 55
4.1.2 SEM OBSERVATIONS 56
4.2 PATENTED C45 SAMPLES PROCESSED BY RT HPT 59
4.2.1 MICROHARDNESS 60
4.2.2 MICROSTRUCTURE 60
4.2.2.1 SEM observations 60
4.2.2.2 TEM microstructure characterization 61
4.2.2.3 Grain size and grain size distribution 63
4.2.3 TENSILE TESTS 65
4.2.4 ANNEALING THE RT HPT PROCESSED SAMPLES 66
4.2.4.1 TEM microstructure characterization 66
4.2.4.2 Tensile test after annealing the RT HPT processed samples 67
4.3 C45 SAMPLES PROCESSED BY WHPT AT 350°C 71
4.3.1 TEM AND ACOM MICROSTRUCTURE CHARACTERIZATION 71
4.3.2 IN-SITU TENSILE TEST 72
4.3.3 OBSERVATIONS DURING IN-SITU TENSILE TEST 73
4.3.4 FRACTURE SURFACE 73
4.3.5 MICROSTRUCTURE OF SAMPLES IN THE VICINITY OF FRACTURE SURFACES 74
4.3.6 IN-SITU TENSILE TEST AT ELEVATED TEMPERATURES 74
4.4 C45 SAMPLES PROCESSED BY WHPT AT 380°C 79
4.4.1 TEM MICROSTRUCTURE CHARACTERIZATION 80
4.4.2 IN-SITU TENSILE TEST 80
4.4.3 OBSERVATIONS DURING IN-SITU TENSILE TEST 81
4.4.4 FRACTURE SURFACE 82
CHAPTER 5 DISCUSSION 85
5.1 GRAIN REFINEMENT AND MICROSTRUCTURE EVOLUTION DURING HPT 87
5.1.1 ROOM TEMPERATURE HPT PROCESSING 87
5.1.2 WARM HPT PROCESSING 89
5.2 MECHANICAL BEHAVIOR 92
5.2.1 STRENGTHENING MECHANISMS 92
5.2.2 DEFORMATION AND FRACTURE BEHAVIOR 95
5.2.2.1 Absence of necking 96
5.2.2.2 Slant fracture 99
5.2.2.3 Fracture surface 102
5.3 MISCELLANEOUS 103
(iii)
5.3.1 CONTINUOUS RECRYSTALLIZATION 103
5.3.2 SUPERPLASTIC BEHAVIOR 104
CHAPTER 6 SUMMARY AND CONCLUSION 107
REFERENCES 117
PUBLICATIONS 132
CONFERENCE CONTRIBUTIONS 133
ACKNOWLEDGEMENT 134
LIST OF ACRONYMS 135
LIST OF FIGURES 136
CURRICULUM VITAE 141
(1)
Chapter 1 Introduction and Motivation
(3)
Chapter 1 Introduction and Motivation
1.1 Introduction
Although nanotechnology is considered a relatively new scientific field, its vision goes back to
December of 1959, and specifically to the famous talk ”There is plenty of room at the bottom”
[1] during the American Physical Society meeting at Caltech, in which Richard Feynman invited
scientists to enter a new field of physics by presenting his vision of designing, building, and
handling things on the atomic scale. He described a technique for writing the Encyclopaedia
Britannica on the head of a pin and suggested miniaturizing computers, electrical components,
as well as machines and motors to a very small scale.
Since nanotechnology involves controlling and manipulating of matters on atomic and molecular
scale, it therefore influences different scientific subjects such as applied physics, chemistry,
engineering, and materials science. Hence, nanotechnology shows huge potential in scientific
research and technological applications, and it can be considered as one of the prominent
technologies of the future.
Almost 20 years after Feynman’s talk and specifically in 1981, Herbert Gleiter [2] discussed the
advantages of ultrafine grains in solids, and came out with the term “nanostructured solids”
during the second Ris international symposium on metallurgy and materials science. Nowadays,
nanomaterials can be considered as one of the most important and hot research fields, which
are, according to the International Organization for Standardization (ISO) and the European
Standardization Committee (CEN), “materials with any external dimension in the nanoscale or
having internal structure or surface feature in the nanoscale.” With a variety of potential
applications, nanomaterials are becoming increasingly relevant in modern research fields and
attract both scientists and industry, as such materials demonstrate a number of unusual features
of mechanical behavior [3, 4]. Nanomaterials can be classified as zero dimension (Clusters), one
dimension (multilayer), two dimensions (ultrafine-grained overlayers), and three dimensions
(bulk nanostructured materials) [5], and presently, many materials science specialists are
focusing their research towards this promising discipline.
In order to synthesize nanostructured materials two approaches have been developed. The first
approach is using individual atoms to assemble nanostructured materials, called the “bottom-
(4)
Chapter 1 Introduction and Motivation
up” approach [6], such as those prepared from the gas phase (inert gas condensation [7]), or
liquid phase [8]. The second approach is called the “top-down” approach in which substantial
grain refinement and nanostructures are produced by processing existing coarse-grained
materials. The most successful technique for this approach is the use of severe plastic
deformation (SPD) processing [6], which can be defined as any method of metal forming under
an extensive hydrostatic pressure that may be used to impose a very high strain on a bulk solid
without causing considerable change in the dimensions of the sample, while at the same time,
having the ability to produce exceptional grain refinement [9].
In the last decade, there has been growing interest in bulk nanocrystalline materials, and
particularly in those processed by SPD method such as high pressure torsion (HPT) and equal
channel angular pressing (ECAP), which have been successfully utilized as novel processing
techniques for producing nano and sub-microcrystalline metallic materials [10, 11]. This interest
is not only because of the unique physical and mechanical properties achieved by various
nanostructured materials, but also due to several advantages of SPD materials as compared to
other nanostructured materials [12]. By implementing the SPD methods, grain refinement, in
addition to phase transformations, will take place during the processing [13] . This results in an
ultrafine grain size and a unique and exceptional chemical, physical, and mechanical properties,
such as high yield stress and high ultimate tensile strength; significantly enhanced in comparison
with the coarse grained counterparts’ properties [14, 15].
Due to its capability of deforming the sample severely (with true strain ≥ 10) [12], and achieving
superior grain refinement with exceptional strength [16], the HPT method, which is processing
the sample by torsional straining under compressive pressure, showed great importance and
became one of the major SPD techniques [9]. However, despite the fact that extensive research
was directed towards this field, the behavior and the fundamental principles are not fully
understood yet [17, 18, 19] as, for example, such materials have extraordinary strength. The
ductility is however disappointingly low, and the distribution of the carbon atoms in the ferrite
matrix is not clear yet [18].
Today, with SPD capabilities of producing bulk nanostructured materials, the utilization of such
materials in electro-conductors and microdevices is very close, while in the field of biomedical
(5)
Chapter 1 Introduction and Motivation
engineering, bulk nanostructured materials have been utilized successfully in a first-application
as materials for dental implants [20]. This opened the door for movement from research to
commercial manufacturing for a wide range of promising products.
1.2 Research motivation
In order to be able to develop innovative processing methods and procedures that can enhance
and improve the mechanical properties and mechanical behavior of structural materials
processed by SPD, it is essential to have deep understanding of the nanostructure formation and
grain refinement mechanisms, as well as deformation and fracture physics.
The goals of this research were as follows:
- To investigate the grain refinement mechanism and to understand the factors that can
contribute, either positively or negatively, to this concern, such as the effect of the initial
microstructure, the applied hydrostatic pressure, the number of processing rotations
(amount of applied strain), and the effect of the processing temperature on the resulting
grain size by using state of the art techniques and devices such as scanning electron
microscope (SEM), electron backscatter diffraction (EBSD), focused ion beam (FIB), and
transmission electron microscopy (TEM), including high resolution TEM (HRTEM), and
automated phase and crystal orientation mapping (ACOM).
- To study the effect of the induced strain on second phase particles (Cementite in C45 steel)
and their concomitant dissolution, in addition to understanding the impact of other boundary
conditions and process parameters on this process, as well as, the effect of the presence and
absence of such particles on the resulting microstructure using the aforementioned methods.
- To systematically investigate the mechanical properties of the processed samples, and the
effect of the process parameters and resulting microstructure on the strength and hardness
using Vickers microhardness, nanoindentation, and a dedicated tensile test stage suitable for
such miniature HPT samples.
- To systematically study and understand the deformation and fracture behavior of the
processed samples, and to reveal the effect of the microstructure and all other process
(6)
Chapter 1 Introduction and Motivation
parameters on this behavior. This part includes strain hardening behavior, superplastic
behavior, ductility, neck formation and instability, crack formation and propagation, in
addition to the fracture surface and fracture type. For this part, in-situ deformation
measurements were carried out on a special tensile stage installed inside the SEM, and the
deformation relief and fracture surfaces were monitored comprehensively, upon which a
sequence of high resolution SEM images of the samples surfaces at different locations and
different magnifications were taken.
- To design and develop processing criteria, based on the obtained results and arrived
understanding and conclusions, about the most suitable process parameters and boundary
conditions that can enhance the strength, while at the same time, keep a reasonable amount
of ductility of C45 steel.
This dissertation is divided into six chapters. Chapter one gives a brief introduction as well as
explains the purposes and objectives of this research, while chapter two describes the principles
of severe plastic deformation with focus on the most important techniques such as HPT, ECAP,
accumulative roll bonding (ARB), and twist extrusion, in addition to the grain refining and
strengthening mechanisms for nanostructured materials. Furthermore, chapter three includes a
description of the materials used, the processing methods, testing and experimental procedures,
and experimental equipment and devices.
The results of this research, as well as the discussion, can be found in chapters four and five.
Chapter four shows the microstructure results obtained through different HPT processing
procedures, the results related to mechanical properties, such as hardness and tensile test
results, the SEM observations during in-situ tensile tests, the results of annealing the room
temperature HPT processed samples, as well as presents the SEM observations during
performing of in-situ tensile test at elevated temperatures. Furthermore, chapter five discusses
the microstructure evolution and the grain refinement mechanism taking place during the
processing, the strengthening mechanisms contributing to the high strength achieved after HPT
processing, the deformation and fracture behavior, such as the absence of necking and the slant
fracture, as well as it studies the features present on the fracture surface, the continuous
(7)
Chapter 1 Introduction and Motivation
recrystallization resulted from the room temperature annealing, in addition to the limitation in
showing superplastic behavior. Finally, Chapter six includes a summary and outlook on future
research.
Chapter 2 Literature Review and General
Principles
(11)
Chapter 2 Literature Review and General Principles
2.1 General overview
Over the last decade, the development of bulk nanostructured materials has proven to be highly
admissible in modern materials science research [6, 21, 22]. The conventional metal forming
processes that impose heavy deformation such as cold rolling, drawing, forging, and extrusion,
cannot exceed a plastic strain of 2.0 [14]. Even when multi-passes are carried out to higher values
the accompanied reduction in some dimensions of the samples will be significant, making them
unsuitable for structural applications. The formed structures are usually substructures of a
cellular type having boundaries with low angle misorientations [12]. Therefore, the use and
development of non-conventional processing techniques, with more powerful grain refinement
capabilities, became increasingly valuable.
Severe plastic deformation (SPD) can be defined as a group of metal forming processes, through
which a very large plastic strain is applied on a bulk solid sample under high hydrostatic pressure,
without introducing any significant change in the original shape or the overall dimensions of the
sample, and having the ability to produce exceptional grain refinement in the range of nano- or
sub micro-meter [6, 9, 12, 14].
Ultrafine grained materials can be defined as polycrystals featuring an average grain sizes
between 1 nm and 1000 nm, and with high angle misorientation for the majority of grain
boundaries. In this respect, SPD has attracted much attention, as it can be considered, nowadays,
a very well established procedure for the production of bulk, ultrafine grained and
nanostructured materials [21].
Early in the 1930’s Percy Williams Bridgman [23] performed a series of experiments dealing with
a large plastic flow, in which he applied torsion straining combined with high quasy-hydrostatic
pressure. In the 80’s these experiments found renewed interest when this method was proposed
for use in microstructure refinement of metallic samples [24]. Following this, significant progress
was made in the field of SPD, and several special and innovative deformation techniques were
introduced to process massive samples in more efficient ways, and to study their obtained
properties.
(12)
Chapter 2 Literature Review and General Principles
Nowadays, numerous SPD techniques have been developed, reported, and comprehensively
studied. These methods vary in their technical characteristics and deformation modes, ranging
from pure shear to simple shear, but commonly impose severe deformation with large total strain
[22], for example the already mentioned HPT derived from the pioneering experiments of
Bridgman [24], ECAP developed by Segal and co-workers in the 1970s [25], ARB proposed by Saito
et al. [26], and twist extrusion (TE) introduced by Beygelzimer et al. [27].
2.2 High pressure torsion (HPT)
In the year 1935 the fundamental principle of HPT was first proposed by Bridgman, when he used
the technique to study the effect of heavy deformation through high shearing strain under high
hydrostatic pressure on 57 different elements [23]. This processing method became of major
importance only 30 years ago after its capabilities of achieving outstanding grain refinement, and
therefore, exceptional high strength, and other favorable properties were recognized by
Smirnova et al. [24].
In general, strength and ductility are fundamental properties that are closely connected with
other engineering properties such as fatigue, fracture toughness, durability, wear resistance,
creep, and superplasticity [20]. Thus, materials with small grain sizes possess several advantages
and superior properties compared to their conventional coarse-grained counterparts.
Processing the same material with HPT and ECAP confirmed that the former method is more
efficient in grain refinement, with a significantly lower fraction of low-angle grain boundaries
[28]. Furthermore, the HPT technique can be used not only for processing bulk metals, but also
for efficient consolidation of powder, as it is possible to prepare samples with densities close to
100% [14].
2.2.1 The principle of high pressure torsion (HPT)
As explained previously, during high pressure torsion (HPT), the ingot in the form of a thin disk,
is placed between two anvils under applied hydrostatic pressure of several GPa, as depicted
schematically in Figure 1(a). Simultaneously, one of the anvils rotates, and the surface frictional
(13)
Chapter 2 Literature Review and General Principles
forces deform the disk by shear. Hence, despite the very high torsional strain imposed on the
sample, it does not break apart due to the presence of high quasi-hydrostatic pressure.
In practice, there are two different types of high pressure torsion processing: unconstrained and
constrained [29]. As shown in Figure 1(b), in the unconstrained type the material is free to flow
outward under the applied pressure. Therefore, the final thickness of the sample after HPT
processing is very small; typically 200 µm. On the other hand, in the constrained type the outward
flow of the material is limited, as the sample is machined to fit perfectly inside a cavity, either in
the lower anvil or in both anvils [16]. In this case, the final sample thickness can be up to 1 mm.
Figure 1: (a) A schematic drawing illustrating the processing procedure of the high pressure torsion technique; (b) A schematic drawing showing the different types of high pressure torsion processing. Figure 1 (b) is republished from reference [29] with permission of Springer; permissions conveyed through Copyrights Clearance Center, Inc.
(14)
Chapter 2 Literature Review and General Principles
2.2.2 Strain imposed by high pressure torsion (HPT)
As illustrated in Figure 2, the displacement dl which results from a small rotation d can be
calculated from the following relation:
𝒅𝒍 = 𝒓𝒅𝜽 (1)
Where r is the radius of the disk.
Therefore, the incremental shear strain d is given by [28]:
𝒅𝜸 =
𝒅𝒍
𝒉=
𝒓𝒅𝜽
𝒉 (2)
Where h is the thickness of the disk
Assuming that the thickness of the disk h is independent of the rotation angle , and since =2N,
it follows that the integration of equation (2) will give the shear strain :
𝜸 =
𝟐𝝅𝑵𝒓
𝒉 (3)
Figure 2: An illustration for the high pressure torsion sample showing the parameters used to estimate the imposed strain. Reprinted from reference [28], with permission from Elsevier; permissions conveyed through Copyrights Clearance Center, Inc.
(15)
Chapter 2 Literature Review and General Principles
For large imposed strains, where 0.8, the equivalent strain 𝜖eq is given by [16]:
𝝐𝒆𝒒 =
𝟐
√𝟑𝒍𝒏 [(𝟏 +
𝜸𝟐
𝟒)𝟏/𝟐 +
𝜸
𝟐] (4)
Furthermore, Degtyarev et al. [30, 31] developed an equation to calculate the equivalent strain
in which they incorporated the contribution of the thickness reduction due to the applied
pressure during HPT. This is given by:
𝝐𝒆𝒒 = 𝒍𝒏 [𝟏 + (𝟐𝝅𝑵𝒓
𝒉)
𝟐
]
𝟏/𝟐
+ 𝒍𝒏 (𝒉𝒐
𝒉) (5)
where ho is the initial thickness of the disk, while h is the final thickness after the HPT treatment.
2.2.3 Important parameters in high pressure torsion (HPT) processes
According to the abovementioned equations used to calculate the shear strain and the
equivalent strain 𝜖eq (equations (3), (4) and (5)), it is clear that the strain has a minimum value at
the center (zero in the case of equations (3) and (4)), which increases moving outward until it
reaches its maximum at the periphery. Therefore, it is logical to expect an inhomogeneous
(gradient) microstructure after processing a sample by HPT.
However, despite the fact that some researchers have reported a variation in the hardness across
processed samples (lower hardness values at the center) [32, 33, 34], other researchers have
confirmed the possibility of achieving a reasonably homogeneous microstructure (including the
center region), with close Vickers microhardness values between the center and the periphery in
different materials if the torsional strain and the imposed pressure were high enough [29, 35, 36,
37]. Hence it is clear that, in addition to the processed material itself, the shear strain, which can
be controlled by the number of rotations, as well as the imposed pressure during the processing
which also influences the flow stress as shown in [38], are the most relevant parameters affecting
the homogeneity of HPT processed samples.
The strain gradient plasticity model was applied successfully by Estrin et al. [39] to simulate the
highly non-uniform deformation taking place during HPT. To do so the first-order strain gradient
model was used, in which the material was considered to be a two-phase “composite”; the
(16)
Chapter 2 Literature Review and General Principles
dislocation cell walls, and the cell interiors. In the former, the climb processes are responsible for
the dynamic recovery, while in the latter, the dynamic recovery is controlled by the dislocation
cross-slip, which also depends on the stacking fault energy of the material [40].
In high stacking fault energy metals such as high-purity aluminum, experiments showed that the
center of the disk exhibited higher microhardness at low strains (low number of rotations) in
comparison with the periphery region due to the advent of dynamic recovery in that outer region.
By increasing the number of rotations, the value of microhardness at the periphery did not
change while the higher microhardness at the central region decreases, and the hardness became
reasonably homogeneous over the whole sample after 5 rotations [41].
Contrasting this, in lower stacking fault energy materials such as high purity nickel, copper, and
commercial purity aluminum, researchers reported that the central region showed lower
microhardness at the central region compared to the periphery region in the early stages of the
processing. Nevertheless, by proceeding with the deformation to a higher number of rotations
under high imposed pressure, a reasonably homogeneous microhardness could be achieved
across the sample [28, 29, 36, 37].
In agreement with the microhardness results, TEM investigations show smaller grain sizes and
highly deformed microstructure in the center region at early stages of HPT processing of high-
purity aluminum [41]. On the other hand, materials with lower stacking fault energies, such as
nickel and copper, have much slower recovery rates, and therefore microstructural evolution will
occur at a slower rate, where higher strains must be applied to reach an acceptable level of
microstructural homogeneity [16]. For both materials, the microstructure will gradually become
more homogeneous by proceeding with the torsional straining to a higher number of revolutions.
2.2.4 Future prospects and innovation potential
Due to the extensive developments in the field of SPD, the unique and attractive properties of
the polycrystalline materials processed by this technique are well recognized, and have become
relevant in modern engineering applications. HPT, as one of the major SPD techniques, has
advantages over other methods, as it yields to finer grain size and higher strength, as well, it is
(17)
Chapter 2 Literature Review and General Principles
very effective in processing some metals which are hard to be processed by other SPD techniques
such as magnesium alloys [42].
Based on the available researches and reports that explore and discuss the HPT processing
technique, and by following the research directions in this field, Zhilyaev et al. [16] identified
several potential applications for this process:
- Using HPT processed materials will be suitable to produce small parts and micro-components
for micro-electro-mechanical systems (MEMS), with the increased strength of the produced
parts being a significant advantage.
- HPT is an attractive processing technique that can be used to increase the strength of some
materials that have excellent biocompatibility, in order to use them in producing mini-
components such as mini-screws and staples, as well as dental and medical implants.
- Some researchers modified this method to be able to process ring and annulus shape
specimens with reasonably homogeneous microstructure and mechanical properties through
their cross- sections [43]. Therefore there is real potential to produce washers with very high
mechanical properties.
- As mentioned previously, HPT is an efficient and ideal tool for consolidating a broad range of
metallic powders and chips into bulk samples with densities close to 100%. Accordingly, this
method can be implemented to develop materials which can be used for specific purposes
such as hydrogen storage, or suitable for some other functional and structural applications.
2.3 Equal channel angular pressing (ECAP)
In 1972, and during their attempts to develop a special metal forming process that uses simple
shear to introduce high strain into metallic billets, Segal and his co-workers introduced and
evolved, for the first time, the ECAP technique [25]. With increasing interest in ultra-fine grained
materials, ECAP began to be a highly relevant process in research topics, and in the 1990s, a
number of studies that reported and discussed the microstructure evolution and the unique
properties of the samples processed by this technique were published [44, 45, 46].
(18)
Chapter 2 Literature Review and General Principles
Figure 3: A drawing illustrating the ECAP principle and showing the die’s angles and .
2.3.1 The principle of equal channel angular pressing (ECAP)
As depicted in Figure 3, a die with a special channel bent at an angle is used. The sample, which
has a special cross-section that fits within the channel, is pressed through the die and exists
without any changes to cross-sectional area dimensions. Simple shear straining conditions are
applied in this process, and since the cross-sectional area remains unchanged, this process can
be repeated with the same sample to attain severe deformation and to accumulate very high
plastic strain.
Furthermore, by rotating the sample in different directions between successive passes, different
slip systems could be activated [47]. In general, as illustrated in Figure 4 [48, 49], there are four
fundamental ECAP routes: A, BA, BC, and C, where in route A the sample is not rotated between
successive passes, in route BA, it is rotated by 90° in alternating directions (90° clockwise then 90°
counterclockwise), in route BC, the sample is rotated by 90° in the same direction and in route C,
the sample is rotated by 180° [47, 50, 51].
(19)
Chapter 2 Literature Review and General Principles
Figure 4: A graph illustrating the ECAP fundamental routes with the associated slip systems for the successive passes. Republished with permission of Elsevier and Korean Institute of Metals and Materials, from references [48] and [49] respectively; permissions conveyed through Copyrights Clearance Center, Inc.
Following the different ECAP routes, active slip systems will change accordingly as depicted in
Figure 4 [48, 49]. In ECAP processing using route A, two shearing planes intersecting at an angle
of 90° are generated, while in route BA the number of shearing planes increases to four with
planes intersecting at an angle of 120°C [52]. Processing using route BC, in its turn, will generate
two shearing planes intersecting at an angle of 120°, repeating themselves with opposite
shearing direction each time. However, in route C, there is only one shearing plane but with
opposite shearing direction after each pass [52].
2.3.2 The strain achieved by ECAP
One of the most important parameters for any method of SPD is the strain introduced into the
sample during this process. In each separate pass through the die, shear strain is imposed on the
ECAP-processed sample and accumulated severely, causing grain refinement. The magnitude of
this shear strain depends on the geometry of the die and specifically on the angle of intersection
(20)
Chapter 2 Literature Review and General Principles
between the two channels , and the angle subtended by the outer arc of curvature where the
two channels intersect , as depicted in Figure 3.
Assuming a fully lubricated specimen, in order to neglect the friction between the specimen and
the die, Iwahashi et al. [53] estimated the shear strain, , according to the following equation:
𝛄 = 𝟐 𝐜𝐨𝐭 (
𝚽
𝟐+
𝚿
𝟐) + 𝚿 𝐜𝐬𝐜 (
𝚽
𝟐+
𝚿
𝟐) (6)
Therefore, the equivalent strain accumulated after N cycles, 𝜖N, can be calculated from the
following general form equation [53]:
𝝐𝑵 = 𝑵 [𝟐 𝒄𝒐𝒕 (
𝜱𝟐
+𝜳𝟐
) + 𝜳 𝒄𝒔𝒄 (𝜱𝟐
+𝜳𝟐
)
√𝟑] (7)
By inspecting the above equations, it can be concluded that, in normal cases where 90°, the
angle has minimal effect on the induced strain. By using a die with = 90°, the equivalent strain
introduced into the sample will be close to one, and the lower the values of and for the die,
the higher values of the strain introduced into the sample [52].
2.3.3 Factors influencing ECAP process
When the ECAP method is used to process samples, there are several factors affecting and
influencing the workability, the microstructure characteristics, and the overall results achieved
through the use of this technique. Besides the die angles , , and the ECAP processing routes,
the processing speed, the pressing temperature, and the use of back-pressure are some factors
that have some influence on the ECAP processed samples. Some of these have significant
influence, while other factors have minor and sometimes negligible effects.
As explained previously, since the angle has a direct influence on the strain introduced into the
sample in each pass, and therefore on the as-processed microstructure, it follows that this angle
is the most significant factor in the ECAP process. Experiments have proven that the smaller the
channels angle is, the easier it is to attain more regular and equiaxed grains with higher angle
(21)
Chapter 2 Literature Review and General Principles
grain boundaries. Furthermore, the application of higher strain in each separate pass is more
important than the total accumulative strain imposed into the sample [54].
By contrast, the angle has a minor influence on the strain imposed on the ECAP processed
samples. Xu and Langdon [55] found that the contour maps for Vickers microhardness of the
samples processed using a die with small value of ( = 20°) show almost the same
homogeneity, and were comparable with the contour maps obtained from the samples
processed using ECAP die with = 0°. Nevertheless, it is almost impossible to machine a solid
ECAP die with = 0°, and since a small value of has insignificant effect on the homogeneity,
therefore, it is more practical and convenient to use solid ECAP dies with low values of .
On the other hand, the pressing speed has no significant influence on the ECAP processed
samples, while the pressing temperature has direct influence on the achieved microstructure, as
higher temperatures will yield to an increase in the obtained grain size, as well as lead to a higher
fraction of low-angle grain boundaries, due to the increasing annihilation of dislocations within
the grains and the increasing rates of recovery in the material after passing a certain temperature
that depends on the alloy itself [56].
Another factor with significant influence on the ECAP processed samples is the application of
back-pressure during the process, which leads to more uniform grain refinement and finer grain
size. In this concern, Lapovok [57] found that, first, the use of back-pressure during ECAP will
improve the workability of the sample, which is important for samples with low ductility because
they can be processed for more cycles without failure or cracks formation. Second, the use of
back pressure enhances the uniformity of the stress-strain conditions of metal flow during the
process. As a consequence, the problem of the dead zone on the outer angle will be solved, and
this corner will be filled. Thus, the processed samples will deform by simple shear and will have
more uniform grain refinement especially at the bottom surface of the sample.
As mentioned previously, the processing routes have strong influence on the ECAP processed
samples. Experiments found route BC, where the shearing directions lie in planes which intersect
at 120°, to be the most effective and preferable ECAP route with the greatest potential to
produce ultrafine equiaxed grains separated by high-angle grain boundaries [58, 59], while the
(22)
Chapter 2 Literature Review and General Principles
evolution of subgrains to grains with high-angle boundaries was found to be less rapid by
processing with route C, and the slowest when processing with route A [58]. By increasing the
die angle from 90° to 120°, the efficiency of route BC in respect of grain refinement is decreased
while the influence of route A is increased. Nevertheless, the influence of route BC is still higher
than that of route A, and route BC continues to be the optimal ECAP processing procedure [59].
2.3.4 ECAP Future prospects for industrial applications
Conventional ECAP processes require repetition of passes in order to achieve high strain in the
billet. That means that the sample must be removed from the die and reinserted again for the
next pass. This procedure requires additional labor and time expenditures, which makes the ECAP
process less attractive for industrial applications. Therefore, several ECAP procedures have been
developed to overcome this limitation. Some are designed to facilitate the processing procedure
in a laboratory, such as the use of rotary ECAP dies [60], the side-extrusion ECAP process [61],
and multi-passes dies [48], while others have been designed for continuous ECAP processing for
large volume production in order to scale to an industrial level, such as continuous confined strip
shearing (C2S2) [62], dissimilar-channel angular pressing (DCAP) [63], equal channel angular
drawing [64], the conshearing process [65], and ECAP-conform processing [66].
Therefore it is clear that, at present, efforts targeted at switching from laboratory scale to
industrial applications are being undertaken, and many companies around the world are involved
in developing new SPD methods with reduced production costs and decreased waste, in order to
make them economically feasible processes [20]. Moreover, there are real attempts to produce
raw materials in the form of billets, and semi-finished products in the form of sheets, rods, and
wires to be used in medical applications [67]. This has culminated into a start-up company that
works in nanostructured titanium rods fabrication for the medical industry [20].
2.4 Accumulative roll bonding (ARB)
In this technique, a conventional rolling setup is used but with a modified procedure. As
illustrated schematically in Figure 5, the surfaces of two stripes are wire brushed and degreased
in advance, then placed neatly on top of each other. After that, the two stripes are joined
(23)
Chapter 2 Literature Review and General Principles
together by conventional roll bonding process, causing the thickness to be reduced into half. The
length of the resulting joined material is sectioned into two halves, then wire brushed, degreased
again, and afterwards placed on top of each other and rolled again to one half thickness. By
repeating the previous steps of rolling, cutting, wire brushing, degreasing, and stacking, a large
strain is accumulating in the workpiece, keeping in mind that this process can be carried out at
room temperature as well as at elevated temperatures, as long as it is carried out below the
recrystallization temperature.
If the thickness reduction is maintained at 50% after each rolling pass the process will result in
no significant geometrical changes. This means that the rolling cycles can be repeated
continuously without any limits, and therefore, ultra-high plastic strain can be introduced in the
sample. Hence, the thickness of the initial stripe after n cycles can be calculated by the following
equation [68]:
𝒕 =𝒕𝟎
𝟐𝒏 (8)
Figure 5: Schematic drawing explaining the accumulative roll bonding ARB procedure. Reproduced with permission of Pergamon, from reference [26]; permission conveyed through Copyrights Clearance Center, Inc.
(24)
Chapter 2 Literature Review and General Principles
Where t0 is the initial stripe’s thickness.
Additionally, the total reduction in the initial stripe’s thickness rt after n cycles is given by [68]:
𝒓𝒕 = 𝟏 −
𝟏
𝟐𝒏 (9)
Furthermore, using Von Mises yield criterion, and assuming plane strain conditions, the
equivalent plastic strain 𝜖 after n cycles can be calculated as [68]:
𝝐 = 𝟎. 𝟖𝒏 (10)
Therefore, if 1.0 mm thickness sheet was processed by ARB for 10 cycles, the equivalent plastic
strain will be 8, while the final thickness of the initial stripe will be less than 1 m with a total
reduction of 99.9%.
Similar to conventional rolling processes, the material processed by ARB is subject to heavy
deformation, however the main difference is that, due to the repetition of procedure, the regions
just below the surface, which experience severe shear deformation, become distributed through
the thickness of the sheets, introducing new interfaces [68, 69]. Such interfaces, as well as oxide
films and inclusions on such surfaces, are dispersed uniformly by repetition and contribute to the
strength and may act as obstacles for grain growth [68].
The microstructure of ARB processed materials is similar to that of materials processed by other
severe plastic deformation methods. They feature ultra-fine grains surrounded by irregular
shaped boundaries with high-angle misorientations. The most prominent feature characterizing
the microstructure of ARB processed materials is elongated grain morphology [69].
2.5 Twist extrusion
In the year 2002, Beygelzimer et al. [27] introduced the twist extrusion processing technique,
illustrated in Figure 6. The main idea of this method is to push a bulk prismatic sample through
an extrusion die that contains a twisted channel with a certain angle around the extrusion
direction, while the shape and size of the sample are maintained without any geometrical
changes during this process. Since the die geometry affirms that the samples dimensions will not
(25)
Chapter 2 Literature Review and General Principles
change after each twist extrusion pass, it follows that this procedure can be repeated on the
same sample, and therefore plastic deformation can be accumulated to obtain grain refinement.
Similar to ECAP and HPT, the corresponding deformation mode in twist extrusion is simple shear,
however there are multiple shear planes instead of one (parallel and perpendicular to the central
axis of the sample) [70]. Furthermore, the sample cross section is deformed in a way that is similar
to that which takes place in a thin disk during HPTs [71], with vortex-like flow and large strain
gradient, stretching and mixing the metal’s particles [70].
The strain accumulated through the cross section of the sample is not homogeneous, as the
minimum strain will be on the axis of the sample, while the maximum strain will be on the
furthest most part from the axis. The minimum strain 𝜖i min, the maximum strain 𝜖i max, and the
average strain 𝜖i avg can be calculated using the following equations [71]:
Figure 6: Schematic drawing illustrates the twist extrusion processing technique and shows the sample before, during, and after the twist extrusion. Reproduced with permission of Springer-Verlag France, from reference [70]; permission conveyed through Copyrights Clearance Center, Inc.
(26)
Chapter 2 Literature Review and General Principles
where the twist angle βmax is shown in Figure 6.
During twist extrusion processing, the grain size decreases strongly by increasing the number of
twist extrusion passes, and after multiple passes, the microstructure can be refined down to the
sub-micrometer range [72, 73], forming high angle grain boundaries [74]. Moreover, in several
experiments [71] a microstructure consisting from a mixture of elongated and equiaxed grains
was reported.
As a result of the grain refinement in twist extrusion, mechanical properties were improved, and
the processed samples showed a good combination of improved strength and acceptable
ductility [71, 72, 73, 74]. On the other hand, some twist extrusion-processed samples suffered
from strong anisotropy in their mechanical properties. This problem could be eliminated by
changing the geometry of the die cross-section, increasing the processing temperature, applying
low temperature annealing, or using conventional post processing deformation such as drawing,
rolling, or direct extrusion [70, 71, 72, 75].
2.6 Characteristics of materials processed by SPD
2.6.1 Grain refinement
Grain size can be considered as one of the key microstructural properties that influence
mechanical behavior, as well as some other physical properties, in addition to chemical and
biochemical response of polycrystalline metals [76]. Because of this, improving the mechanical
properties by controlling grain size continues to be highly relevant in recent materials research.
One of the most powerful techniques that can be used for grain fragmentation and refinement
has proven to be SPD, with its various processing methods and techniques [12, 14, 20, 77].
𝝐𝒊 𝒎𝒊𝒏 = 𝟎. 𝟒 + 𝟎. 𝟏 𝒕𝒂𝒏 𝜷𝒎𝒂𝒙 (11)
𝝐𝒊 𝒎𝒂𝒙 = 𝐭𝐚𝐧 𝜷𝒎𝒂𝒙 (12)
𝝐𝒂𝒗𝒈 =
(𝝐𝒊 𝒎𝒊𝒏 + 𝝐𝒊 𝒎𝒂𝒙)
𝟐 (13)
(27)
Chapter 2 Literature Review and General Principles
Figure 7: An illustration of the grain refinement during severe plastic deformation (SPD) showing the different sequence processes; (a) early dislocations development, (b) elongated structure formation, (c)
subgrain boundaries formation, (d) blocking dislocation propagation by subgrain boundaries, (e) fine structure formation. This Figure is reproduced with permission of Elsevier S.A, from reference [78]. permission conveyed through Copyrights Clearance Center, Inc.
The mechanism of grain fragmentation and refinement by the application of plastic deformation
was comprehensively investigated for different materials by many researchers [79, 80, 78].
Furthermore, many models describing this phenomena have been suggested, with the most
accepted ones being based on the idea that the dislocation cell structure gradually transforms to
an ultrafine grained structure under large strains [76].
In general, during SPD the processes involved in grain fragmentation and refinement can be
depicted schematically in Figure 7, and summarized with the following sequence of events [76,
78, 81]. In the initial stages of straining, a structure consisting of elongated dislocation cells is
developed. With increasing applied strain, more dislocation cells and dense dislocation walls are
produced, which subdivide the original coarse grains into more refined blocks, and gradually
transform to subgrain boundaries. With further straining, the dislocations are blocked by the
subgrain boundaries, as they are considered barriers to dislocation propagation, and the
(28)
Chapter 2 Literature Review and General Principles
misorientation between adjacent (sub)grains continue to increase because of such dislocation
incorporation. As a consequence, the subgrain boundaries convert to boundaries with large
misorientations.
2.6.2 Mechanical properties of nanostructured materials
One of the major features that has placed nanomaterials in the forefront of modern materials
science research in the past 20 years is their outstanding mechanical properties compared to
their coarse-grained counterparts [15, 77]. Many research efforts have demonstrated great
enhancement of hardness, as well as yield strength, and ultimate tensile strength, as a result of
strong grain refinement down to nanometer range [14, 52, 82]. On the other hand, many
nanomaterials exhibit limited ductility and very limited uniform elongation in tensile testing [16],
which is very similar to heavily strained materials produced by conventional processing
techniques, such as cold rolling and extrusion [52].
Thus, because strength and ductility are the most important and fundamental mechanical
properties for industrial applications of structural materials, considerable efforts have been
devoted to investigating and understanding the mechanical behavior and deformation
mechanisms of nanostructured materials, as well as to improve the ductility of ultrafine-grained
materials (UFG) without sacrificing their outstanding high strength.
2.6.3 Deformation mechanisms
Based on the concept of dislocations as crystal defects, proposed by Taylor [83, 84], Orowan [85,
86, 87], and Polanyi [88], the plastic deformation in coarse grained materials occurs due to the
motion and slip of large number of dislocations as they are the basic carrier of plastic flow. The
nanocrystalline materials processed by SPD undergo grain refinement processes and result in a
grain size in the sub-micro- or the nano- range. Upon that, the percentage of atoms in grain
boundaries increases dramatically. Therefore, it is hardly expected that intragranular dislocation
sources can operate, and pile-up can exist in such materials [89]. New deformation mechanisms
have been found to take place, and play a major role in the mechanical behavior of
nanocrystalline materials. [90].
(29)
Chapter 2 Literature Review and General Principles
In nanostructured materials, where the grain size is less than 100 nm, deformation is essentially
controlled by partial dislocation activity [52, 89]. Since the resulting grains after SPD have non-
equilibrium grain boundaries [12, 91], and their grain size is typically between 50 nm and 100 nm,
the grain boundaries emit partial dislocations leading to the formation of deformation twins [89,
92], which was proven through the use of high resolution TEM [93].
Furthermore, with such fine grains, the number of grain boundaries per unit volume increases,
and new deformation mechanisms, as grain boundary sliding and grain rotation, are expected to
become more important [52]. Hahn et al. [94, 95] proposed a model for the deformation of
nanocrystalline material based on grain boundary sliding and the formation of mesoscopic glide
planes. Many researchers reported such mesoscopic shear planes in different materials using
atomic force microscopy (AFM) [96, 97], while Chinh et al. [98] demonstrated experimental
evidence of grain boundary sliding in pure aluminum processed by ECAP. On the other hand, the
random grain orientation distribution, and the absence of a rolling texture in a nanocrystalline
palladium sample after cold rolling, in comparison with the coarse grained palladium, provides
strong evidence for the importance of grain boundary rotation during the deformation of
nanocrystalline metals. [99].
2.6.4 Strengthening mechanisms
As discussed previously, plastic deformation occurs due to the motion of a large number of
dislocations. Therefore the strength of any material depends on the ability of dislocations to
move. This means that the easily moving dislocations will result in a soft and weak material, while
constrained motion of dislocations will imply that higher stress is required to initiate the plastic
flow, which leads to increase of the hardness and the strength of that material. Hence, the
strengthening mechanisms, in general, depend on suppressing and impeding the movement of
dislocations [100].
There are four main mechanisms used to strengthen metals: grain size strengthening, solid
solution strengthening, precipitation hardening, and work hardening [100]. In general, any
mechanism which hinders the mobility, the generation, and the multiplication of dislocations can
be considered as strengthening mechanism. Therefore, the transformation hardening in dual
(30)
Chapter 2 Literature Review and General Principles
phase (DP) and transformation induced plasticity (TRIP) steels is a strengthening mechanism too.
Nevertheless, due to the huge grain refinement in metals processed by severe plastic
deformation (SPD), grain size reduction is considered to be the main strengthening mechanism
as it has a huge effect on the mechanical behavior of materials [12, 101], while dislocation storage
and precipitation of nanoparticles inside the grains can be used to further improve strength and
ductility of such materials.
In general, during the plastic deformation of polycrystalline materials, dislocations are unable to
cross the grain boundaries and pile up. During their movement, motion direction will change due
to the different grain orientations, and discontinuity in the slip planes will take place between
the two adjacent grains. Therefore, grain boundaries are considered as barriers because they
block the motion of dislocations [100].
With refinement of the grain size, the average number of dislocations per grain decreases, and
the total grain boundary area increases. A high density of grain boundaries means more barriers
for the movement of dislocations, which requires higher stresses. Hence, by decreasing the grain
size, the yield strength increases according to the Hall-Petch relationship [102, 103]:
𝝈𝒚 = 𝝈𝒐 + 𝑲𝒅𝒅−𝟎.𝟓 (14)
where σy is the yield strength, σo and Kd are constants depending on the material, and d is the
average grain size diameter.
Another strengthening mechanism is solid solution hardening. When alloying elements are added
to a metal in amounts less than the solubility limit, either substitutional or interstitial solid
solution is formed. Substitutional solute atoms with a smaller diameter than that of the base
metal will cause tensile stresses in the lattice, while larger solute atoms will cause compressive
stresses. Such strains and stresses will cause lattice distortion, and therefore, will restrict the
motion of dislocations resulting in higher strength [100]. The critical shear stress required to
move dislocations through the stress field of solute atoms can be estimated from the following
equation based on Fischer’s formula [104]:
∆𝝉 = 𝑮𝒃√𝒄𝝐𝟑/𝟐 (15)
(31)
Chapter 2 Literature Review and General Principles
where G is the shear modulus, b is the magnitude of the burger’s vector, c is the concentration
of the solute atoms and 𝜖 is the strain on the base material caused by the solute atoms.
For alloying element concentrations higher than the solubility limit, the particles of the second
phase will precipitate. If the particle size is smaller than a critical value, dislocations will cut
through it, but with larger particles sizes, dislocations will penetrate the array by bowing between
the obstacles [105]. In both cases, the movement of dislocations will be impeded, and as a result,
the strength will be increased.
The fourth main strengthening mechanism is strain or dislocation hardening, where a ductile
metal becomes harder as it is plastically deformed due to dislocation storage [83]. After
exceeding the yield stress, the dislocation movement will be blocked by some obstacles, and
more stress is required to proceed with the deformation [106]. During plastic deformation, the
number of dislocations rapidly increases, new dislocations nucleate from grain boundaries,
intragranular sources, as well as from surface defects. By increasing the number of dislocations,
the average distance between them decreases and their interaction with each other increases,
hindering their mobility. Thus more stress is required to proceed with the deformation [100].
2.6.5 The strength-ductility paradox
As mentioned previously, strength and ductility are the most important mechanical properties,
but in general, they oppose each other such that higher strength is usually accompanied by lower
ductility [16, 107]. In most cases, nanostructured materials follow this behavior, as they always
possess high strength and hardness, suffer from the poor and limited ductility [82, 101], and in
many cases, fail very soon after the onset of yielding [108].
In contrast to the limited ductile behavior taking place on the macroscopic scale, nanostructured
materials show ductile behavior on the microscopic scale, recognizable by the dimpled formation
on fractured surfaces [77, 109]. In general such dimples have much larger sizes than the average
grain size [101].
The high strength and limited ductility behavior of nanostructured materials can be predicted by
understanding the mechanism of plastic deformation; by impeding the motion of dislocations,
(32)
Chapter 2 Literature Review and General Principles
the strength will increase, however ductility of the sample will decrease. The low tensile ductility
can also be attributed either to tensile instability (necking), which takes place at the maximum
load during tensile testing, when the decrease in the cross sectional area is faster than the
increase in strain hardening, or to the crack nucleation and propagation instability, which takes
place when the applied stress concentration on a flaw exceeds the maximum toughness of the
material [110].
Chung et al. [111] stated that tensile instability is affected by the strain hardening rate in such a
way that higher rates of strain hardening will delay necking. Therefore, by improving the strain
hardening rate, which can be controlled by some parameters like grain size, dislocations density,
low/high angle boundaries, the tensile ductility can be improved.
In the past years, considerable efforts have been devoted to devising innovative methods and
techniques that can improve the tensile ductility and the uniform elongation of nanocrystalline
materials, including those prepared by SPD, in order to make them useful for structural and
engineering applications; where ductility is of great importance and plays a major role.
In this concern, Wang et al. [112] proposed the use of bimodal or multimodal grain size
distribution in order to improve the tensile ductility without sacrificing the high strength. In this
technique, a certain thermomechanical treatment was used to process a pure copper sample,
resulting in micrometer grains embedded in a matrix of ultrafine grains. Hence, the ultrafine
grains, according to Hall-Petch relationship (equation (14)), provide the expected high strength
due to fine grain size, while the micrometer size grains provide strain hardening capacity through
inducing strain hardening mechanisms, which delay the instability and improve tensile ductility
[112, 113].
The strain rate can also affect uniform tensile elongation as it suppresses dynamic recovery and
improves the strain hardening rate [108]. The contribution of this parameter is known to be small,
but for strain rate-sensitive materials, the strain rate hardening can offer additional help, as
higher values of strain rate sensitivity will contribute to delay instability and improve the uniform
elongation. This is also recognized by Hart’s criterion [108, 114]:
(33)
Chapter 2 Literature Review and General Principles
𝟏
𝝈(
𝝏𝝈
𝝏𝝐)
�̇�+ (
𝝏 𝒍𝒐𝒈 𝝈
𝝏 𝒍𝒐𝒈 �̇�)
�̇�
− 𝟏 ≤ 𝟎 (16)
where the first term is the normalized strain hardening rate, and the second term is the strain
rate sensitivity.
Another method of improving the ductility of nanostructured materials was proposed by Valiev
et al. [115], by repeatedly processing a copper sample using the ECAP technique until the fraction
of high-angle grain boundaries was enough to facilitate grain boundary sliding and grain
rotations; leading to high strength as well as high ductility. In the past ten years similar behavior
has been reported by many researchers using different materials (aluminum, copper, nickel, and
titanium) processed by different SPD techniques, such as ECAP, HPT, and ARB, where high
strength accompanied with large enough ductility were achieved [52].
Moreover, strength and ductility can be improved through the formation of second phase
particles in the nanostructured matrix [77, 82]. A good combination of high strength and ductility
was reported for ultrafine grained C-Mn steel [116], as well as for bainitic steel [109], due to the
presence of second phase particles, as they help the dislocations to accumulate due to a strong
pinning effect at high strains in addition to suppressing the dynamic recovery, therefore
improving the ductility [117].
There are several other methods that can be used to improve the ductility of nanostructured
materials, while at the same time maintaining high strength, such as low temperature annealing
[118], the use of the pulse electrodeposition (PED) technique to synthesize high density
nanoscale twins in the sample [119], or even by controlling the deformation parameters such as
deformation temperature, rather than controlling the grain structure [108]. In considering the
above techniques, it is clear that high strength accompanied with high ductility in nanostructured
materials can be achieved by controlling the processing techniques, tailoring the microstructure,
or controlling the deformations parameters, making them suitable and viable for engineering and
structural applications.
Chapter 3 Experimental Procedures
(37)
Chapter 3 Experimental Procedures
3.1 Materials used in this research
Steel, as a fundamental and essential material in structural and industrial applications, continues
to be the most widely used metallic material. The ability to produce steel in large quantities with
different shapes, forms, and sizes at relatively low costs and precise specifications that cover a
wide range of physical, chemical, and mechanical properties are some of the factors contributing
to its importance.
Among steels, carbon steels are by far the most frequently used, and can be divided into four
classes as follows [120]:
1. Low carbon steels: which have a carbon content of less than 0.30%. The manganese
content varies from less than 0.40% for high formability steels which is typical for
automobile body panels, and can go up to 1.5% for rolled structural plates which is ideal
for stampings and forgings.
2. Medium carbon steel: In this class, the carbon ranges between 0.30% and 0.60% and the
manganese between 0.60% and 1.65%. Depending on the carbon and manganese content
steels in this class can be used as quenched and tempered, as well as used for shafts,
couplings and rails and rail wheels.
3. High carbon steels: these steels are suitable for spring materials and high strength wires.
They have 0.60% to 1.00% of carbon with a manganese content that varies between
0.30% and 0.90%.
4. Ultrahigh carbon steels: which are experimental alloys with 1.25% to 2.00% of carbon.
Ultrafine, equiaxed grains of ferrite with a uniformly distributed fine, spherical,
discontinuous proeutectoid carbide particle microstructure is produced by processing the
steel thermomechanically, leading to superplastic behavior.
In this study, two types of steel were used; St45 steel which falls in the low carbon steel
category, while the other one is the C45 steel, which falls in the medium carbon steel
category. The as-delivered microstructure of both steels consists of ferrite-pearlite, as
depicted in Figure 8, while Table 1 is showing the chemical composition of the two steels.
(38)
Chapter 3 Experimental Procedures
Figure 8: SEM images showing that the microstructure consists of pearlite and ferrite; (a) for as delivered St45 steel; (b) for as delivered C45 steel.
Table 1: showing the chemical composition for St45 and C45 steels in wt.%
C Si Mn S P Cr Mo Ni
St45 0.08-0.2 0.40 0.90-1.5 0.015 0.025 0.30 0.08 0.30
C45 0.42-0.5 0.24 0.74 0.024 0.011 0.17 0.05 0.12
3.2 Processing methods
3.2.1 Patenting treatment
It is well known that the initial microstructure affects the process of microstructure refinement
during HPT, as the more homogenous initial microstructure will yield to more homogenous final
microstructure, and consequently, better mechanical properties. Thus, initial thermal treatment
was applied on the as delivered ferrite-pearlite microstructure to modify it to more homogenous
one.
As mentioned previously, the as-delivered microstructure of the C45 steel is ferritic-pearlitic as
depicted in Figure 9(b), and the target is to modify this microstructure to an upper bainitic one
in order to eliminate the amount of ferrite grains. The first step in planning and setting the
thermal treatment procedure to attain this goal is to analyze the Isothermal Transformation
Diagram (IT) for the C45 steel shown in Figure 9(a) [121], also referred to as time-temperature-
(39)
Chapter 3 Experimental Procedures
transformation (TTT) diagram. The TTT diagram shows the development of microstructure when
the steel is held at a certain temperature for a certain period of time, and it could be generated
by checking the development and measuring the amount of phase formation in the
microstructure of a group of samples hold in a lead or salt bath then quenched one at a time after
increasing the holding time [122].
By checking the TTT diagram shown in Figure 9(a), the thermal treatment procedure was set as
follows:
1. Heat the C45 sample up to 900°C and hold it at this temperature for 1 hour.
2. After the first step, cool down the sample from 900°C to 450°C in less than 5 seconds by
quenching it in a molten lead bath kept at 380°C.
3. After keeping the sample at 380°C for 5 seconds, move the sample to another molten lead
bath but this time at a temperature of 450°C and hold it for 30 minutes.
As a result of the above mentioned thermal treatment, the microstructure was modified to upper
bainitic microstructure as shown in Figure 9(b), with a more homogeneous distribution of
cementite than that in the initial ferritic-pearlitic one.
Figure 9: (a) The time-temperature-transformation (TTT) diagram for the C45 steel; (b) SEM micrograph showing the bainitic microstructure of the C45 steel sample after the patenting treatment.
(40)
Chapter 3 Experimental Procedures
Figure 10: (a) Standard compression-torsion testing machine Schenck with a capacity of 400 kN; (b) specially designed Bridgmen type anvils equipped with the machine. This Figure is republished from reference [123].
3.2.2 High pressure torsion (HPT) and warm high pressure torsion (WHPT)
As mentioned previously, HPT is a severe plastic deformation technique in which samples are
subjected to torsional strain under high quasy-hydrostatic pressure [16]. To achieve this, samples
are cut into disks, and then placed between a fixed anvil and a rotating anvil under pressure as
explained comprehensively in [section 2.2.1]. The machine setup is similar to the one shown in
Figure 10 [123].
Furthermore, the processing regimes such as the processing temperature, strain, and hydrostatic
pressure have an essential and major contribution in successful HPT processing and must fulfill
certain requirements [124, 125]. The processing temperature must be below 0.4Tm where Tm is
the melting temperature. With such low temperatures, the conservative motion of dislocations
will be the dominant deformation mechanism, while using higher processing temperatures will
activate alternative mechanisms such as dislocation climb and diffusional creep [124] which will
(41)
Chapter 3 Experimental Procedures
affect negatively the grain refinement. Moreover, the samples must be processed to a true strain
not less than 6 – 8 in order to produce high angle grain boundaries. Besides, the hydrostatic
pressure must not be less than 1 GPa so as to enhance the deformability and to increase the
dislocations density by suppressing the defects annihilation.
Based on the above, the processing temperature, strain, as well as the hydrostatic pressure for
all processing procedures used in this study were selected to fulfill the aforementioned
requirements ( 𝜖>8, T<0.4Tm, P>1 GPa). Therefore, HPT was applied to 10 mm diameter and 0.7
mm thick disks of St45 and C45 steels according to the following procedures:
HPT technique was applied to St45 and C45 steel disks in as-delivered state for 4 and 6 cycles
under a quasi-hydrostatic pressure of 4.5 GPa, where one cycle means back-and-forth rotation
for 90°, inverting the rotation direction after each cycle with identical magnitudes of rotation
rate.
St45 and C45 steel samples in the as-delivered state were processed by HPT for 5 continuous
rotations of 360° each, under quasi-hydrostatic pressures of 4.5 GPa and 6 GPa.
C45 steels samples, after the patenting treatment, were processed for 5 rotations under quasi-
hydrostatic pressure of 6 GPa.
The novel processing technique of warm high pressure torsion (WHPT) was implemented on
C45 steel samples after patenting treatment for 5 and 10 rotations, under a quasi-hydrostatic
pressure of 6 GPa using temperatures of 350°C and 380°C.
3.3 Microstructure characterization
The microstructure of the samples, which is the structure observed under the microscope after
polishing and etching the surface, was investigated before any treatment (as-delivered) and after
each processing step, as well as after the in-situ tensile test using SEM, EBSD, and TEM.
3.3.1 Scanning electron microscopy (SEM)
SEM, which can detect and measure sub-micrometer features by applying a highly focused beam
of electrons to scan the sample surface and recording the scattered electrons [126], was used to
(42)
Chapter 3 Experimental Procedures
check the microstructure of the as delivered St45 and C45 samples, as well as the C45
microstructure after the patented treatment.
The samples were ground using silicon carbide grinding papers with different sizes (grit 1200,
2500 and 4000), and then polished using OP-U suspension (Colloidal silica suspension). After the
polishing, all samples were etched using Nital with 3% nitric acid concentration. The ground,
polished and etched samples were installed inside a LEO 1550 high resolution field emission
scanning electron microscope (FE-SEM) to investigate the microstructure.
3.3.2 Electron backscatter diffraction (EBSD)
EBSD is a technique used to obtain crystallographic information such as orientation mapping and
crystal type [127]. In this method, an electron beam scans the flat surface of a tilted sample inside
the SEM (as depicted in Figure 11(a)). Some of the backscattered electrons will form a certain
pattern called electron backscattered diffraction pattern (EBSP) which can be detected with a
fluorescent screen. Each crystal structure and orientation will generate different diffraction
patterns [128]. By moving the electron beam across a grid of positions on the surface of the
sample, the diffraction patterns at each location will be obtained, and then compared to a
computer database for indexing and determining the crystals orientations. Based on that, EBSD
maps can be generated, and data about the grains shapes, sizes, orientations, and boundaries
Figure 11: (a) schematic drawing for the EBSD system; (b) EBSD sample extraction through FIB.
(43)
Chapter 3 Experimental Procedures
can be obtained. A LEO 1550 (FE-SEM) equipped with INCA Crystal EBSD-System (Suite version
4.09 Issue 17b) was used.
For our EBSD analysis, FEI Quanta FEG 3D focused ion beam (FIB) was used to extract the samples
parallel to the shear direction from suitable locations near the fracture surface, as shown in
Figure 11(b). The final polishing was carried out with a 5 kV Ga+ beam to minimize FIB damage
of the sample surface.
3.3.3 Transmission electron microscopy (TEM)
TEM is a kind of microscopy technique whereby a beam of electrons penetrates a thin sample,
diffracts due to the atomic planes inside the material, generates transmission electron diffraction
patterns, and captures a high resolution image by focusing the transmitted electrons onto a
special screen, or by using a camera [129].
The microstructures of all the samples, (including the as-delivered, patented heat treated, room
temperature HPT processed, annealed, warm HPT processed, as well as the deformed samples
after the in-situ tensile test), were investigated by TEM. For conventional bright field (BF TEM),
dark field (DF TEM), and high resolution (HRTEM) images, a Philips CM20 TEM operated at 200
KV, in addition to a Philips CM30 and an FEI Titan 80 – 300 both operated at 300 KV were used.
The ACOM analysis (see [Section 3.3.3.2] below) was performed using an FEI Tecnai F20 TEM
operated at 200 KV equipped with an ASTAR system. TEM specimens were extracted and thinned
to a suitable thickness using FIB lift-out (Zeiss NVision 40 Ar and FEI Strata 400).
3.3.3.1 TEM specimen preparation using Focused ion beam (FIB)
A FIB is an instrument similar to an SEM, but in contrast to a focused beam of electrons, uses a
focused beam of ions. Accelerated ions will sputter neutral and ionized substrate atoms when
they hit the surface of a solid sample, as they will lose their energy to the atoms and the electrons
of that sample [130].
FIB was used to extract and prepare specimens for the TEM investigations as shown in Figure
12(a). For HPT processed disks, TEM specimens were extracted from a location 3 mm away from
the center of the disk, which corresponds to the midpoint of the gauge length of the dog bone
(44)
Chapter 3 Experimental Procedures
samples for the in-situ tensile test. For the sample after the in-situ tensile test, TEM specimens
were extracted from a location that is as close as possible to the fracture surface, as shown in
Figure 12(b). For all the FIB- prepared samples, the final polishing was carried out with a 5 kV Ga+
beam to minimize FIB damage of the TEM lamella.
3.3.3.2 Automated phase and crystal orientation mapping TEM (ACOM)
ACOM is an automated technique for mapping nanocrystalline phases and their orientations in a
TEM. It can be used to identify and localize phases and crystal orientations with nanometer
spatial resolution, and works as depicted in Figure 13, by collecting the electron diffraction spot
patterns through a CCD camera mounted in front of the TEM screen during the scanning and the
processing of the sample area [131]. A digital unit (precession electron diffraction device) is used
to control the primary electron beam and its precession movements.
The ACOM technique was used to characterize the microstructure of the WHPT processed
samples, and also post in-situ tensile test in order to identify the crystal orientations and the
phases, as this method is particularly suited to detect one phase dispersed in another with a good
detection threshold within the limits of its crystallinity [109]. Orientation maps have been
obtained with an acquisition speed of 100 frames per second. FIB also was used in the same
aforementioned procedure to extract and prepare the samples for ACOM. For in-situ tensile
Figure 12: (a) A micrograph using the electron beam showing the TEM sample extraction using the FIB; (b) SEM micrograph showing the location of the TEM specimen from the tensile sample after in-situ tensile test. Figure 12(b) is republished with permission of Elsevier S.A, from reference [109]; permission conveyed through Copyrights Clearance Center, Inc.
(45)
Chapter 3 Experimental Procedures
Figure 13: (a) Scanning the area of the specimen with 1 nm spots with a scan step size of tens of nm; (b) Electron diffraction spot patterns; (c) ACOM setup for FEI Tecnai F20 ST electron microscope. Republished with permission of Oldenbourg Wissenschaftsverlag GmbH, from reference [131]; permission conveyed through Copyrights Clearance Center, Inc.
samples, TEM specimens were extracted from the clamping area and the gauge length parallel
and perpendicular to the tensile axis near the fracture surface.
3.4 Mechanical properties
Mechanical properties, which are important for identifying, classifying, and fabricating materials,
are those properties related to the behavior and the reaction of these materials to external
physical loads, [100]. The most important mechanical properties are hardness, strength, and
ductility.
(46)
Chapter 3 Experimental Procedures
3.4.1 Hardness
Hardness is a very important mechanical property that can be defined as the resistance to
localized plastic deformation, such as a small indent or scratch [100]. Because this test is
considered to be simple, nondestructive, and relatively inexpensive to setup, in addition to the
fact that other mechanical properties (such as the strength) can be estimated from it, hardness
tests are performed more often than other mechanical tests.
One of the early ways to measure the hardness was the Mohs hardness test. It works by
scratching the sample with another material with a known hardness from the Mohs scale, which
ranges from 1 on the soft end for talc to 10 on the hard end for diamond. Later on, quantitative
hardness measuring techniques were developed, by which a small indenter is forced to penetrate
the surface of a sample under controlled conditions of load and loading rate, while the
penetration depth is being measured and used to calculate a certain hardness index number
depending on the technique being used. The most common hardness techniques are Brinell,
Rockwell, Vickers, and Nanoindentation. In this study, two methods were used which are Vickers
microhardness and Nanoindentation.
3.4.1.1 Vickers microhardness
Vickers hardness is a standard hardness method that was developed as a modification on Brinell
hardness in order to measure the hardness of thin samples and thin films. In this technique, a
small diamond square-based pyramid whose opposite sides meet at the apex at an angle of 136°,
Figure 14: (a) A drawing shows the shape of the square-based diamond pyramid used in Vickers hardness test; (b) A drawing shows the shape of the resulting indent and d1 and d2 are the diagonals.
(47)
Chapter 3 Experimental Procedures
as shown in Figure 14(a), is forced into the surface of the specimen by applying a load ranging
between 1 gram and 1000 grams for a certain standard time. The diagonals of the indent are
measured using optical microscope similar to Figure 14(b), and the hardness number is
determined by the ratio of the load applied on the indenter tip (P) in kilograms-force over the
surface area of the resulting indent (A) in square millimeters, where (A) can be calculated by the
formula:
𝑨 =
𝒅𝟐
𝟐 𝐬𝐢𝐧(𝟏𝟑𝟔°/𝟐) (17)
where d is the average length of the diagonals of the resulting indent.
By calculating the sin term, formula (17) can be approximated to:
𝑨 ≈
𝒅𝟐
𝟏. 𝟖𝟓𝟒 (18)
Hence,
𝑯𝑽 =
𝑷
𝑨 ≈
𝟏. 𝟖𝟓𝟒𝑷
𝒅𝟐 (19)
The units of the HV is kilograms-force per square millimeter, which are equivalent to the units of
pressure, however it is not a pressure, as the HV is calculated by dividing the load over the surface
area of the indent and not the area normal to the load.
In this study a Micromet 5103 Vickers microhardness device from the Buehler company was used
to investigate the hardness. The samples were ground and polished before testing according to
the aforementioned procedure [Section 3.3.1].
3.4.1.2 Nano-indentation
Nanoindentation is a group of indentation hardness testing techniques developed in the early
70’s to measure the hardness of small volume materials. In addition to the hardness, it is used to
extract and measure the elastic properties (modulus of elasticity) of the specimen [132]. In
nanoindentation, a load is applied to a Berkovich (three sided pyramid shown in Figure 15(a))
indenter which is in contact with the surface of the specimen. During application of the load the
(48)
Chapter 3 Experimental Procedures
depth of penetration is measured. By knowing also the geometry of the indenter, the contact
area is determined and the hardness is calculated by dividing the load by the contact area.
The load-displacement curve is generated during the loading as well as during the unloading
process throughout the experiment as shown in Figure 15(b). Due to the fact that the main part
of the displacement recovered during the unloading process is elastic, the elastic punch theory
can be applied to determine the modulus of elasticity E [133].
In the year 1992, W. C. Oliver and G. M. Pharr suggested an improved technique to determine
the hardness and the modulus of elasticity E using a load and displacement sensing indentation
experiment [134]. Since it was found that the load-displacement curves for the tested materials
were not linear during the unloading as well as at the initial stages, the use of the flat punch
approximation is therefore not suitable. Hence, the improved Oliver-Pharr method suggested a
paraboloid geometry punch, and accounts for the curvature in the unloading data to measure
the hardness and the elastic modulus within 5%.
For our measurements, a Nanoindenter XP machine from MTS was used. The hardness and the
Young’s modulus of the steel samples were determined from the load versus displacement curve
in a complete load/unload cycle. The nanoindenter was calibrated using a fused silica sample. A
Berkovich diamond indenter with total included angle of 130.6°, and initial tip radius of less than
Figure 15: (a) SEM image showing a Berkovich tip (from University of Nebraska-Lincoln website, accessed 3rd April 2017, http://bm3.unl.edu/graphics/nanoindenter_Berkovich_tip_01.jpg); (b) A schematic representation of load versus indenter displacement into the surface of the sample for nanoindentation experiment showing both curves of loading and unloading while S is the initial unloading stiffness.
(49)
Chapter 3 Experimental Procedures
20 nm, was used for all the measurements. The load and displacement were measured directly
by the machine, and the data was processed using the MTS software to produce load–
displacement curves. The mechanical properties were then calculated using the Oliver and Pharr
method.
3.4.2 In-situ tensile test
For specification and analytical purposes, the uniaxial tensile test is considered to be the most
substantial test for measuring the plastic properties of materials [135]. Through this test, yield
strength, work hardening, ultimate tensile strength, and ductility can be found, in addition to
elastic modulus and toughness.
Normally, the tensile test is carried out on bars and stripes with wide ends and a narrow middle,
commonly called “dog bone”, as shown in Figure 16. The sample is placed firmly from the wide
ends between two sets of grips, as depicted in Figure 16(a). During the test, the grips will move
at a constant speed and deform the sample. During the deformation of the sample, as a result of
the midsection (called “gauge area”) being narrower than the grip section, the applied stress
becomes highest in the middle, and subsequently the fracture and most of the strain occurs in
this area. During this test, the load is plotted versus sample elongation and converted to stress
and strain.
Figure 16: (a) Schematic drawing for a standard tensile test setup; (b) Schematic representation showing cutting position of tensile specimens from HPT disks.
(50)
Chapter 3 Experimental Procedures
The stress and the strain are calculated using the following relations:
𝝈 =
𝑷
𝑨𝒐
(20)
𝝐 =
∆𝑳
𝑳𝒐
(21)
Where:
𝝈 is the normal stress on a plane perpendicular to the longitudinal axis of the specimen.
P is the applied load.
Ao is the original cross sectional area.
𝝐 is the normal strain in the longitudinal direction.
L is the change in the specimen’s original gauge length – elongation of the specimen.
L is the original specimen’s gauge length.
The curve resulting from the above equations is called the engineering stress-strain diagram, and
it is based on the original cross sectional area and the original gauge length of the specimen.
These values continuously change during the test, however, for some materials, these changes
are negligible for all but the final stage (after the ultimate tensile strength). If the instantaneous
values of cross sectional area and gauge length are used, the true stress and true stain values can
be calculated from the following relations, and therefore, the generated curve is called the true
stress-strain diagram:
𝝈𝒕𝒓𝒖𝒆 = 𝝈𝒆𝒏𝒈(𝟏 + 𝝐𝒆𝒏𝒈) (22)
𝝐𝒕𝒓𝒖𝒆 = 𝒍𝒏(𝟏 + 𝝐𝒆𝒏𝒈) (23)
SEM in-situ tensile test are a very powerful and crucial tool as it enables researchers to
characterize and comprehensively study the tensile behavior of the investigated samples, as well
as observe the initiation and the propagation of cracks leading to the fracture of the specimen.
The in-situ tensile test was performed using a MicroDAC (Kammrath and Weiss GmbH) tensile
test stage which was installed inside a LEO 1550 high resolution field emission scanning electron
(51)
Chapter 3 Experimental Procedures
microscope (FE-SEM). The test was carried out until the fracture of the samples under constant
strain rates of 0.2 and 0.02 µm per second, equivalent to 10-4 and 10-5 s-1. This test was performed
at room temperature, 370°C, and 440°C. Dog-bone-shaped tensile specimens with a gauge length
of 1.4 mm, a width of 1 mm and thicknesses between 0.3 and 0.5 mm were cut out from HPT-
disks as shown in Figure 16(b).
Chapter 4 Experimental Results
(55)
Chapter 4 Experimental Results
4.1 As-delivered samples processed by RT HPT
Both St45 and C45 samples were cut into 10 mm diameter disks with 0.7 mm thickness, then
processed by HPT at room temperature (RT HPT) under a quasi-hydrostatic pressure of 4.5 GPa.
The St45 samples were processed for 1, 2, 4, 6, 8, and 10 cycles, while the C45 ones were
processed for 4 and 6 cycles. As mentioned in [section 3.2.2], 1 cycle means back and forth
rotation for 90°. Furthermore, the C45 samples where RT HPT processed under 4.5 GPa and 6
GPa of quasi-hydrostatic pressure for 3 and 5 continuous rotations (360°). The shear strain at a
certain point on the sample is a function of the distance of this point from the sample center, R
and represented by the following equation:
𝜸 =
𝜽𝑹
𝒉𝒑
(24)
Where is the torsional strain angle in radians, and equals to 2N in the case of using complete
rotations (N is the number of rotations), R is the distance of the point of interest from the
sample’s center, and hp is the thickness of the sample after applying the pressure.
4.1.1 Microhardness
Vickers microhardness of the processed samples was measured along their diameter and plotted
in Figure 17. By increasing the strain, either by increasing the number of cycles or by increasing
the distance from the center of the samples (according to Equation (24)), the microhardness
increased, as it is clearly seen in Figure 17(a) and (b), while the center area indicates minimum
enhancement, also correlating with equation (24). Furthermore, by comparing the hardness
results of the St45 and C45 steels depicted in Figure 17(c), it is evident that the hardness of the
C45 is higher than the St45 under the same processing conditions which is expected due to the
higher carbon content in the C45 steel.
By comparing the hardness results of the St45 steel processed by the cyclic method (each cycle
corresponds to 180° straining as described in [section 3.2.2]) with these of samples processed by
continuous complete rotations, it is found that both methods reveal hardness values around the
same level with slightly higher hardness value for the samples processed by the continuous
(56)
Chapter 4 Experimental Results
rotation HPT. The hardness after 5 rotations under 4.5 GPa was found to be 672 HV, compared
to 655 HV for the samples processed for 10 cycles (equivalent to 5 rotations) at the same 4.5 GPa
of quasi-hydrostatic pressure, as shown in Figure 17(d). The scanning electron microscopy images
reveal that continuous HPT processing leads to more homogeneous samples, due to the stronger
flow of the ferrite.
Zhilyaev et al. [28] stated that by increasing the pressure applied, there will be an increment in
the average hardness. It is then possible to obtain homogeneous microstructure throughout the
sample when the applied pressure and the torsional strain are both sufficiently high. Thus, C45
steel samples were processed by continuous rotation HPT for 3 and 5 rotations under 6 GPa of
pressure instead of 4.5 GPa. As depicted in Figure 17(e), after 3 rotations, which is equivalent to
shear strain ( 140), the hardness increased by around 2.5 times more as compared with that
of the as-delivered C45 steel hardness. It reached a value of 670 HV, while after processing the
sample under the same conditions for 5 rotations (equivalent to a shear strain 235), the
hardness was found to be only 700 HV. It is clearly seen that between 3 rotations and 5 rotations
processing, the hardness enhancement is minimal due to work hardening saturation.
4.1.2 SEM observations
All SEM samples were prepared, ground, and polished according to the procedure mentioned in
[section 3.3.1]. For both C45 and St45 steels processed using both cyclic and continuous HPT
modes, the SEM images revealed that, by increasing the strain on the samples during the
processing, the pearlite colonies become more and more deformed; elongated in one dimension
which is the shear direction, as depicted in Figure 18. Additionally, SEM images show some areas
that could not be revealed by etching (resistant to etching similar to the martensite case) (Figure
18(e)) due to the cementite dissolution as carbon normally increases the resistance to etching.
By comparing the surface structure of the C45 steel samples processed by HPT under pressure of
4.5 GPa with those samples processed under pressure of 6 GPa for the same number of rotations,
it is clearly seen that the former sample surface shows curved and bended elongated structure,
as depicted in Figure 18(f), while the later one shows more homogeneous elongated structure in
one direction without bends and curves, as presented in Figure 18(g).
(57)
Chapter 4 Experimental Results
Figure 17: (a) a diagram showing the microhardness of the cyclic HPT processed St45 steel samples through their diameter after 1, 4, and 6 cycles; (b) a diagram showing the microhardness of the cyclic HPT processed C45 steel samples through their diameter after 4, and 6 cycles; (c) bar chart comparing the microhardness of both C45 and St45 steels before and after the HPT processing for 4 and 6 cycles under 4.5 GPa of pressure; (d) bar chart showing the hardness of St45 samples after cyclic HPT processing and continuous rotations HPT processing; (e) bar chart showing the microhardness of C45 samples Processed by Continuous rotations HPT under 6 GPa of pressure for 0, 3, and 5 rotations.
(58)
Chapter 4 Experimental Results
Figure 18: (a) and (b) SEM images showing the surface structure of St45 steel after 4 and 6 cycles of HPT processing at 4.5 GPa of pressure; (c) and (d) SEM images showing the surface structure of C45 steel after 4 and 6 cycles of HPT processing at 4.5 GPa of pressure; (e) SEM image showing etching resistant area (surrounded by the ellipse) on the surface of the C45 sample processed by HPT for 6 cycles under 4.5 GPa of pressure; (f) SEM image showing the surface structure of the C45 sample processed by HPT for 5 rotations under 4.5 GPa of pressure; (g) SEM image showing the surface structure of the C45 sample processed by HPT for 5 rotations under 6 GPa of pressure.
(59)
Chapter 4 Experimental Results
Figure 19: Microhardness of RT HPT processed samples measured along the HPT sample diameter. Obtained from Dr. Jiang-li Ning, KIT.
4.2 Patented C45 samples processed by RT HPT
C45 samples were patented using the same thermal treatment explained previously in
[Section 3.2.1], then processed by HPT at Ufa State Aviation Technical University. The patented
10 mm diameter C45 disks were HPT processed at room temperature for 5 rotations under 6 GPa
of hydrostatic pressure.
Figure 20: SEM observations of the cementite morphology with HPT deformation; (a) 0 (HPT sample
center); (b) 190; (c) 310.
(60)
Chapter 4 Experimental Results
4.2.1 Microhardness
After HPT processing the patented C45 samples at room temperature for 5 complete rotations,
the microhardness was measured along the diameter of the samples. Microhardness versus
distance from the center of 4 arbitrary samples was plotted in Figure 19. The inhomogeneous
values for hardness along the diameter can be clearly seen. In samples #1 and #2, the center
shows minimal enhancement; slightly higher than the patented state value. The hardness
increases rapidly while moving away from the center towards the edges, achieving values
between 600-800 HV. This variation along the diameter correlates well with Equation (24).
However, samples #3 and #4 have low hardness that is comparable with the as-patented ones,
indicating non-successful processing, probably due to slipping occurred between the samples and
the anvils. Thus the microhardness of all samples was tested after HPT, and only samples with
hardness larger than 600 HV were chosen for mechanical tests and other investigations.
4.2.2 Microstructure
4.2.2.1 SEM observations
As predicted, the central area of the RT HPT processed samples exhibits minimal effect, as the
bainitic structure in this area was nearly unaffected, as depicted in Figure 20(a). Again, this is
compatible with Equation (24). By moving away from the center of the sample to the location of
Figure 21: TEM images showing the structure of ferrite phase after HPT processing to shear strain 190: (a) the cellular structure with SAED pattern; (b) the banded nano-granular structure with SAED pattern; (c) the corresponding dark field image in the same area with (b) created using a part of 110α and 200α rings. Republished with permission of Trans Tech Publications, from reference [136]; permission conveyed through Copyrights Clearance Center, Inc.
(61)
Chapter 4 Experimental Results
Figure 22: TEM images in an area with banded-structure after HPT processing to shear strain 190; (a) the bright field image; (b) the dark field image of the same area created using some of the cementite reflections within the 110α ring. Republished with permission of Trans Tech Publications, from reference [136]; permission conveyed through Copyrights Clearance Center, Inc.
shear strain 190 (3 mm away from the center after 3 HPT rotations), strong alignment of
cementite lamellae in the shear direction is observed, as shown in Figure 20(b). Within the
smooth and structureless area, some lamellae still can be resolved. At a strain of 310 (3 mm
away from the center after 5 HPT rotations), separate cementite lamellae are no longer resolved,
and the surface is covered with large smooth areas, elongated in the shear direction similar to
Figure 20(c)
4.2.2.2 TEM microstructure characterization
TEM measurements for those samples were performed by Dr. Di Wang from Karlsruhe Institute
of Technology (KIT) and reported in [136]. For the areas subjected to a shear strain of 190,
TEM results were compatible with SEM results as they both revealed inhomogeneous
microstructure. As depicted in Figure 21, the ferrite phase consists of a mixture of cellular and
granular grains, with some regions having relatively large cellular structure. Figure 21(a) shows
that the cells are approximately equiaxed with dimensions between 150 – 400 nm, while the
bright field TEM image depicted in Figure 21(b), and the dark field image illustrated in Figure
21(c) (created using reflections from a part of the 110α and 200α diffraction rings together and
selected by a suitably small aperture), both show some elongated grains with a width between
20 - 60 nm and a length between 50 - 150 nm. These grains are separated by dense dislocation
(62)
Chapter 4 Experimental Results
walls that are very similar to the lamellar-type boundaries in the banded structure found in
heavily deformed metals [13, 136].
At this level of strain the cementite lamellas in the same location were broken and fragmented
mainly into fine equiaxed particles, with a noticeable alignment of such fragmented cementite
particles along the shear direction, as shown in the bright field TEM image of Figure 22(a) and
the corresponding dark field TEM image of Figure 22(b). This was created using some of the
cementite reflections from a selected area in the SAED pattern within the 110α diffraction ring,
and selected by a suitably small aperture.
The bright field TEM image depicted in Figure 23(a), the dark field TEM image created using
reflections from a part of the 110α and 200α diffraction rings together, selected by a small
aperture, shown in Figure 23(b), in addition to the dark field TEM image created using some of
the cementite reflections within the 110α ring, selected by a small aperture, shown in Figure
23(c), belong to the HPT processed C45 steel sample which experienced a shear strain of 310.
It is clearly seen from Figure 23(a) and Figure 23(b) that the ferrite structure is homogeneous and
equiaxed, with an estimated mean grain size on the order of 10 nm, while the SAED patterns
illustrated in Figure 23(a) confirms that the structure is polycrystalline with nanoscaled crystallite
sizes and high angle grain boundaries. Furthermore, as depicted in Figure 23(c), it is clear that the
Figure 23: TEM images of the C45 steel after HPT processing to shear strain 310; (a) the bright field TEM image and the SAED pattern; (b) the dark TEM field image of ferrite in the same area in (a) created using a part of the reflections of 110α and 200α rings; (c) the dark field TEM image of cementite in the same area with (a) created using some of the cementite reflections within the 110α ring. Republished with permission of Trans Tech Publications, from reference [136]; permission conveyed through Copyrights Clearance Center, Inc.
(63)
Chapter 4 Experimental Results
Figure 24: Size distribution of the remaining cementite particles in the samples, (a) 190; (b) 310. Republished with permission of Trans Tech Publications, from reference [136]; permission conveyed through Copyrights Clearance Center, Inc.
cementite phase is in the form of very fine equiaxed particles distributed homogeneously in the
ferrite structure, however in this case, with no recognized alignment as in the previous case.
4.2.2.3 Grain size and grain size distribution
Languillaume et al. [137] suggested that the easiest way to measure the size distribution of
precipitates in thin foils is the Dark field imaging. Thus, by using different cementite reflections
within the 110α diffraction ring selected by a suitably small aperture, Ning et al. [136] created
several dark field images for HPT processed samples with both shear strains, 190 and 310.
Figure 25: Tensile specimens: (a,b) schematic representation showing cutting position of tensile specimens from HPT disks; (c) View of the specimens before the tensile test, special coating with TiO2 is applied on the upper surface of the specimen. Figure 25(a) is republished with permission of Elsevier S.A, from reference [109]; permission conveyed through Copyrights Clearance Center, Inc.
(64)
Chapter 4 Experimental Results
Figure 26: Cracks resulted from RT HPT in tensile specimens of C45 steel.
Figure 27: Stress/strain diagrams of RT HPT processed C45 steel samples.
The size distribution of the fragmented cementite was characterized by checking around 100
particles from approximately the same location on all samples.
The size distribution of the grains was also in agreement with the aforementioned SEM and TEM
results, as the samples with a shear strain of 190 were inhomogeneous, with coarser grains and
wider distribution, and cementite particles with sizes more than 30 nm still could be resolved, as
depicted in Figure 24(a). On the other hand, samples with a shear strain of 310 were more
homogeneous and showed a narrower distribution with finer grains, as well as the sizes of the
coarser grains decreased dramatically as it is clearly seen in Figure 24(b).
(65)
Chapter 4 Experimental Results
4.2.3 Tensile tests
A dedicated tensile stage for miniature specimens was used to perform the uni-axial tensile test
at room temperature on the above mentioned patented and HPT processed samples, as well as
Figure 28: Microstructure of C45 steel after RT HPT and subsequent annealing for two hours; (a, b) at 400°C, bright field TEM images, (c) at 400°C, bright field (left) and dark field (right) TEM images showing arrangements of dislocations, (d) HR TEM image of C45 steel after RT HPT and subsequent annealing for two hours at 400°C, (e, f) bright field TEM images after annealing at 450°C for 2 hours.
(66)
Chapter 4 Experimental Results
on the “as patented” samples using a strain rate of 5×10-4 s-1. Dog-bone shaped specimens with
a gauge length of 1.4 mm, a width of 1 mm, and a thickness of 0.3 mm were cut out from the HPT
disks, as depicted in Figure 25(a) and (b), then their surfaces were mechanically ground and
polished to eliminate surface defects as they are potential source of stress concentration and
subsequent failure. P-50 laser extensometer from Fiedler Optoelectronics was used to precisely
measure the elongation by reading TiO2 marks that had been created on the surface of the tensile
specimens, as shown in Figure 25(c). At least three tensile samples were tested for each HPT
conditions in order to obtain statistically valid results.
Most of the samples with hardness more than 600 HV were found to have cracks, which were
revealed after the preparation of the tensile specimens as depicted in Figure 26. Therefore, it
was difficult to obtain enough samples suitable for tensile testing.
Stress-strain diagrams of the RT HPT processed specimens as well as for the “as-patented” sample
were generated as shown in Figure 27. Despite the high strength of the RT HPT processed
samples, which reached a value up to 2400 MPa, it is clear that they suffer from brittleness as
they show very limited ductile behavior, and the failure occurs immediately after the plastic flow
onset.
4.2.4 Annealing the RT HPT processed samples
Because of the aforementioned brittleness characteristics, and in order to increase the low
ductility of the RT HPT processed C45 samples, to make them suitable to be used in producing
mini components for possible applications, annealing treatments at two different temperature
for two hours were carried on, the first temperature being 400°C and the second 450°C.
4.2.4.1 TEM microstructure characterization
After completing the annealing procedure, FIB was used to extract and prepare specimens for
TEM investigations from a distance of 2.5 mm away from the center of the RT HPT processed and
annealed disk.
TEM investigations of the RT HPT sample after it was annealed for two hours at 400°C revealed
that its microstructure consisted of mixture of elongated and equiaxed grains, as shown in the
(67)
Chapter 4 Experimental Results
bright field TEM images in Figure 28(a) and (b), with some grains containing a high density of
dislocations, depicted in Figure 28(c). Moreover, the HR TEM image in Figure 28(d) shows
recrystallized cementite with very good crystallinity and a strongly disordered crystalline ferrite
area. The mean grain size of ferrite was found to be 63 nm in the shear direction and 26 nm in
the radial direction, with more distinct grain boundaries.
After annealing the RT HPT processed C45 sample at 450°C for two hours, it was revealed that
the microstructure still comprised of elongated and equiaxed grains, but in this instance, some
grain growth had taken place as shown in Figure 28(e) and (f). Here the mean grain size was found
to be 213 nm in shear direction and 130 nm in radial direction.
4.2.4.2 Tensile test after annealing the RT HPT processed samples
Dog bone specimens for the tensile test were cut out from the annealed disks similar to the
drawings depicted in Figure 25(a) and (b), and prepared according to the previously explained
grinding and polishing procedure. Tensile tests were carried out on samples both having been
Figure 29: Stress/strain diagrams of RT HPT processed and annealed C45 steel samples.
(68)
Chapter 4 Experimental Results
Figure 30: (a, b, c, d) The microstructure of C45 steel after WHPT processing at 350°C for 5 rotations under 6 GPa of pressure; (a) BF TEM image taken in tangential direction, an arrow indicates shear direction at WHPT; (b) BF TEM image showing traces of cementite lamellas along the interfaces; (c) ACOM 2-phases image, red color indicates ferrite while green color indicates cementite; (d) orientation maps created in radial and tangential directions; (e) Grain size distribution obtained from a population of 260 grains. Republished with permission of Elsevier S.A, from reference [109]; permission conveyed through Copyrights Clearance Center, Inc.
(69)
Chapter 4 Experimental Results
annealed at 400°C and 450°C for 2 hours. The stress-strain diagrams for both samples are
demonstrated in Figure 29.
The sample annealed at 400°C for 2 hours showed a yield strength of 1277 MPa, UTS 1384 MPa,
and an elongation of 3%. On the other hand, the sample annealed at 450°C showed a vast
improvement in the elongation, with a value of around 11%, without any considerable effect on
the yield stress, in the range of 1077 MPa, while it shows almost similar UTS which was in the
order of 1397 MPa. Moreover, Vickers micro-hardness was performed on both samples and
found to be HV665 for the sample annealed at 400°C, and HV556 for the sample annealed at
450°C.
Nevertheless, it is clear from such obtained results that there was a strong decrease in strength
after annealing as compared with even RT HPT and WHPT processed samples. This renders the
improvement of ductility by annealing to be a rather unmotivated pursuit, as such a combination
of properties can be achieved using conventional thermal treatment.
Figure 31: The engineering and the true stress-strain curve for the WHPT processed sample at 350°C. Republished with permission of Elsevier S.A, from reference [109]; permission conveyed through Copyrights Clearance Center, Inc.
(70)
Chapter 4 Experimental Results
Figure 32: SEM images taken during the in-situ tensile test of WHPT processed C45 steel; (a) Tensile specimen immediately before fracture, no necking is recognized; (b) Surface flaws one of which (shown with large white arrow) gave an origin of the crack; (c) Shear bands near surface defects; (d) Enlarged SEM image of the surface flaw taken immediately before fracture showing development of two cracks spreading at an angle of 35° to the sample edge (shown with arrows); (e) The sample after failure showing a slant through the thickness fracture surface. Republished with permission of Elsevier S.A, from reference [109]; permission conveyed through Copyrights Clearance Center, Inc.
(71)
Chapter 4 Experimental Results
4.3 C45 samples processed by WHPT at 350°C
Patented C45 rods were cut into disks and prepared for normal HPT processing. The same
procedure for the RT HPT process was used, with the exception that, before the process started,
induction heaters were used to heat the anvils of the HPT machine up to 350°C, and maintained
this temperature throughout the processing period with an accuracy of ±5°C. Those results were
reported in [109].
4.3.1 TEM and ACOM microstructure characterization
From the BF TEM images shown in Figure 30(a) and (b), in addition to the 3D-orientation map
depicted in Figure 30(c), it is clearly seen that the microstructure of the WHPT sample processed
Figure 33: SEM images of WHPT-processed C45 sample taken at different magnifications showing different features of the fracture surface; (a) overview image of dimples; (b) enlarged view of an area in (a); (c) large oval shaped dimple with an array of inclusions on the bottom; (d) deformation marks on the wall of a large dimple resulting from shear bands. Republished with permission of Elsevier S.A, from reference [109]; permission conveyed through Copyrights Clearance Center, Inc.
(72)
Chapter 4 Experimental Results
at 350°C for 5 rotations under 6 GPa of quasi-hydrostatic pressure, consists of a mixture of
equiaxed and elongated (in shear direction) ferritic lamellas in the nanoscale size. Furthermore,
as a result of the fragmentation of the initial cementite lamella due the WHPT heavy deformation,
the ACOM 2-phase image in Figure 30(d) shows small cementite islands homogeneously
distributed in the ferrite matrix, mostly in between ferritic lamellae at the grain boundaries, with
a small distribution of cementite particles in the grains interiors, in addition to a number of
irregularly shaped and strongly fragmented cementite particles.
The secant method was applied on both radial and tangential orientation maps, which works by
counting the number of grain boundary intercepts per unit length of the secant drawn normal to
the shear direction, in order to estimate the mean lamella thickness of ferrite, which was found
to be in the range of 42 nm. Moreover, based on the ACOM map analysis, the average grain size
for the 350°C WHPT processed sample was found to be 120 nm with a standard deviation of 77
nm, measured from a population of 260 grains; Figure 30(e) shows the grain size distribution of
this sample [117]. As well, the misorientation distribution measurements confirmed that 86% of
the grain boundaries in a population consisted of 178 boundaries, were found to be high angle
boundaries with misorientation angle more than 15°.
4.3.2 In-situ tensile test
Dog bone specimens were cut from the WHPT processed samples as shown in Figure 25, then
ground and polished according to the procedure described in [section 3.3.1] in order to get rid of
surface defects. The in-situ tensile test was carried out under a constant strain rate of 0.2 μm per
a second until the fracture of the sample, while simultaneously recording the stress-strain curve,
in addition to a sequence of SEM images at different stress-strain values and different locations
on the surface with varying levels of magnification.
Figure 31 shows the obtained engineering stress-strain curve plotted with the true stress-strain
curve, calculated using Equations (22) and (23) mentioned in [section 3.4.2]. It is clearly seen from
this curve that the yield strength of this sample is around 1400 MPa, the UTS more than 2100
MPa, with a total elongation of around 4.5% (plastic elongation of 3.7%). Moreover, it is evident
that the sample deformed homogeneously with no signs of instability until the final fracture.
(73)
Chapter 4 Experimental Results
4.3.3 Observations during in-situ tensile test
As mentioned previously, the behavior of the samples was studied carefully and comprehensively
during performing the in-situ tensile test. A series of SEM images at different locations on the
surface of the samples, under different stress and strain levels were recorded at varying levels of
magnifications.
The in-situ observations during the tensile test were in agreement with the obtained stress-strain
curve (Figure 31), as the SEM images did not reveal any traces of strain instability or diffused
necking until the final fracture as shown in Figure 32(a) (taken just before fracture occurred).
Comprehensive and precise inspection of the sample’s surface using SEM revealed that, at a
stress level of around 1480 MPa and 2.1% elongation, a flaw appeared on the specimen’s surface
near the edge as depicted in Figure 32(b), followed by facilitated plastic flow near all the surface
defects of the sample. This was visualized by shear bands formation as shown in Figure 32(c).
By increasing the load during the tensile test, and therefore, moving to higher levels of stress and
strain values, SEM micrographs affirmed the developing of two cracks inclined at 35° angle on
both sides of one of the surface flaws, as depicted in Figure 32(d). Proceeding with the tensile
test the two cracks propagated for some distance, however one of them, after propagation of
about 150 µm, changed its direction to 90° (to the tensile axis), causing the failure of the sample
by a slant through the thickness fracture at a true stress level of 2174 MPa, and a true plastic
strain of 3.7% as depicted in Figure 32(e).
4.3.4 Fracture surface
As shown in Figure 33(a), the careful inspection of the fracture surface using SEM affirmed that
it was covered with very fine dimples. This indicated that the failure occurred by microvoids
coalescence, as they typically nucleate in regions of localized strain discontinuity, and, as the
strain increases, grow, coalesce, and form a continuous fracture surface [138].
As depicted in Figure 33(b), further analysis of the fracture surface’s images at higher
magnifications revealed that dimples possessed equiaxed shape. Moreover, as shown in Figure
33(c), larger microvoids were detected on the fracture surface. These were most likely formed at
(74)
Chapter 4 Experimental Results
arrays of inclusions or fine precipitates, as this is typical for ductile fracture. Figure 33(d)
illustrates a dimple with wavy lines on its wall, which are probably shear bands in the orientation
of the major stress direction.
4.3.5 Microstructure of samples in the vicinity of fracture surfaces
After the in-situ tensile test, TEM cross sections were extracted from the gauge area, and
specifically from a location near the fracture surface, using the FIB as shown previously in Figure
12(b). They were then thinned to a suitable thickness and polished according to the procedure
mentioned in [section 3.3.1]. TEM investigation revealed that neither the mean grain size nor the
current orientation of ferrite lamellae were changed (Figure 34(a)), which is unsurprising, taking
into account that the sample was fractured at a relatively low plastic strain of 3.7% (especially
when compared with the strain due to WHPT). However, dislocation density inside the grains has
notably increased (Figure 34(b)).
4.3.6 In-situ tensile test at elevated temperatures
For these experiments, dog bone samples were cut out of C45 disks processed by WHPT at 350°C,
under 6 GPa of quasi-hydrostatic pressure for 6 rotations, then ground and polished according to
Figure 34: (a) BF TEM image showing that the microstructure of the C45 steel after the in-situ tensile test had not changed, an arrow indicates shear direction at WHPT and tensile direction; (b) STEM image of dislocations in the ferrite after tensile test. Republished with permission of Elsevier S.A, from reference [109]; permission conveyed through Copyrights Clearance Center, Inc.
(75)
Chapter 4 Experimental Results
Figure 35: (a) image showing the in-situ tensile test setup. (b) Image showing the dummy sample with the thermocouple attached to its surface.
Figure 36: (a) Stress-strain diagram obtained through in-situ tensile test; (b) SEM image at a strain level of 25% showing necking as well as shear bands; (c) SEM image at a strain level of 27% and the arrows is indicating two micro-cracks; (d) SEM image showing the propagation of the crack during the in situ tensile test experiment.
(76)
Chapter 4 Experimental Results
Figure 37: (a) Stress-strain curve obtained from the in-situ tensile experiment at 370°C using a strain rate of 0.02 µm; (b) Stress-strain curve obtained from the in-situ tensile experiment at 440°C using a strain rate of 0.02 µm; (c) SEM image obtained from the 370°C tensile experiment showing necking and shear bands; (d) SEM image obtained from the 440°C tensile experiment showing necking and shear bands; (e) SEM image showing the crack propagation during tensile testing the sample at 370°C; (f) SEM image showing the final fracture after operating the tensile test at 440°C.
(77)
Chapter 4 Experimental Results
Figure 38: (a, b, c) Microstructure for the WHPT sample processed at 380°C for 10 rotations; (a) BF TEM image; (b) ACOM orientation map created in a tangential direction to the shear plane; (c) corresponding ACOM 2-phases image, red color represents ferrite while green color represents cementite; d) Grain size distribution obtained through EBSD from a population of 505 grains.
(78)
Chapter 4 Experimental Results
the aforementioned procedures to eliminate surface defects. Such defects are potential sources
for stress concentration which can affect results and cause earlier failure.
The same MicroDAC (Kammrath and Weiss GmbH) room temperature in-situ tensile test set up
was also used, as shown in Figure 35(a), however, this time after installing a special heater with
1.3 mm tip, suitable to heat miniature tensile specimens with 1.4 mm gauge length. Prior to
experimentation, temperature measurement calibration between the heater thermocouple and
actual sample temperature was performed through preparation of a dummy sample, with an
external thermocouple attached directly to its surface, as depicted in Figure 35(b). The in-situ
tensile stage was installed inside the SEM, the special heater was turned on, and the readings of
the heater’s thermocouple were compared with the readings obtained directly from the dummy
sample’s thermocouple.
During the experiments, the first in-situ tensile test was performed at 370°C, maintaining a
deforming rate of 0.2 µm per second, which is equivalent to a strain rate of 10-4 s-1. Figure 36(a)
displays the engineering stress-strain curve obtained from this experiment, showing a yield
strength of more than 600 MPa and a tensile strength of 655 MPa. Moreover, the total elongation
was around 30% with relatively short uniform elongation of around 9%.
In agreement with the stress-strain curve, instability and necking were clearly recognized by the
SEM images captured during the test, as well as shear bands, as depicted in Figure 36(b). The
failure began with a micro-crack appearing near the edge of the tensile sample at a strain level
of 25% followed by another micro-crack in the middle part of the sample at a strain level of 27%,
as shown in Figure 36(c). The two micro-cracks propagated, approached each other, and met at
a strain of 29%, as illustrated in Figure 36(d). This culminated in the complete failure of the
sample at a total strain of around 30%.
Another two in-situ tensile test experiments were performed using a slower deforming rate of
around 0.02 µm per second, equivalent to the strain rate of 10-5 s-1, using two different
temperatures of 370°C and 440°C during the test. The results in Figure 37, which were obtained
during the experiment, as well as the SEM observations and micrographs analysis, show the same
behavior as the previous sample. This began with the stress-strain curve’s behavior, which shows
(79)
Chapter 4 Experimental Results
relatively short uniform deformation, followed by necking and failure at around 30% of total
strain as illustrated in Figure 37(a) and (b). Furthermore, beside the strain instability, the SEM
micrographs shown in Figure 37(c) to (f) affirm similar deformation behavior, shear bands
formation, and fracture behavior in comparison with the previous sample, tested using a strain
rate of 10-4 s-1.
4.4 C45 samples processed by WHPT at 380°C
In order to obtain more homogeneous samples, the processing temperature during the WHPT
was increased from 350°C to 380°C, and the number of rotations was increased from 5 to 10,
while the quazi-hydrostatic pressure was kept at 6 GPa. After the patented C45 disks had been
processed, the aforementioned procedures were used to cut and polish the dog bone samples,
as well as extract and prepare TEM specimens, in order to inspect the mechanical properties,
investigate the fracture behavior, and characterize the microstructure. (These results are being
prepared for submission to Acta Materialia).
Figure 39: The engineering and the true stress-strain curve for the WHPT processed sample at 380°C.
(80)
Chapter 4 Experimental Results
4.4.1 TEM microstructure characterization
Based on the BF TEM image shown in Figure 38(a), and the ACOM map illustrated in Figure 38(b),
the microstructure of the sample processed by WHPT at 380°C for 10 rotations under 6 GPa of
quasi-hydro- static pressure was found to be a mixture of equiaxed and elongated (in the shear
direction) ferritic lamellas in the nanoscale. However, with this sample, the elongated lamellas
were less elongated and fewer than the ones found in the WHPT sample processed at 350°C for
5 rotations. Furthermore, the ACOM 2-phase image in Figure 38(c) shows heavily fragmented and
irregularly shaped cementite particles distributed in the ferrite matrix, mostly in between the
ferritic lamellae, which is also similar to the WHPT processed sample at 350°C but with a more
homogeneous cementite distribution.
EBSD measurement technique was implemented on the current sample. From a population of
505 grains, the average grain size was found to be 127 nm with a standard deviation of 73 nm,
which is very close to the WHPT sample processed at 350°C but with a more homogeneous and
narrower grain size distribution, as depicted in Figure 38(d). Moreover, the misorientation
measurements for this sample also confirms that 86% of the grain boundaries were found to be
high angle boundaries with misorientation angle more than 15°.
4.4.2 In-situ tensile test
Similar to the previous WHPT sample, the in-situ tensile test samples were prepared according
to the same procedure, and the test was carried out under the same conditions, maintaining the
same constant strain rate of 0.2 μm per a second until the fracture of the sample. A series of SEM
images at different locations and magnifications, as well as the stress-strain curve, were recorded
during the test.
This time Equations (22) and (23) were also implemented on the recorded engineering stress-
strain curve to calculate the true stress and strain values. As illustrated in Figure 39, the yield
strength of this sample shows considerable improvement, with more than 2300 MPa, the UTS
shows good improvement at 2600 MPa, while there was some drop back in the ductility with a
total elongation of 2.8%.
(81)
Chapter 4 Experimental Results
Furthermore, the stress-strain curve shows that the deformation was homogeneous until the
fracture, with no signs of strain instability or diffused necking, which is comparable with the
stress-strain curve of the WHPT sample processed at 350°C.
4.4.3 Observations during in-situ tensile test
Similar to the previous sample, the same procedure was used during the in-situ tensile test, and
a number of SEM images were recorded to study and understand the behavior of the processed
samples.
In agreement with the behavior obtained through studying the stress-strain curve (Figure 39),
SEM observations also revealed no signs of strain instability until the final separation, as shown
Figure 40: SEM micrographs showing; (a) Tensile specimen immediately before fracture, no necking is recognized; (b) Surface flaws one of which gave an origin of the crack; (c) Surface defect with no shear bands recognized; (d) the sample after failure showing a slant through the thickness fracture surface.
(82)
Chapter 4 Experimental Results
in the SEM micrograph in Figure 40(a). Moreover, surface flaws were recognized near the edges
of the tensile specimen, as illustrated in Figure 40(b), which is similar to the 350°C WHPT
processed sample behavior. In the current case the flaws were not revealed until very late into
testing, as the stress level was on the order of 2500 MPa, which was very near to fracture. Unlike
the 350°C WHPT processed sample, no signs of facilitated plastic flow were detected near any of
the surface defects as, Figure 40(c), which was recorded immediately before fracture, shows one
of the surface defects without any existence of facilitated flow or shear bands near it, for
example.
The failure initiated by a crack nucleation from one of the surface flaws near the edge, similar to
the previous WHPT sample, then propagated in a perpendicular direction to the tensile axis,
causing a slant through the thickness fracture in the same manner of the previous sample, as
depicted in Figure 40(d).
4.4.4 Fracture surface
Comprehensive inspection of the fracture surface affirmed that the failure was due to microvoids
growth and coalescence, and subsequently formed a continuous fracture surface, as the SEM
micrograph depicted in Figure 41(a) shows that this surface is covered with very fine dimples,
similar to the fracture surface of the 350°C WHPT sample.
Figure 41: (a) SEM micrograph showing that the fracture surface is covered with very fine dimples; (b) Higher magnification SEM micrographs of the fracture surface showing that the dimples have equiaxed shape.
(83)
Chapter 4 Experimental Results
Figure 41(b) is a higher magnification SEM micrograph of the fracture surface, and shows clearly
that the fine dimples have equiaxed shape. Normally equiaxed dimples form under conditions of
uniform plastic flow in the direction of applied tensile stress, and are typically produced under
conditions of tensile overload [138]. Since the fracture surfaces of both WHPT samples have
equiaxed dimples, the fracture is considered to be ductile and tough, and occurred due to stress
concentration at surface flaws.
Chapter 5 Discussion
(87)
Chapter 5 Discussion
5.1 Grain refinement and microstructure evolution during HPT
As was mentioned previously, and is similar to other SPD techniques, the grain refinement
mechanism in C45 steel during HPT processing is based on the transformation of dislocation cell
walls, which are formed during extensive plastic deformation, to high angle grain boundaries
through various dislocation activities that can be outlined as follows [78, 76, 81]: At an early stage
after applying the strain, dislocation cell structure, followed by elongated grains, will develope.
With further straining, more dislocation cells will be generated, forming dense dislocation walls
that will transform into subgrain boundaries, and therefore, subdivide the original coarse grains
into finer blocks. In continuing straining of the sample, the subgrain boundaries will work as a
barrier for dislocation propagation, as they will block them, and the misorientation between
adjacent grains will continue to accumulate until the subgrain boundaries convert to
conventional ones with large misorientation angle.
5.1.1 Room temperature HPT processing
The microstructure evolution of C45 steel during HPT followed the above mentioned dislocation-
based mechanism, as the samples processed for 3 HPT rotations show inhomogeneous
microstructure consisting of equiaxed and elongated grains, with wide grain size distribution
separated by dense dislocation walls. After 5 HPT rotations, the dislocation walls were developed
into conventional grain boundaries, and the inhomogeneous structure was transformed into a
homogeneous one with very fine and narrow grain size distribution, consisting of equiaxed
ferritic grains and separated by high angle grain boundaries.
On the other hand, before HPT processing, the carbon in the C45 sample was well partitioned
due to the initial aforementioned patenting treatment. Initially, most of the carbon atoms were
in Fe3C carbides, and less than 0.1% were in solid solution in the bcc ferrite matrix [139]. During
the heavy deformation by HPT, and since cementite is quite brittle when compared with ductile
ferrite, cementite is fragmented severely to segments and particles [136]. Therefore, after 3
rotations of RT HPT processing, the cementite lamellas were broken and fragmented mainly into
fine equiaxed particles, however there was also a notable alignment along the shear direction as
shown in Figure 22(b), which is comparable to the cementite fragments alignment along the wire
(88)
Chapter 5 Discussion
axis in cold drawn pearlitic steel [140, 141, 142, 143]. In general, the original orientation of the
cementite lamellae (before processing), with respect to the shear direction, will affect the
cementite evolution during HPT. Ivanisenko et al. [13] found that lower inclined lamellae to the
shear direction were strongly deformed, and their lamellae spacing decreased, while the spacing
of the perpendicular to shear direction lamellae would increases, and such unfavorably oriented
lamellae would become wavy and bendy. Moreover, lamellae with low inclination angle become
gradually aligned in the shear direction with increasing the strain [13, 144]. Since the aligned
cementite particles in RT HPT processed C45 samples have very close orientations to each other
(similar to those appearing in the dark field TEM image of Figure 22(b)), it is suggested that they
fragmented from the same original cementite lamellae [136]. Alignment could no longer be
recognized by further straining, and the cementite appeared in the form of very fine equiaxed
particles, spread homogeneously in the ferrite structure, as can be seen clearly in the dark field
TEM image of Figure 23(c).
Concurrently with the fragmentation, a dissolution process is taking place at an initial stage, as
Sauvage et al. [144] used the 3D atom-probe tomography (3D-APT) to study the pearlitic steel
after HPT. They found that dissolution of 40% of the cementite took place at an early stage during
HPT processing. In particular, the APT analysis, which was conducted after only one HPT rotation,
affirms two results: (i) on the cementite side, there was an off-stoichiometry cementite (along
the interfaces with ferrite) with carbon concentration in a range between 20 to 25 at.% over 10
nm thick layer, (ii) on the ferrite side, there was a carbon concentration gradient of about 2 at.%
along the ferrite/cementite interfaces, which decreases slowly to zero over a range of about 8
nm. As soon as the binding energy of carbon atoms in the cementite lattice is (0.4-0.42 eV), which
is less than the (0.5-1.0 eV) for dislocations with interstitials, carbon atoms can therefore jump
from the cementite near the interface boundaries to the dislocations in the ferrite, as this process
is energetically favorable. Then some of these atoms will diffuse along the core of dislocations
collected at the interfaces, which will cause cementite decomposition and dissolution [13, 136,
140, 141, 145, 146, 147, 148]. Furthermore, the “non-equilibrium” grain boundaries that usually
resulted after HPT processing tend to minimize their energy state, so, they are considered as
perfect segregation sites [149]. Subsequently, due to this decomposition during the HPT
(89)
Chapter 5 Discussion
processing, the amount of cementite decreases effectively until it is almost entirely decomposed
and only a few nanoclusters remain [13, 140, 150, 151, 152].
Following the careful TEM investigations and the 3D-APT analysis, X-ray diffraction (XRD)
measurements did not show any significant changes in the lattice parameters of the ferrite, even
after 5 rotations of HPT processing. This suggests that the aforementioned carbon gradient in the
cementite/ferrite interfaces was not due to the carbon atoms diffusion to interstitial spaces in
the BCC ferrite lattice, as such diffusion will be accompanied by a shift in the ferrite peaks due to
the changes in the lattice parameters [13].
Based on the above, it is evident that the decomposed cementite did not form a non-equilibrium
supersaturated carbon solid solution in ferrite, however the carbon atoms segregated into the
induced dislocations and grain boundaries, giving a pinning effect to these dislocations, and
initiated the formation of new dislocation boundaries causing further grain refinement [136]. It
is apparent that both, the dislocations grain refinement process in ferrite, and the dissolution of
cementite, are promoting each other [136], and such pinning effect of carbon atoms can explain
the finer grain size in such steel than in pure iron [13, 151].
5.1.2 Warm HPT processing
Similar to the RT HPT, dislocations mechanism is the dominant mechanism for WHPT grain
refinement. However, WHPT samples show coarser grain size which can be attributed to the
higher recovery rate that was promoted by the higher processing temperatures. During HPT
deformation, accumulation of dislocations will generate subgrain walls which then transform to
high angle boundaries by absorbing more dislocations. With the higher recovery rate, however,
dislocation annihilation within the subgrain walls by cross slip and climb will become easier and
more favorable. This will cause a drop in the number of dislocations absorbed by the subgrain
walls, which will make their evolution to high angle grain boundaries more difficult and end up
with coarser grain size [52, 56, 153, 154, 155].
Cementite in WHPT processed samples was severely fragmented and repartitioned due to the
heavy resulting strain, but unlike the RT HPT samples, in which cementite was almost fully
(90)
Chapter 5 Discussion
decomposed, TEM and ACOM investigations show that the cementite in the WHPT processed
samples consisted of irregularly shaped and strongly deformed particles, resulting from the
fracture and fragmentation of the initial cementite lamellae, located mostly at the grain
boundaries, with some particles in the grain interiors. Moreover, the area fraction of cementite
was estimated from the ACOM two-phase image of Figure 30(d) and found to be 6.2% [109]. By
taking into consideration that in the ACOM image we see only a projection of the cementite
grains in the ferrite matrix, in addition to the fact that ACOM missed some of the smallest
particles, we can say that this value can be considered high, as it is slightly lower than the
equilibrium cementite volume fraction of 7.8 vol.% (estimated from the phase diagram).
Therefore, it is suggested that the cementite dissolution during WHPT processing was limited,
showing very different behavior from C45 steel as well as from pearlitic steel processed by RT
HPT [13, 136, 144, 150].
Beside the above cementite calculations, Sauvage et al. [139] studied the WHPT processed C45
samples by means of APT. The APT 3D reconstructed volume depicted in Figure 42(a) shows both
nanoscaled and 20 nm elongated carbides, in addition to some diffused segregation along some
grain boundaries. Additionally, these carbides were found to be partially dissolved with carbon
concentration in the range of 15-20 at.%, as depicted in the composition profile computed across
the carbide/matrix interface of Figure 42(b). This is below the 25 at.% expected for stoichiometric
cementite (Fe3C carbide), however, the crystal structure of these carbides is similar to that of
Fe3C as they are detected well by ACOM. Following this partial decomposition, the released
carbon atoms into the matrix segregate along defects, such as boundaries, where the local
concentration was found to be up to 2 at% [139] (see the concentration profile Figure 42(c)).
In general, the diffusion process affects the rate of segregation, and when carbon atoms (in our
C45 steel samples) are able to diffuse faster through the matrix, the segregation process will
speed up and vice versa. Hence, the higher temperature during WHPT will increase the rate of
the segregation process. On the other hand, the equilibrium of the system is determined by its
total free energy, and by increasing the temperature the entropy will also increase, thus,
unsegregated impurities dispersed randomly throughout the grains will become more favorable,
and the rate of the segregation process will be reduced [156]. This thoroughly explains the limited
(91)
Chapter 5 Discussion
cementite dissolution in the WHPT C45 processed samples, in comparison with the RT HPT
processed ones, and justifies the differences in the cementite volume fractions.
Based on the above, it is clear that the higher processing temperature improved the recovery
rate, which sped up the dislocation annihilation process, and had a significant contribution in the
resulting coarser grain sizes obtained through WHPT processing. Furthermore, the limited
cementite dissolution (also caused by the higher processing temperature), means that fewer
carbon atoms are segregating to the dislocations and grain boundaries. This can reduce the
pinning effect on strain induced dislocations, and therefore, will affect the grain refinement
process and ending up with coarser grains.
To sum up, the grain refinement mechanism and microstructure evolution in response to both
RT HPT and WHPT processing can be outlined as follows:
Figure 42: (a) APT 3D reconstructed volume showing the heterogeneous distribution of carbon atoms (red color). (b) Carbon concentration profile across a carbide/matrix interface (direction labeled A on image (a), sampling volume thickness 2 nm). (c) Carbon concentration profile across a segregation along a boundary (direction labeled B on image (a), sampling volume thickness 2 nm). Republished with permission of John Wiley and Sons, from reference [139]; permission conveyed through Copyrights Clearance Center, Inc.
(92)
Chapter 5 Discussion
At the very beginning stage of HPT processing, and after applying the strain, dislocation cell
structure will be developed in the sample followed by the formation of elongated structure. At
this early stage, and due to the energy of mechanical strain, cementite lamellae will become
fragmented and elongated in the shear direction, and 40% of the cementite will dissolute during
this stage, as carbon atoms will jump from the cementite boundaries interface to nearby
dislocations in the ferrite matrix [144], while those diffused carbon atoms will be trapped by the
developed dislocation cell boundaries. Otherwise, for the WHPT sample, and because of the
elevated temperatures, the dissolution of cementite will be limited, as the segregation of carbon
atoms to dislocations and grain boundaries will become unfavorable due to the increase in the
entropy [156]. By further straining of the room temperature processed samples, the cementite
dissolution will continue, and the outcoming carbon atoms will frequently pin the newly
generated dislocations, which will initiate new cell boundary formation at a closed distance to
each other. At the same time, non-pinned dislocations will be trapped by those boundaries,
increasing misorientation between neighbor cells and resulting in finer grain size [144]. On the
other hand, for WHPT processed samples, and since the cementite dissolution is limited, the
pinning effect will not be so significant. Besides, the faster recovery rate and dislocations
annihilation processes will also reduce the number of dislocations trapped by the dislocation cell
boundaries, leading to coarser grains.
5.2 Mechanical behavior
5.2.1 Strengthening mechanisms
Whether the sample is processed at room temperature or elevated temperature, all obtained
results after HPT show a significant increase in the yield strength, as well as in the UTS. In the
room temperature processed samples, high dislocation density induced by severe deformation,
segregated C atoms, in addition to the few remaining nanoscale cementite particles, all
contributed to the obtained high strength. Nevertheless, the grain size strengthening
mechanism, according to Hall-Petch relationship (equation (14)), has the largest effect as due to
the finer grains, the total number of grain boundary area increases. Therefore it provides more
barriers to the movement of dislocations, meaning higher stress is required for deformation. On
(93)
Chapter 5 Discussion
the other hand, RT HPT processed samples suffer from considerably limited ductility during
tensile tests, as they fail immediately after the plastic flow onset. Such extremely brittle behavior
is not surprising and can be attributed to the high, or even nearly saturated, dislocation density
[157, 158], in addition to the poor dislocation accumulation and storage capability inside such
extremely fine grains [112, 159].
The WHPT processed samples exhibited almost the same level of UTS, but with better ductile
behavior, which is also unsurprising as the elevated temperature increased the recovery rate,
and therefore, reduced the dislocation density, which was in agreement with our microstructure
results. Besides, the larger grain sizes, in comparison with RT HPT samples, provide more space
within the grains for significant numbers of dislocations to intersect with each other, and
consequently, accumulate during deformation [160]. Hence, with more than 2100 MPa of UTS
and 4.5% of total elongation obtained by the sample processed at 350°C, and 2600 MPa of UTS
with 2.8% of total elongation obtained by the sample processed at 380°C, it is clear that strength
and ductility are better than those obtained after RT HPT, and even comparable with some grades
of AHSS as shown in Table 2.
Similar to RT HPT samples, the high strength achieved after WHPT processing can be induced by
grain size strengthening, as the WHPT samples were found to have fine grain size. Moreover, the
C45 steel is not high alloyed steel, in addition, the cementite particles are located mainly along
the grain boundaries of ferrite, as depicted in Figure 30(b) and (d) and also in Figure 38(c).
Subsequently, they contribute in a way similar to grain boundaries.
Table 2: Comparison of mechanical properties of various high strength steels and SPD processed steels.
Material and processing UTS, MPa Elongation, % Reference
C45
bainite 800 10 present work
WHPT @ 350°C 2150 4.5 present work
WHPT @ 380°C 2600 2.8 present work
Mart 1250/1520 1500 4 - 6 [161]
Nb-micro-alloyed AHSS
as-quenched 1750 7 [162]
tempered 1500 13
Rail steel R260, ECAP 1540 - [163]
Alloyed ferritic steel, HPT 1850 2.5 [164]
Alloyed austenitic steel, HPT 1850 1.5 [164] Republished with permission of Elsevier S.A, from reference [109]; permission conveyed through Copyrights Clearance Center, Inc.
(94)
Chapter 5 Discussion
In order to estimate the contribution of grain size strengthening in the yield strength of the WHPT
samples processed at 350°C, the Hall-Petch relationship (equation (14) ) can be applied, where K
is a material depending constant, and characteristic length d is the average grain diameter (only
for textureless material with equiaxed granular microstructure). Since the microstructure after
WHPT is complex and consists of a mixture of equiaxed and elongated lamellae, it is unclear as
to what suitable value of characteristic length should be assigned. It was shown by Langford [143]
that for elongated structures with strong ⟨110⟩ texture, like drawn pearlite, d is equal to double
the ferrite lamellae width. Hence, because the majority of grains in the microstructure are
elongated, this equation for drawn pearlite can be used for d. By taking a value of k=12.6 MPa
mm1/2, measured for mild steel behind the yield point [165], and d= 2x42 nm, the contribution of
grain size strengthening to the yield strength is estimated to be in the range of 1375 MPa, which
is very close to the measured yield strength.
By comparing the strength of the WHPT sample processed at 380°C with the one at 350°C, it is
obvious that the former sample has higher yield strength, and higher UTS. As shown in Figure 30
and Figure 38, the microstructure of the 380°C WHPT sample is more homogeneous than the
350°C WHPT sample, with more homogeneous cementite distribution. The grain size distribution
of the 380°C WHPT sample, depicted in Figure 38(d), does not contain any grains with sizes larger
than 400 nm, while the grain size distribution of the 350°C WHPT sample depicted in Figure 30(e)
shows that some grains with sizes larger than 550 nm still exist, giving the 380°C WHPT sample
enhanced mechanical properties when compared with the 350°C samples. Moreover, ACOM
analysis revealed that most of the cementite particles are located mainly along the ferrite grain
boundaries, with fewer and finer cementite particles in the range of 5 -10 nm inside the ferrite
grains. As soon as second-phase particles are located inside the grains, they have more effect on
the strength, as they represent obstacles for dislocations according to the Orowan mechanism.
Furthermore, since SEM images of the fracture surface confirmed that the 380°C sample has
smaller dimples, indicating more homogeneous cementite distribution (will be discussed in
[section 5.2.2.3]), we can expect more particles inside the ferrite grains compared to the 350°C
sample. Accordingly, similar to what was suggested by Ning et al. [117], one cementite particle
with diameter of less than 5 nm (which cannot be detected by ACOM, as it is below its limits)
(95)
Chapter 5 Discussion
inside each ferrite grain will end up with a strength enhancement that can be estimated using
the Orowan equation:
∆𝝈𝑶𝒓 = 𝟏. 𝟐𝑮𝒃𝒇𝟏/𝟑𝒅−𝟏 (25)
where G is the shear modulus for the ferrite matrix, b is the burger vector’s length of the 1/2
<111> dislocation in ferrite, f is the cementite/ferrite volume fraction, and d is the average
cementite particle diameter. f can be calculated by:
𝒇 = (
𝒅
𝑫)
𝟑
(26)
where d is the average cementite particle diameter and is equal to 5 nm, and D is the average
ferrite grain size and is equal to 127 nm, thus f is equal to 6.1x10-5. Knowing that for carbon steel
G = 80 GPa and b = 0.248 nm, the additional strengthening will be in the range of 187 MPa, and,
taking into consideration the standard deviation of the ferrite grain size, this value can range
between 119 MPa and 440 MPa. If we keep in mind that there is a possibility for more particles
and/or with finer sizes to exist, even more strength enhancement could be expected, which can
justify, (with the more homogeneous microstructure), the additional strength in the 380°C WHPT
processed sample measured in the tensile test.
Moreover, for UFG materials, it has been reported that their strength can be considerably higher
than what predicted by the Hall-Petch relationship [166, 167], which can be attributed to the
deformation induced unusual grain boundary segregations [139, 167]. With UFG materials, it is
firmly confirmed that the intragranular motion of dislocation is the main deformation mechanism
[12, 168]. In addition, grain boundaries are considered to be a source as well as a sink for
dislocations. Hence, the segregated atoms at the grain boundaries can suppress the dislocations
generation there and higher stresses are required for dislocations emission [166, 167], which can
be considered as a major additional mechanism for strengthening the WHPT processed C45 steel
and increases the strength effectively.
5.2.2 Deformation and fracture behavior
By reviewing the obtained mechanical properties, and analyzing the SEM observations for
deformation and fracture during in-situ tensile tests for WHPT-processed, it is clear that the
(96)
Chapter 5 Discussion
deformation was uniform until the final failure of the samples, with no signs of defused necking
or strain instability while the fracture was slant through the thickness, rather than perpendicular
to the plane of the sheet.
As mentioned previously, the RT HPT samples possess a limited amount of cementite particles
due to the dissolution processes that took place during processing, while there was limited
cementite dissolution during WHPT processing. Therefore, more cementite particles could be
resolved in the resulting microstructure. When Song et al. [116] studied the effect of such
cementite on the strength of steel, it was revealed that, at high strains, these particles put a
strong pinning effect on the dislocations, giving them a chance to accumulate and reduces their
dynamic recovery.
During the tensile test for the HPT processed samples, two dynamic processes are taking place:
the process of accumulation of newly generated dislocations, as well as the process of dynamic
recovery. The competition between these two processes affect the strain hardening behavior,
and therefore, the strength and ductility. Generally, when dislocations interact with rigid
cementite particles, they try to bypass them according to the aforementioned Orowan
mechanism by bowing around them, which will cause geometrically necessary dislocations (GND)
to generate and accumulate [169]. Since there is a lower recovery rate due to the pinning effect,
in addition to the increased dislocation accumulation due to GND formation, the strain hardening
and the strain hardening rate for the WHPT samples will be improved, which additionally will
improve strength and ductility, as it will sustain uniform elongation to a higher value [117].
5.2.2.1 Absence of necking
Such behavior in WHPT processed C45 steel was not surprising, and is also comparable to some
grades of advanced high strength steels AHSS, as Nasser et al. [170] reported that in some grades
of AHSS, (such as the DP 780-HY steel), it is difficult to identify the instability point. Also, Yan [171]
stated that fracture in AHSS my occur with minimal necking, while other researchers have
demonstrated that there is a high probability of fracture occurring before necking [172], as some
of the twinning induced plasticity (TWIP) steels, such as TWIP 940 steel, as well as Hadfield steel,
fail in the uniform deformation zone before any instability takes place [111, 173].
(97)
Chapter 5 Discussion
It is known that necking can be affected by some parameters, for example, a high strain hardening
rate can delay necking during the tensile test [111, 174]. Thus, in order to understand the absence
of necking in the C45 samples processed by WHPT, it is necessary to understand when this
phenomenon will start and predict its occurrence.
In general, during the tensile test, the stress 𝝈 increases while the cross sectional area 𝑨
decreases. The tensile force 𝑭 can be written as:
𝑭 = 𝝈. 𝑨 (27)
From the above equation, a relation between the instability and the strain was derived by [175]
to predict at which strain necking will appear, beginning with the fact that if strain hardening can
compensate the decreasing area, then homogeneous deformation will be possible. On the other
hand, when the decrease of area wins over the strength increase, instability will initiate necking
formation, and the stress-strain curve shows a maximum [175]. At that point:
𝒅𝑭 = 𝒅(𝝈. 𝑨) = 𝟎 (28)
As we know, the volume of the sample during the test is constant, so:
𝒅𝑽 = 𝒅(𝒍. 𝑨) = 𝟎 (29)
where l is a defined length on the tensile sample.
At the UTS of the engineering stress-strain curve, we can rewrite equation (28) and equation (29)
as the following:
𝝈. 𝒅𝑨 + 𝑨. 𝒅𝝈 = 𝟎 (30)
𝒍. 𝒅𝑨 + 𝑨. 𝒅𝒍 = 𝟎 (31)
From the above two equations, we get:
𝝈
𝒍=
𝒅𝝈
𝒅𝒍 (32)
or
𝒅𝝈
𝒅𝝐= 𝝈 (33)
Knowing that the plastic hardening in metals can be well characterized by:
(98)
Chapter 5 Discussion
Figure 43: Curves for the natural logarithm of the true stress vs. the natural logarithm of the true strain for: a) The 350°C WHPT processed sample. b) The 380°C WHPT processed sample.
𝝈 = 𝒌𝝐𝒏 (34)
where 𝒌 and 𝒏 are the strength coefficient and the strain hardening exponent, respectively.
By combining equations (33) and (34) we get: (valid at the UTS point)
𝒅(𝒌𝝐𝒏)
𝒅𝝐= 𝒌𝝐𝒏 (35)
𝒏𝒌𝝐𝒏−𝟏 = 𝒌𝝐𝒏 (36)
𝒏𝒌𝝐𝒏
𝝐= 𝒌𝝐𝒏 (37)
or
𝝐𝒊𝒏𝒔𝒕𝒂𝒃. = 𝒏 (38)
Equation (38) shows that the strain of instability (when necking starts to form) is equal to the
strain hardening exponent. Of course, this equation can be refined to include the strain rate
sensitivity, but for common structural materials tested at room temperature its effect is very
small and negligible [174]. Therefore, by calculating the strain hardening exponent of the C45
samples processed by WHPT, prediction for the strain of instability can be estimated. To do so,
equation (34) was used, written as the following:
𝒍𝒏 𝝈 = 𝒍𝒏 𝒌 + 𝒍𝒏 𝝐𝒏 (39)
(99)
Chapter 5 Discussion
or
𝒍𝒏 𝝈 = 𝒍𝒏 𝒌 + 𝒏 𝒍𝒏 𝝐 (40)
Furthermore, equation (40) can be simplified and written as the following:
𝒚 = 𝒂𝒙 + 𝒃 (41)
where:
y = ln(), x = ln(ϵ), a = n, and b = ln(k)
Hence, for some selected regions between the yield strength and the UTS on the engineering
stress-strain curves shown in Figure 31 and Figure 39, for the WHPT sample processed at 350°C
and 380°C respectively, the true stress-strain values were found using equations (22) and (23).
Their natural logarithms were then calculated and plotted, linearly fitted as depicted in Figure 43
(a) and (b), and the slopes of the linear fits calculated. Therefore, according to the above analysis,
the slopes of the linear fits of Figure 43(a) and (b) are equal to the strain hardening exponent n.
From the above mentioned figures, the strain hardening exponents 𝒏 for the 350°C and the 380°C
samples were found to be 0.33 and 0.14 respectively. Therefore, by recalling equation (38), the
predicted strain of instability (when neck forming will start) for the 350°C and the 380°C samples
are also 0.33 and 0.14 respectively.
By checking the stress-strain curves for both samples shown in Figure 31 and Figure 39, it is clearly
seen that the total strains are much less than the strains of instability. This clarifies the absence
of necking in our samples during the in-situ tensile tests, as with such strain hardening rates and
strain hardening exponents, the samples reach the fracture stresses before the decrease of
sample areas win over the strengths increase. Therefore, they fail without necking.
5.2.2.2 Slant fracture
The SEM micrographs depicted in Figure 32(e) and Figure 40(d) confirmed that both samples
processed by WHPT failed by a slant through the thickness fracture during the in-situ tensile test.
Such slant fracture was frequently observed in some grades of AHSS steel [172, 176], and takes
place in samples that fail without necking [111], which is also similar to the fracture behavior of
(100)
Chapter 5 Discussion
both WHPT processed samples. Moreover, Hudgins et al. [177] reported that many conventional
high strength steels can fail in this mode.
Let us recall that the tensile specimen shape with a gauge 11.40.3 mm3 resembles a shape of
a thin sheet. Until now, the process of failure in thin sheets is not well understood [178, 179,
180], but in general, it is known that due to the little constraints in thin sheets, the stresses
through the thickness are minimized even when plane-strain conditions exist. Therefore,
complete slant shear fracture may occur, especially when tensile triaxial (hydrostatic) stresses
are minimal [174], as stress triaxiality was found to influence the equivalent strain at the fracture,
and particularly, in low stress triaxiality regimes [181, 182].
Due to its importance in different modern engineering applications, such as the automotive
industry, many researchers’ efforts were utilized to study thin sheet fracture processes, with
increasing attempts to reproduce experimental results through proven models, in order to use
them as failure evaluation tools. A diversity of approaches were used to understand thin sheet
fracture mechanisms; one extreme approach is based on microscopic fracture processes while
the other is based on the continuum mechanics [183].
The microscopic fracture processes approach deals with the processes taking places during
failure, such as the role of dislocations on crack nucleation, and the effect of microvoids growth
and coalescence on the failure mechanism. As an example of following this approach, the effect
of void nucleation and growth in a shear band was analyzed by Fleck et al. [184], and the impact
of void distribution, whether uniformly, clustered or nonuniformly distributed, on the strain
localization was studied in [185], while the “void sheet” formation resulting from the intense
strain localization during ductile failure was discussed by Bandstra et al. [186].
On the other hand, the continuum mechanics approach is used to predict the behavior of the
bulk material by dealing with it as a continuous mass instead of dealing with particles and grains
separately. Because of this, the continuum mechanics approach is more attractive, as no
assumptions are required, neither about microscopic details such as voids, second phase
particles, grains and grains size, nor about the processes and mechanisms taking place such as
the mechanisms of fracture initiation by inclusions cracking or debonding. As an example of
(101)
Chapter 5 Discussion
following this approach, Needleman [187] discussed the shear bands instability and its
dependence on strain hardening, strain rate sensitivity, and on the material’s response to a
change in loading path.
Since the engineering properties used for the ductile fracture modeling are determined by the
interaction of stress and strain fields with the microstructure of the material, it is clear that any
solution to this problem involves both mechanics and microstructural aspects [183].
Nevertheless, the relation between strain localization and fracture process is not well
understood, as Rice [188] stated that it is not clear if the small rupture cavities (voids) nucleated
then led to strain localization, or the plastic flow localized first and then nucleated the small
cavities. Besson [189] put an open question regarding the localization band and the generated
void sheets in his review article, asking whether the localization occurs first and then the
secondary voids are nucleated within the band, or the secondary voids produce and generate
localization.
Figure 44: from reference [23]; (a) image showing the 3-D Von Mises equivalent strain fields and final crack pattern in the material after load step 4, (b) laminography image showing the 3-D rendering of damage/void evolution and the slant fracture. Republished with permission of Elsevier S.A, from reference [190]; permission conveyed through Copyrights Clearance Center, Inc.
(102)
Chapter 5 Discussion
Recently, Morgeneyer et al. [190] used high resolution X-ray laminography, combined with digital
volume correlation, to study in situ strain localization during the failure of a thin Al alloy (2198)
sheet, in which, after the first load step, and before any damage nucleation is taking place, two
slanted (through the thickness) strain localization bands could be recognized. These bands
continued to grow by increasing the load until one of them became more distinct than the other,
as shown in Figure 44(a), and the incremental strain field became more localized in the same
slanted (through the thickness) band.
Furthermore, Morgeneyer et al. [190] also studied the void/damage evolution during loading
until fracture. At load stages one and two, no damage evolution was recognized, while at load
stages three and four, some limited damage appears within the strain localization bands. After
that, the damage was spread very fast causing a slant through the thickness fracture, as depicted
in Figure 44(b). Therefore, it is clear that the strain localization in the slanted bands occurred first,
then initiated damage and void nucleation and growth, which began at a later stage. Hence, in
order to solve and model such fractures in thin sheets, researchers must direct their work
towards simulating and reproducing such early strain localization in slanted strain bands [190].
5.2.2.3 Fracture surface
As it was shown in [section 4.3.4] and [section 4.4.4], the SEM micrographs (Figure 33 and Figure
41) revealed that the fracture surfaces of both samples were found to be covered with fine
equiaxed dimples, but with finer ones for the sample processed by WHPT at 380°C for 10
rotations. Normally, grain boundaries, dislocation pile-ups, inclusions, and second phase particles
(i.e. cementite in steel) are considered to be localized strain discontinuity zones, and can act as
nucleation sites for microvoids [138]. When vast plastic flow takes place near such particles, they
cannot deform easily with the matrix and lose coherence, causing the formation of tiny voids
which grow and form a continuous fracture surface [178]. Furthermore, the shape of the dimples
is very important as it reflects the state of stress applied to the sample. Since the dimples on both
fracture surfaces have equiaxed shape, it is clear that the failure occurred due to uniaxial tensile
overload, as such equiaxed or hemispheroidal dimples formed under conditions of uniform
plastic flow in the direction of applied tensile stress [138].
(103)
Chapter 5 Discussion
It is important to study and compare the fracture surface features as they correlate well with the
mechanical properties [191]. In our samples, cementite particles are considered to be nucleation
sites, as they cannot deform easily with the ferrite matrix, losing coherence and causing the
formation of tiny voids which grow by slipping. In general, the average size of the dimples on a
fracture surface is of great importance, as it is affected by the number and distribution of the
nucleated microvoids, such that when few nucleation sites are available, the distance between
them will be wide allowing microvoids to grow to large sizes before they coalesce. Hence, the
dimples on the fracture surface will be large. On the other hand, with a larger number of
nucleation sites, a higher void density will be created in the microstructure with a smaller
distance between them. Therefore voids, instead of growing, will coalesce with each other after
smaller strains [138, 191]. This correlates with the obtained tensile elongation, as the sample
with larger dimples sizes (WHPT processed at 350°C) showed plastic deformation of 3.7%
compared to 1.5% for the sample with finer dimples (WHPT processed at 380°C). Moreover, this
correlates with other research efforts [192, 193, 194, 195], as they found that the average
distance between the second phase particles is comparable with the average dimple size. Hence,
it is reasonable to say that the 380°C WHPT processed sample has more nucleation sites
compared to the 350°C, and therefore, has more homogeneous cementite distribution with a
lower average distance between cementite particles.
5.3 Miscellaneous
5.3.1 Continuous recrystallization
Generally, conventional (discontinuous) recrystallization, which takes place in almost all cold
worked metallic metals during annealing above a certain critical temperature, consists of three
stages: relieving the internal strains which called the recovery stage, followed by forming new
grains which called the recrystallization stage, and finally the growing of the newly formed grains
which consume the originally deformed ones, called the grain growth stage [122].
On the other hand, in fine-grained microstructures produced by severe deformation, continuous
recrystallization is more likely to take place, causing rapid recovery and homogeneous nucleation
of recrystallized grains, which quickly transformed into new grains. Until now this recrystallization
(104)
Chapter 5 Discussion
mechanism has been poorly studied [196], and continues to be a topic debated among materials
scientists [11]. Nevertheless, similar to the conventional recrystallization behavior, three stages
can be recognized during the continuous recrystallization.
As explained by Belyakov et al. [197], during the early stages of annealing of SPD-processed UFG
materials, as a result of a recovery process, high internal stresses are reduced, which is
associated with a relaxation in the boundaries of the strain-induced grains. This leads to the
formation of narrow and straight grain boundaries, which gives the chance for many grains to
transform into a great number of nucleation sites for static recrystallization at the end of this
stage (stage 1). With further heating, those nucleation sites will begin to grow, consuming the
high dislocation density in adjacent grains. However, this grain growth is limited, as the growing
grains will touch each other and stop. Thus, this recrystallization is considered to be “transient
recrystallization”, and leads to a lower density of dislocations compared to the previous stage, in
addition to sharper and clearer grain boundaries. After the transient recrystallization, the third
stage, which is conventional normal grain growth, will take place.
By checking the microstructure of the C45 samples after annealing, Figure 28(a) and (b), and
recalling the microstructure before annealing, it is clearly seen that, for both samples, annealed
at 400°C and 450°C for two hours, the average grain size did not change significantly, and
remained relatively small compared to the grain size before annealing. The grain boundaries are,
however, narrow and more distinct. Hence, it is concluded that continuous recrystallization,
which is typical for severely deformed ferritic steel [11, 196, 197], has taken place in the annealed
samples.
5.3.2 Superplastic behavior
The Superplasticity phenomenon in metals can be defined as a flow process in which
polycrystalline materials exhibit very high ductility (more than 500% of uniform elongation
without any incipient neck development) during the tensile test. Normally, superplasticity
requires small grain sizes (less than 10 m) and a high forming temperature (greater than 0.5Tm,
where Tm is the melting temperature) [12, 90, 198]. Ultrafine grains with high angle grain
boundaries are essential factors for this phenomenon as they are responsible for grain boundary
(105)
Chapter 5 Discussion
sliding being a dominant deformation mechanism in such materials, and therefore, exhibiting
superplastic capabilities [91, 199, 200].
It is confirmed that within the superplastic deformation regime, the rate of flow is inversely
proportional to the grain size squared, which means that one order of magnitude decrease in the
grain size will lead to two order of magnitude improvement in the superplastic forming rate [90,
201]. Hence, it is expected that the ultrafine-grained materials (UFG) processed by SPD may show
considerable improvement in superplastic behavior at relatively low temperatures and/or high
strain rates [202, 203]. Furthermore, grain boundary sliding (the main deformation mechanism
for superplastic behavior) is a diffusion controlled process, and with the highly non-equilibrium
grain boundaries formed by SPD, the diffusion can become faster and may lead to improved
superplastic deformation [91].
Through our attempts to investigate superplastic behavior for the 350°C warm high pressure
torsion (WHPT) processed C45 steel samples at 6 rotations, under 6 GPa of quasi-hydrostatic
pressure, a series of in-situ tensile test measurements under the SEM were carried out at 370°C
using different strain rates of 10-4 s-1 and 10-5 s-1, and additionally at 440°C using a strain rate of
10-5. It was not possible to use higher temperatures as, during the previous annealing processing,
some grain growth was observed to take place at a temperature of 450°C (see [section 4.2.4.1]).
Unfortunately, as shown in [section 4.3.6], all our tested samples did not show any low
temperature superplastic behavior at the aforementioned strain rates. Since it is well known that
carbon improves the cohesion of grain boundaries in steels [204], and taking into consideration
that during our WHPT processing it was proven through APT measurements that there was
carbon segregation along the grain boundaries of C45 steel [139], it is therefore reasonable to
expect reduction in superplastic deformation. Moreover, there are a limited number of
publications reporting true superplastic behavior for materials processed by HPT, compared with
other SPD methods. This can be explained by the small thickness within the gauge length of the
tensile sample [16]. Horita et al. [198] performed a series of tensile measurements on HPT
processed samples with higher thicknesses, and achieved higher superplastic properties than the
same samples after ECAP processing, leading to the conclusion that the previously reported weak
(106)
Chapter 5 Discussion
superplastic deformation of the HPT processed metals can be attributed to the very small gauge
thicknesses.
Chapter 6 Summary and Conclusion
(109)
Chapter 6 Summary and Conclusion
In this study, the behavior of C45 steel processed by RT HPT and WHPT was investigated based
on a series of experimental measurements, including Vickers microhardness, nanoindentation,
SEM, TEM, ACOM, EBSD, and in situ tensile test. Based on the measurements, the differences in
the microstructure evolution, the deformation behavior, the mechanical properties, as well as
the fracture behavior, were studied, explained, and discussed comprehensively.
In the first part of this study, and in order to achieve good grain refinement, a special patenting
treatment was carried out in order to modify the ferritic-pearlitic microstructure to a more
homogeneous upper bainitic one. Then, the RT HPT technique was implemented to process the
samples for 3 and 5 rotations under 6 GPa of hydrostatic pressure.
TEM investigations of the microstructure of the RT HPT processed samples confirmed that this
technique is very powerful in grain refinement. However, mechanical tests showed that the
processed samples suffer from very limited ductility, as they failed without recognized plastic
deformation. The grain refinement can be attributed to the generation and accumulation of
dislocations during severe deformation, and their transformation from dense dislocations walls
to subgrain- and, then to conventional grain boundaries with high misorientation angle through
capturing more dislocations.
Moreover, the intensive deformation caused severe fragmentation of cementite lamellae as, in
comparison with ferrite, cementite is considered to be brittle. Concurrently with fragmentation,
cementite dissolution also took place. Since the binding energy of carbon atoms in cementite
lattice is less than that of dislocations with interstitials, carbon atoms in the cementite near the
interface boundaries are able to jump to the dislocations in the ferrite and subsequently diffuse
along the core of dislocations collected at the interfaces. This causes nearly complete cementite
decomposition and dissolution, as only very few nanoscale cementite particles could be resolved
after RT HPT.
This segregation of carbon atoms into the induced dislocations and grain boundaries imparted a
pinning effect to those dislocations, and therefore, motivated the formation of new dislocation
boundaries which improved the grain refinement. With such very fine grains and saturated
dislocation density, the high strength, as well as the limited ductility, can be understood. Finer
(110)
Chapter 6 Summary and Conclusion
grains mean that the sample has more grain boundaries with which to impede the mobility of
dislocations, thus, more stress is required to maintain the dislocations motion, and therefore,
higher yield stress will be obtained. Moreover, the saturated dislocations density, in addition to
the fine grain size will limit the dislocations mobility, causing a reduction in the sample’s ductility.
Annealing processes at 400°C and 450°C for 2 hours were implemented to improve the ductility
of the RT HPT processed samples. The 400°C annealing caused continuous recrystallization with
no grain size change, and some improvement in the ductility. The 450°C annealing resulted in
limited grain growth (still consists of ultrafine grains in the sub-micrometer range) with notably
improved ductility, however, there was a considerable drop in strength (UTS decreased by half in
comparison with the non-annealed value).
Since there is huge demand and motivation to find new stronger materials to be used in different
structural and engineering applications, UFG materials continue to be one of the most promising
group as they are characterized by their outstanding strength compared to their coarse-grained
counterparts, but on the other hand, huge efforts are made to improve their extremely poor
ductility without sacrificing the high strength [205]. To do so, the WHPT technique was
implemented to process the patented C45 samples using two different temperatures, 350°C and
380°C, processed for 5 and 10 rotations, respectively, both under 6 GPa of quasi-hydrostatic
pressure. The WHPT processed samples exhibit high strength comparable to that of the RT HPT
processed samples, and at the same time, this strength was accompanied by reasonable ductility,
which makes such samples attractive and suitable for some practical applications.
Table 3: mechanical properties of C45 steel under different processing procedures.
Yield strength (MPa)
Tensile strength (MPa)
Total elongation (%)
As delivered 350 800 ≥ 16
As patented 350 800 ≥ 16
RT HPT for 5 rotations under 6 GPa 2200 1
RT HPT for 5 rotations under 6 GPa and annealed at 400°C for 2 hours
1277 1384 3
RT HPT for 5 rotations under 6 GPa and annealed at 450°C for 2 hours
1077 1397 11
WHPT at 350°C for 5 rot. under 6 GPa 1400 2100 4.5
WHPT at 380°C for 10 rot. under 6 GPa 2300 2600 2.8
(111)
Chapter 6 Summary and Conclusion
Figure 45: Flowchart summarizes the microstructure evolution during RT HPT process.
(112)
Chapter 6 Summary and Conclusion
Figure 46: Flowchart summarizes the microstructure evolution during WHPT process.
(113)
Chapter 6 Summary and Conclusion
ACOM images for the WHPT samples processed at both temperatures showed that the
microstructure consists of a mixture of equiaxed and elongated (in shear direction) ferritic
lamellae in nanoscale size. The average grain size was larger, as compared with that obtained by
RT HPT processing. This can be attributed to the higher processing temperatures as they increase
the dislocation recovery rate, and therefore, reduce the accumulation of dislocations and the
formation of finer grains.
Furthermore, ACOM images showed that small cementite particles still exist in the
microstructure, homogeneously distributed in between the ferrite matrix, with cementite/ferrite
area fraction of 6.2%. These cementite particles were located mostly at the grain boundaries,
with the existence of some particles in the grains interiors. 3D APT analysis affirmed the presence
of off- stoichiometric cementite, which indicated that a partial dissolution process was taking
place. Also, 3D APT showed carbon concentration up to 2 at.% along some grain boundaries, as
the carbon atoms, released through the partial decomposition of cementite, segregated on
dislocations and grain boundaries. This limited cementite dissolution process is very different
from that found in the RT HPT sample, and can also be attributed to the higher processing
temperatures which, increase the entropy. Because of this, the segregation rate was reduced.
With coarser grain size (in the range of 120 nm), lower dislocation density, and fine cementite
particles dispersed in the ferrite matrix, WHPT processed samples show attractive mechanical
properties which are superior to all grades of AHSS steels. The coarser grain size provides more
space within the grains for dislocations to accumulate during deformation, while the fine
cementite particles impede the motion of dislocations, and promote their accumulation
according to the Orowan mechanism. At the same time, they reduce the recovery rate during
tensile tests due to their pinning effect, which increase the strain hardening rate, and therefore,
increase the uniform elongation and delay neck formation.
Moreover, Grain boundary tailoring and engineering can improve the mechanical properties of
metals [91]. In this manner, we can consider the unusual segregation of the carbon atoms to the
grain boundaries as a grain boundary engineering process which can affect both strength and
ductility, therefore, further investigations on the structural parameters that can influence
(114)
Chapter 6 Summary and Conclusion
positively the WHPT C45 ductility is important in order to understand the contribution of each
parameter.
In-situ tensile tests have not only proven that the mechanical properties of the WHPT samples
were comparable with some AHSS grades, but also affirmed that their deformation and fracture
behaviors were comparable too. During tensile tests, they deformed uniformly with no signs of
instability or neck formation until the final fracture. This fracture was attributed to origin due to
stress concentration at one of the surface flaws. Furthermore, samples broke perpendicular to
the loading direction in a slant through the thickness fracture.
The fracture surfaces of WHPT processed samples after in-situ tensile tests were studied
comprehensively using high resolution SEM. The obtained micrographs revealed that the
surfaces were covered with dimples, which indicated that the fracture occurred due to
microvoids coalescence. Fine cementite particles, in WHPT processed samples, can cause
localized strain discontinuity under tensile load, therefore, they are considered to be nucleation
sites. By further loading, the microvoids grew and coalesced, and formed a continuous fracture
surface. The WHPT sample processed at 380°C showed a smaller dimple size compared to the
350°C processed sample, indicating more homogeneous cementite particle distribution, and a
smaller distance between them.
Some attempts to study superplastic behavior at low temperatures for WHPT samples were
carried out using SEM in-situ tensile tests at temperatures of 370°C and 440°C, at strain rates of
10-4 s-1 and 10-5 s-1. Unfortunately, superplasticity could not be achieved in WHPT-processed C45
steel samples, and this drawback is believed to be caused by the segregated carbon atoms to
grain boundaries as they prevent grain boundary sliding; the dominant deformation mechanism
in metals showing superplastic behavior.
Based on all the obtained mechanical properties that are summarized in Table 3, and the
microstructure evolution of RT HPT and WHPT that are summarized in the flowcharts of Figure
45 and Figure 46, respectively, in considering all aforementioned discussions, it is clear that using
severe plastic deformation techniques at elevated temperatures (such as WHPT), can be
considered as a promising and attractive processing technique for carbon steels. Besides the
(115)
Chapter 6 Summary and Conclusion
outstanding yield and tensile strength resulting from implementation of this technique, C45 steel
samples processed by WHPT also show reasonable ductile behavior, unlike RT HPT processed
ones, making them attractive for use in production of micro-components. Nevertheless, further
attempts and experiments should be performed in order to discover the full potential of this
technique. For example, since most of the cementite particles were located at the grain
boundaries of the ferrite, it follows that their contribution to the mechanical properties, as well
as their influence on the mechanical behavior, is minimal. Therefore, it will be worthwhile to
make some attempts to try to obtain better and more homogeneous intragranular distribution
of cementite particles.
In summary, it is affirmed that WHPT is an efficient method for grain refinement that will improve
the strength of metals effectively, and at the same time, will end up with an acceptable amount
of ductility. In this research, the yield strength of C45 steel was improved up to 6 times, the UTS
was also improved up to more than 3 times, and at the same time, good amount of ductility was
maintained (up to 4.5% of total strain).
(117)
References
[1] R. P. Feynman, "There's Plenty of Room at the Bottom," Caltech Engineering and Science, vol. 23,
no. 5, pp. 22-36, 1960.
[2] H. Gleiter, "Materials with Ultra-Fine Grain Sizes," in Proceedings of the Second Risø International
Symposium on Metallurgy and Materials Science, Roskilde, 1981.
[3] H. J. Fecht, "Nanostructure formation and properties of metals and composites processed
mechanically in the solid state," Scripta Materialia, vol. 44, no. 8-9, pp. 1719-1723, 2001.
[4] R. Valiev, "Structure and mechanical properties of ultrafine-grained metals," Materials Science
and Engineering: A, Vols. 234-236, pp. 59--66, 1997.
[5] R. Siegel, "Nanostructured materials -mind over matter-," Nanostructured Materials, vol. 4, pp.
121--138, 1994.
[6] Y. T. Zhu, T. C. Lowe and T. G. Langdon, "Performance and applications of nanostructured
materials produced by severe plastic deformation," Scripta Materialia, vol. 51, pp. 825-830, 2004.
[7] H. Gleiter, "Nanocrystalline materials," Progress in Materials Science, vol. 33, pp. 223-315, 1989.
[8] H. J. Fecht, "Defect-induced melting and solid-state amorphization," Nature, vol. 356, pp. 133-
135, 1992.
[9] R. Z. Valiev, Y. Estrin, Z. Horita, T. G. Langdon, M. J. Zehetbauer and Y. T. Zhu, "Producing Bulk
Ultrafine-Grained Materials by Severe Plastic Deformation," JOM, Vols. 33-39, p. 58, 2006.
[10] Y. Ivanisenko, R. Wunderlich, R. Valiev and H.-J. Fecht, "Annealing behaviour of nanostructured
carbon steel produced by severe plastic deformation," Scripta Materialia, vol. 49, pp. 947--952,
2003.
[11] A. Belyakov, R. Kaibyshev, Y. Kimura and K. Tsuzaki, "Recrystallization Mechanisms in Severely
Deformed Dual-Phase Stainless Steel," Materials Science Forum, Vols. 638 - 642, pp. 1905-1910,
2010.
[12] R. Valiev, R. Islamgaliev and I. Alexandrov, "Bulk nanostructured materials from severe plastic
deformation," Progress in Materials Science, vol. 45, no. 2, pp. 103--189, 2000.
[13] Y. Ivanisenko, W. Lojkowski, R. Valiev and H.-J. Fecht, "The mechanism of formation of
nanostructure and dissolution of cementite in a pearlitic steel during high pressure torsion," Acta
Materialia, vol. 51, pp. 5555--5570, 2003.
(118)
[14] A. Azushima, R. Kopp, A. Korhonen, D. Yang, F. Micari, G. Lahoti, P. Groche, J. Yanagimoto, N.
Tsuji, A. Rosochowski and A. Yanagida, "Severe plastic deformation (SPD) processes for metals,"
CIRP Annals - Manufacturing Technology, vol. 57, pp. 716--735, 2008.
[15] H. Gleiter, "Nanostructured materials: basic concepts and microstructure," Acta Materialia, vol.
48, pp. 1--29, 2000.
[16] A. P. Zhilyaev and T. G. Langdon, "Using high-pressure torsion for metal processing:
Fundamentals and applications," Progress in Materials Science, vol. 53, pp. 893-979, 2008.
[17] I. Choi, R. Schwaiger, L. Kurmanaeva and O. Kraft, "On the effect of Ag content on the
deformation behavior of ultrafine-grained Pd-Ag alloys," Scripta Materialia, vol. 61, pp. 64-67,
2009.
[18] Y. Ivanisenko, X. Sauvage, I. MacLaren and H.-J. Fecht, "Nanostructuring and dislocation of
cemintite in pearlitic steel during severe plastic deformation," in Nanoscale Phenomena -
Fundamentals and Applications, Springer Verlag, 2008.
[19] M. Dao, L. Lu, R. Asaro, J. De Hosson and E. Ma, "Toward a quantitative understanding of
mechanical behavior of nanocrystalline metals," Acta Materialia, vol. 55, pp. 4041-4065, 2007.
[20] . R. Z. Valiev, I. Sabirov, A. P. Zhilyaev and T. G. Langdon, "Bulk Nanostructured Metals for
Innovative Applications," JOM, vol. 64, no. 10, pp. 1134-1142, 2012.
[21] R. Valiev, M. Zehetbauer, Y. Estrin, H. Höppel, Y. Ivanisenko, H. Hahn, G. Wilde, H. Roven, X.
Sauvage and T. Langdon, "The Innovation Potential of Bulk Nanostructured Materials," Advanced
Engineering Materials, vol. 9, pp. 527-533, 2007.
[22] V. Segal, "Severe plastic deformation: simple shear versus pure shear," Materials Science and
Engineering: A, vol. 338, no. 1-2, pp. 331-344, 2002.
[23] P. W. Bridgman, "Effects of High Shearing Stress Combined with High Hydrostatic Pressure,"
Phsycal Review, vol. 48, pp. 825-847, 1935.
[24] N. Smirnova, V. Levit, V. Pilyugin and V. Sazonova, "Evolution of structure of F.C.C single crystals
during strong plastic deformation," The Physics of Metals and Metallography, vol. 61, no. 6, pp.
127-134, 1986.
[25] V. Segal, "Equal channel angular extrusion: from macromechanics to structure formation,"
Materials Science and Engineering: A, vol. 271, no. 1-2, pp. 322-333, 1999.
[26] Y. Saito, N. Tsuji, H. Utsunomiya, T. Sakai and R. Hong, "Ultra-fine grained bulk aluminum
produced by accumulative roll-bonding (ARB) process," Scripta Materialia, vol. 39, no. 9, pp.
1221-1227, 1998.
[27] Y. Beygelzimer, D. Orlov and V. Varyukhin, "A new severe plastic deformation method: Twist
extrusion," in Ultrfine grained materials II, Seattle, 2002.
(119)
[28] A. Zhilyaev, G. Nurislamova, B.-K. Kim, M. Baro´, J. Szpunar and T. Langdon, "Experimental
parameters influencing grain refinement and microstructural evolution during high-pressure
torsion," Acta Materialia, vol. 51, pp. 753-765, 2003.
[29] A. P. Zhilyaev, T. R. McNelley and T. G. Langdon, "Evolution of microstructure and microtexture in
fcc metals during high-pressure torsion," Journal of Materials Science, vol. 42, no. 5, pp. 1517-
1528, 2007.
[30] M. V. Degtyarev, T. I. Chashchukhina, L. M. Voronova, L. S. Davydova and V. P. Pilyugin,
"Deformation strengthening and structure of structural steel upon shear under pressure," The
Physics of Metals and Metallography, vol. 90, no. 6, pp. 604-611, 2000.
[31] M. V. Degtyarev, T. I. Chashchukhina, M. Y. Romanova and L. M. Voronova, "Correlation between
the copper structure and temperature-rate parameters of pressure-induced shear deformation,"
Doklady Physics, vol. 49, no. 7, pp. 415-418, 2004.
[32] A. Vorhauer and R. Pippan, "On the homogeneity of deformation by high pressure torsion,"
Scripta Materialia, vol. 51, no. 9, pp. 921-925, 2004.
[33] H. Jiang, Y. Zhu, D. P. Butt, I. V. Alexandrov and T. C. Lowe, "Microstructural evolution,
microhardness and thermal stability of HPT-processed Cu," Materials Science and Engineering: A,
vol. 290, no. 1-2, pp. 128-138, 2000.
[34] Z. Yang and U. Welzel, "Microstructure-microhardness relation of nanostructured Ni produced by
high-pressure torsion," Materials Letters, vol. 59, no. 27, pp. 3406-3409, 2005.
[35] G. Sakai, Z. Horita and T. G. Langdon, "Grain refinement and superplasticity in an aluminum alloy
processed by high-pressure torsion," Materials Science and Engineering: A, vol. 393, no. 1-2, pp.
344-351, 2005.
[36] Z. Horita and T. G. Langdon, "Microstructures and microhardness of an aluminum alloy and pure
copper after processing by high-pressure torsion," Materials Science and Engineering: A, Vols.
410-411, pp. 422-425, 2005.
[37] A. Zhilyaev, S. Lee, G. Nurislamova, R. Valiev and T. Langdon, "Microhardness and microstructural
evolution in pure nickel during high-pressure torsion," Scripta Materialia, vol. 44, no. 12, pp.
2753-2758, 2001.
[38] M. Zehetbauer, H. Stüwe, A. Vorhauer, E. Schafler and J. Kohout, "The role of hydrostatic
pressure in severe plastic deformation," Advanced Engineering Materials, vol. 5, no. 5, pp. 330-
337, 2003.
[39] Y. Estrin, A. Molotnikov, C. Davies and R. Lapovok, "Strain gradient plasticity modelling of high-
pressure torsion," Journal of the Mechanics and Physics of Solids, vol. 56, no. 4, p. 1186–1202,
2008.
(120)
[40] P. Mckenzie, R. Lapovok and Y. Estrin, "The influence of back pressure on ECAP processed AA
6016: Modeling and experiment," Acta Materialia, vol. 55, no. 9, pp. 2985-2993, 2007.
[41] C. Xu, Z. Horita and T. G. Langdon, "The evolution of homogeneity in processing by high-pressure
torsion," Acta Materialia, vol. 55, no. 1, pp. 203-212, 2007.
[42] . M. Kai, Z. Horita and T. G. Langdon, "Developing grain refinement and superplasticity in a
magnesium alloy processed by high-pressure torsion," Materials Science and Engineering: A, vol.
488, no. 1-2, pp. 117-124, 2008.
[43] Y. Harai, Y. Ito and Z. Horita, "High-pressure torsion using ring specimens," Scripta Materialia, vol.
58, no. 6, pp. 469-472, 2008.
[44] R. Valiev, N. Krasilnikov and N. Tsenev, "Plastic deformation of alloys with submicron-grained
structure," Materials Science and Engineering: A, vol. 137, pp. 35-40, 1991.
[45] J. Wang, M. Furukawa, Z. Horita, M. Nemoto, R. Z. Valiev and T. G. Langdon, "Enhanced grain
growth in an Al-Mg alloy with ultrafine grain size," Materials Science and Engineering: A, vol. 216,
no. 1-2, pp. 41-46, 1996.
[46] R. Z. Valiev, "Structure and mechanical properties of ultrafine-grained metals," Materials Science
and Engineering: A, Vols. 234-236, pp. 59-66, 1997.
[47] V. Segal, "Materials processing by simple shear," Materials Science and Engineering: A, vol. 197,
no. 2, pp. 157-164, 1995.
[48] K. Nakashima, Z. Horita, M. Nemoto and T. G. Langdon, "Development of a multi-pass facility for
equal-channel angular pressing to high total strains," Materials Science and Engineering: A, vol.
281, no. 1-2, pp. 82--87, 2000.
[49] M. Nemoto, Z. Horita, M. Furukawa and T. G. Langdon, "Equal-channel angular pressing: A novel
tool for microstructural control," Metals and Materials International, vol. 4, no. 6, pp. 1181-1190,
1998.
[50] S. Ferrasse, V. Segal, K. Hartwig and R. Goforth, "Microstructure and properties of copper and
aluminum alloy 3003 heavily worked by equal channel angular extrusion," Metallurgical and
Materials Transactions A, vol. 28, no. 4, pp. 1047-1057, 1997.
[51] M. Furukawa, Y. Iwahashi, Z. Horita, M. Nemoto and T. G. Langdon, "The shearing characteristics
associated with equal-channel angular pressing," Materials Science and Engineering: A, vol. 257,
no. 2, pp. 328-332, 1998.
[52] R. Z. Valiev and T. G. Langdon, "Principles of equal-channel angular pressing as a processing tool
for grain refinement," Progress in Materials Science, vol. 51, no. 7, pp. 881-981, 2006.
(121)
[53] . Y. Iwahashi, J. Wang, Z. Horita, M. Nemoto and T. G. Langdon, "Principle of equal-channel
angular pressing for the processing of ultra-fine grained materials," Scripta Materialia, vol. 35, no.
2, pp. 143-146, 1996.
[54] K. Nakashima, Z. Horita, M. Nemoto and T. G. Langdon, "Influence of channel angle on the
development of ultrafine grains in equal-channel angular pressing," Acta Materialia, vol. 46, no.
5, pp. 1589-1599, 1998.
[55] C. Xu and T. G. Langdon, "The development of hardness homogeneity in aluminum and an
aluminum alloy processed by ECAP," Journal of Materials Science, vol. 42, no. 5, pp. 1542-1550,
2007.
[56] A. Yamashita, D. Yamaguchi, Z. Horita and T. G. Langdon, "Influence of pressing temperature on
microstructural development in equal-channel angular pressing," Materials Science and
Engineering: A, vol. 287, no. 1, pp. 100-106, 2000.
[57] R. Y. Lapovok, "The role of back-pressure in equal channel angular extrusion," Journal of
Materials Science, vol. 40, no. 2, pp. 341-346, 2005.
[58] Y. Iwahashi, Z. Horita , M. Nemoto and T. G. Langdon, "The process of grain refinement in equal-
channel angular pressing," Acta Materialia, vol. 46, no. 9, pp. 3317-3331, 1998.
[59] M. Furukawa, Z. Horita and T. G. Langdon, "Factors influencing the shearing patterns in equal-
channel angular pressing," Materials Science and Engineering: A, vol. 332, no. 1-2, pp. 97-109,
2002.
[60] Y. Nishida, H. Arima, J.-C. Kim and T. Ando, "Rotary-die equal-channel angular pressing of an Al - 7
mass% Si - 0.35 mass% Mg alloy," Scripta Materialia, vol. 45, no. 3, pp. 261-266, 2001.
[61] A. Azushima and K. Aoki, "Properties of ultrafine-grained steel by repeated shear deformation of
side extrusion process," Materials Science and Engineering: A, vol. 337, no. 1-2, pp. 45-49, 2002.
[62] J.-H. Han, H.-K. Seok, Y.-H. Chung, M.-C. Shin and J.-C. Lee, "Texture evolution of the strip cast
1050 Al alloy processed by continuous confined strip shearing and its formability evaluation,"
Materials Science and Engineering: A, vol. 323, no. 1-2, pp. 342-347, 2002.
[63] J.-C. Lee, J.-Y. Shu and J. P. Ahn, "Work-softening behavior of the ultrafine-grained Al alloy
processed by high-strain-rate, dissimilar-channel angular pressing," Metallurgical and Materials
Transactions A, vol. 34, no. 3, pp. 625-632, 2003.
[64] U. Chakkingal, A. B. Suriadi and P. Thomson, "Microstructure development during equal channel
angular drawing of Al at room temperature," Scripta Materialia, vol. 39, no. 6, pp. 677-684, 1998.
[65] H. Utsunomiya, K. Hatsuda, T. Sakai and Y. Saito, "Continuous grain refinement of aluminum strip
by conshearing," Materials Science and Engineering: A, vol. 372, no. 1-2, pp. 199-206, 2004.
(122)
[66] G. J. Raab, R. Z. Valiev, T. C. Lowe and Y. T. Zhu, "Continuous processing of ultrafine grained Al by
ECAP-Conform," Materials Science and Engineering: A, vol. 382, no. 1-2, pp. 30-34, 2004.
[67] R. Z. Valiev, "Some New Trends in SPD Processing for Fabrication of Bulk Nanostructured
Materials," Materials Science Forum, Vols. 503-504, pp. 3-10, 2006.
[68] . Y. Saito, H. Utsunomiya, N. Tsuji and T. Sakai, "Novel ultra-high straining process for bulk
materials—development of the accumulative roll-bonding (ARB) process," Acta Materialia, vol.
47, no. 2, pp. 579-583, 1999.
[69] N. Tsuji, Y. Saito, S.-H. Lee and Y. Minamino, "ARB (Accumulative Roll-Bonding) and other new
Techniques to Produce Bulk Ultrafine Grained Materials," Advanced Engineering Materials, vol. 5,
no. 5, pp. 338-344, 2003.
[70] Y. Beygelzimer, V. Varyukhin and S. Synkov, "Shears, Vortices, and Mixing During Twist Extrusion,"
International Journal of Material Forming, vol. 1, no. 1, pp. 443-446, 2008.
[71] D. Orlov, V. Stolyarov, H. Salimgareyev, E. Soshnikova, A. Reshetov, Y. Beygelzimer, S. Synkov and
V. Varyukhin, "Microstructure, mechanical properties and anysotropy of pure Ti processed by
twist extrusion and cold rolling," in Ultrafine Grained Materials III, Charlotte, North Carolina,
2004.
[72] V. Varyukhin, Y. Beygelzimer, S. Synkov and D. Orlov, "Application of Twist Extrusion," Materials
Science Forum, Vols. 503-504, pp. 335-340, 2006.
[73] V. V. Stolyarov, Y. E. Beigelzimer, D. V. Orlov and R. Z. Valiev, "Refinement of Microstructure and
Mechanical Properties of Titanium Processed by Twist Extrusion and Subsequent Rolling," The
Physics of Metals and Metallography, vol. 99, no. 2, p. 204–211, 2005.
[74] M. Berta, D. Orlov and P. B. Prangnell, "Grain refinement response during twist extrusion of an Al-
0.13% Mg alloy," International Journal of Materials Research, vol. 98, no. 3, pp. 200-204, 2007.
[75] Y. Beygelzimer, D. Orlov, A. Korshunov, S. Synkov, V. Varyukhin, I. Vedernikova, A. Reshetov, A.
Synkov, L. Polyakov and I. Korotchenkova, "Features of Twist Extrusion: Method, Structures &
Material Properties," Solid State Phenomena, vol. 114, pp. 69-78, 2006.
[76] Y. Estrin and A. Vinogradov, "Extreme grain refinement by severe plastic deformation: A wealth
of challenging science," Acta Materialia, vol. 61, no. 3, pp. 782-817, 2013.
[77] M. Haddad, Y. Ivanisenko and H.-J. Fecht, "Behavior of C45 steel Subjected to Different High
Pressure Torsion (HPT) Procedures," International Journal of Applied Science and Technology, vol.
5, no. 1, pp. 144-153, 2015.
[78] A. Mishra, V. Richard, F. Gregori, R. Asaro and M. Meyers, "Microstructural evolution in copper
processed by severe plastic deformation," Materials Science and Engineering: A, Vols. 410-411,
pp. 290-298, 2005.
(123)
[79] Y. M. Wang, M. W. Chen, H. W. Sheng and E. Ma, "Nanocrystalline grain structures developed in
commercial purity Cu by low-temperature cold rolling," Journal of Materials Research, vol. 17, no.
12, pp. 3004-3007, 2002.
[80] N. Tao, Z. Wang, W. Tong, M. Sui, J. Lu and K. Lu, "An investigation of surface nanocrystallization
mechanism in Fe induced by surface mechanical attrition treatment," Acta Materialia, vol. 50, no.
18, pp. 4603-4616, 2002.
[81] K. Wang, N. Tao, G. Liu, J. Lu and K. Lu, "Plastic strain-induced grain refinement at the nanometer
scale in copper," Acta Materialia, vol. 54, no. 19, pp. 5281-5291, 2006.
[82] C. Koch, "Optimization of strength and ductility in nanocrystalline and ultrafine grained metals,"
Scripta Materialia, vol. 49, no. 7, pp. 657-662, 2003.
[83] G. I. Taylor, "The Mechanism of Plastic Deformation of Crystals. Part I. Theoretical," Proceedings
of the Royal Society of London: Series A, vol. 145, no. 855, pp. 362-387, 1934.
[84] G. I. Taylor, "The Mechanism of Plastic Deformation of Crystals. Part II. Comparison with
Observations," Proceedings of the Royal Society of London: Series A, vol. 145, no. 855, pp. 388-
404, 1934.
[85] E. Orowan, "Zur Kristallplastizität. I," Zeitschrift für Physik, vol. 89, no. 9-10, pp. 605-613, 1934.
[86] E. Orowan, "Zur Kristallplastizität. II," Zeitschrift für Physik, vol. 89, no. 9-10, pp. 614-633, 1934.
[87] E. Orowan, "Zur Kristallplastizität. III," Zeitschrift für Physik, vol. 89, no. 9-10, pp. 634-659, 1934.
[88] M. Polanyi, "Über eine Art Gitterstörung, die einen Kristall plastisch machen könnte," Zeitschrift
für Physik, vol. 89, no. 9-10, pp. 660-664, 1934.
[89] H. Van Swygenhoven, "Grain Boundaries and Dislocations," Science, vol. 296, pp. 66-67, 2002.
[90] Y. T. Zhu and T. G. Langdon, "The fundamentals of nanostructured materials processed by severe
plastic deformation," JOM, vol. 56, no. 10, pp. 58-63, 2004.
[91] R. Valiev, "Nanostructuring of metals by severe plastic deformation for advanced properties,"
Nature materials, vol. 3, no. 8, pp. 511-516, 2004.
[92] M. Chen, E. Ma, K. J. Hemker, H. Sheng, Y. Wang and X. Cheng, "Deformation Twinning in
Nanocrystalline Aluminum," Science, vol. 300, no. 5623, pp. 1275-1277, 2003.
[93] X. Z. Liao, Y. H. Zhao, S. G. Srinivasan, Y. T. Zhu, R. Z. Valiev and D. V. Gunderov, "Deformation
twinning in nanocrystalline copper at room temperature and low strain rate," Applied Physics
Letters, vol. 84, no. 4, pp. 592-594, 2004.
[94] H. Hahn, P. Mondal and K. Padmanabhan, "Plastic deformation of nanocrystalline materials,"
Nanostructured Materials, vol. 9, no. 1-8, pp. 603-606, 1997.
(124)
[95] H. Hahn and K. A. Padmanabhan, "A model for the deformation of nanocrystalline materials,"
Philosophical Magazine Part B, vol. 76, no. 4, pp. 559-571, 1997.
[96] A. Vinogradov, S. Hashimoto, V. Patlan and K. Kitagawa, "Atomic force microscopic study on
surface morphology of ultra-fine grained materials after tensile testing," Materials Science and
Engineering: A, Vols. 319-321, pp. 862-866, 2001.
[97] Y. Huang and T. G. Langdon, "Using atomic force microscopy to evaluate the development of
mesoscopic shear planes in materials processed by severe plastic deformation," Materials Science
and Engineering: A, vol. 358, no. 1-2, pp. 114-121, 2003.
[98] N. Q. Chinh, P. Szommer, Z. Horita and T. G. Langdon, "Experimental Evidence for Grain-Boundary
Sliding in Ultrafine-Grained Aluminum Processed by Severe Plastic Deformation," Advanced
Materials, vol. 18, pp. 34-39, 2006.
[99] J. Markmann, P. Bunzel, H. Rösner, K. Liu, K. Padmanabhan, R. Birringer, H. Gleiter and J.
Weissmüller, "Microstructure evolution during rolling of inert-gas condensed palladium," Scripta
Materialia, vol. 49, no. 7, pp. 637-644, 2003.
[100] W. D. Callister, Materials Science and Engineering, an Introduction, USA: John Wiley & Sons,
2007.
[101] M. Meyers, A. Mishra and D. Benson, "Mechanical properties of nanocrystalline materials,"
Progress in Materials Science, vol. 51, no. 4, pp. 427-556, 2006.
[102] E. O. Hall, "The deformation and ageing of mild steel: III discussion of results," Proceedings of the
Physical Society. Section B, vol. 64, no. 9, pp. 747-753, 1951.
[103] N. J. Petch, "The cleavage strength of polycrystals," Journal of the Iron and Steel Institute, vol.
174, pp. 25-28, 1953.
[104] R. Fleischer, "Substitutional solutes in AlRu-II. Hardening and correlations with defect structure.,"
Acta Metallurgica et Materialia, vol. 41, no. 4, pp. 1197-1205, 1993.
[105] D. Hull and D. J. Bacon, Introduction to dislocations, Oxford: Pergamon Press, 1984.
[106] Verein Deutscher Eisenhüttenleute VDEh, Steel, a handbook for materials research and
engineering. Volume 1: Fundamentals, Berlin: Springer-Verlag, 1992.
[107] R. Valiev, "Materials science: Nanomaterial advantage," Nature, vol. 419, no. 6910, pp. 887-889,
2002.
[108] Y. Wang and E. Ma, "Three strategies to achieve uniform tensile deformation in a nanostructured
metal," Acta Materialia, vol. 52, no. 6, pp. 1699-1709, 2004.
[109] M. Haddad, Y. Ivanisenko, E. Courtois-Manara and H.-J. Fecht, "In-situ tensile test of high strength
nanocrystalline bainitic steel," Materials Science and Engineering: A, vol. 620, pp. 30-35, 2014.
(125)
[110] C. Koch, D. Morris, K. Lu and A. Inoue, "Ductility of nanostructured materials," MRS Bulletin, vol.
24, no. 2, pp. 54-58, 1999.
[111] K. Chung, K. Ahn, D.-H. Yoo, K.-H. Chung, M.-H. Seo and S.-H. Park, "Formability of TWIP (twinning
induced plasticity) automotive sheets," International Journal of Plasticity, vol. 27, no. 1, pp. 52-
81, 2011.
[112] Y. Wang, M. Chen, F. Zhou and E. Ma, "High tensile ductility in a nanostructured metal," Nature,
vol. 419, no. 6910, p. 912–915, 2002.
[113] C. Youngdahl, J. Weertman, R. Hugo and H. Kung, "Deformation behavior in nanocrystalline
copper," Scripta Materialia, vol. 44, no. 8-9, pp. 1475-1478, 2001.
[114] M. J. Zehetbauer and Y. T. Zhu, Bulk Nanostructured Materials, Weinheim: WILEY-VCH Verlag
GmbH, 2009.
[115] R. Valiev, I. Alexandrov, Y. Zhu and T. Lowe, "Paradox of strength and ductility in metals
processed by severe plastic deformation," Journal of Materials Research, vol. 17, no. 1, pp. 5-8,
2002.
[116] R. Song, D. Ponge and D. Raabe, "Mechanical properties of an ultrafine grained C-Mn steel
processed by warm deformation and annealing," Acta Materialia, vol. 53, pp. 4881-4892, 2005.
[117] J. Ning, E. Courtois-Manara, L. Kurmanaeva, A. Ganeev, R. Valiev, C. Kübel and Y. Ivanisenko,
"Tensile properties and work hardening behaviors of ultrafine grained carbon steel and Armco
iron processed by warm high-pressure torsion," Materials Science and Engineering: A, vol. 581,
pp. 8-15, 2013.
[118] R. Valiev, A. Sergueeva and A. Mukherjee, "The effect of annealing on tensile deformation
behavior of nanostructured SPD titanium," Scripta Materialia, vol. 49, no. 7, pp. 669-674, 2003.
[119] M. Dao, L. Lu, Y. Shen and S. Suresh, "Strength, strain-rate sensitivity and ductility of copper with
nanoscale twins," Acta Materialia, vol. 54, no. 20, pp. 5421-5432, 2006.
[120] ASM Handbook, Properties and Selection: Irons, Steels, and High-Performance Alloys, Volume 1,
ASM International, 1993.
[121] Edelstahl witten-krefeld GMBH, THYROFORT heat-treatable steel manual, Witten, Germany.
[122] ASM Handbook, Heat treating, volume 4, USA: ASM International, 1991.
[123] L. Kurmanaeva, "Design and Mechanical Performance of Fully Dense Nanocrystalline Noble
Metals on the Basis of Pd, Pd-Ag and Pd-Au, Phd. dissertation," Ulm university, 2011.
[124] T. G. Langdon, "Twenty-five years of ultrafine-grained materials: Achieving exceptional properties
through grain refinement," Acta Materialia, vol. 61, no. 19, pp. 7035-7059, 2013.
(126)
[125] K. Edalati and Z. Horita, "A review on high-pressure torsion (HPT) from 1935 to 1988," Materials
Science and Engineering: A, vol. 652, pp. 325-352, 2016.
[126] M. Davidson and A. E. Vladar, "The physics of metrology instruments," Solid State Technology,
vol. 41, pp. 135-142, 1998.
[127] A. J. Wilkinson and T. B. Britton, "Strains, planes, and EBSD in materials science," Materials Today,
vol. 15, no. 9, pp. 366-376, 2012.
[128] T. Maitland and S. Sitzman, "Backscattering Detector and EBSD in Nanomaterials
Characterization," in Scanning Microscopy in Nanotechnology, New York, Springer, 2006, pp. 41-
75.
[129] R. F. Egerton, Physical principles of electron microscopy, USA: Springer, 2005.
[130] S. Reyntjens and R. Puers, "A review of focused ion beam applications in microsystem
technology," Journal of Micromechanics and Microengineering, vol. 11, no. 4, p. 287–300, 2001.
[131] E. F. Rauch, J. Portillo, S. Nicolopoulos, D. Bultreys, S. Rouvimov and P. Moeck, "Automated
nanocrystal orientation and phase mapping in the transmission electron microscope on the basis
of precession electron diffraction," Zeitschrift für Kristallographie, vol. 225, p. 103–109, 2010.
[132] B. Poon, D. Rittel and G. Ravichandran, "An analysis of nanoindentation in linearly elastic solids,"
International Journal of Solids and Structures, vol. 45, pp. 6018-6033, 2008.
[133] G. M. Pharr, W. C. Oliver and F. R. Brotzen, "On the generality of the relationship among contact
stiffness, contact area, and elastic modulus during indentation," Journal of Materials Research,
vol. 7, pp. 613-617, 1992.
[134] W. Oliver and G. Pharr, "An improved technique for determining hardness and elastic modulus
using load and displacement sensing indentation experiments," Journal of Materials Research,
vol. 7, pp. 1564-1583, 1992.
[135] H. Czichos, T. Saito and L. Smith, Handbook of Materials Measurement Methods, Germany:
Springer-Verlag, 2006.
[136] J. Ning, Y. Ivanisenko, D. Wang, M. Murashkin and H. Fecht, "Obtaining a Homogeneous Fe-C
Nanostructure from a Ferritic-Pearlitic Dual-Phase Steel by High Pressure Torsion," Materials
Science Forum, Vols. 667-669, pp. 199-204, 2011.
[137] E. Courtois, T. Epicier and C. Scott, "Characterisation of Niobium Carbide and Carbonitride
Evolution within Ferrite: Contribution of Transmission Electron Microscopy and Advanced
Associated Techniques," Materials Science Forum, vol. 500, pp. 669-676, 2005.
[138] ASM handbook, Fractography, Volume 12, 9th ed., 1992.
(127)
[139] X. Sauvage, A. Ganeev, Y. Ivanisenko, N. Enikeev, M. Murashkin and R. Valiev, "Grain boundary
segregation in UFG alloys processed by severe plastic deformation," Advanced Engineering
Materials, vol. 14, p. 968–974, 2012.
[140] J. Languillaume, G. Kapelski and B. Baudelet, "Cementite dissolution in heavily cold drawn
pearlitic steel wires," Acta Materialia, vol. 45, pp. 1201-1212, 1997.
[141] X. Sauvage, W. Lefebvre, C. Genevois, S. Ohsaki and K. Hono, "Complementary use of
transmission electron microscopy and atom probe tomography for the investigation of steels
nanostructured by severe plastic deformation," Scripta Materialia, vol. 60, pp. 1056-1061, 2009.
[142] J. Embury and R. Fisher, "The structure and properties of drawn pearlite," Acta Metallurgica,
Vols. 147-159, p. 14, 1966.
[143] G. Langford, "Deformation of pearlite," Metallurgical Transactions A, vol. 8A, pp. 861-875, 1977.
[144] X. Sauvage and Y. Ivanisenko, "The role of carbon segregation on nanocrystallisation of pearlitic
steels processed by severe plastic deformation," Journal of Material Science, vol. 42, pp. 1615-
1621, 2007.
[145] V. N. Gridnev, V. V. Nemoshkalenko, Y. Meshkov, V. G. Gavrilyuk, V. G. Prokopenko and O. N.
Razumov, "Moessbauer effect in deformed Fe-C alloys," Physica Status Solidi (A), vol. 31, pp. 201-
210, 1975.
[146] V. N. Gridnev, V. G. Gavrilyuk, I. Y. Dekhtyar, Y. Y. Meshkov, P. S. Nizin and V. G. Prokopenko,
"Investigation of carbide phase in strained steel by the method of nuclear gamma resonance,"
physica status solidi (a), vol. 14, no. 2, pp. 689-694, 1972.
[147] V. Gavriljuk, "Decomposition of cementite in pearlitic steel due to plastic deformation," Materials
Science and Engineering: A, vol. 345, no. 1-2, pp. 81-89, 2003.
[148] Y. Li, P. Choi, C. Borchers, S. Westerkamp, S. Goto, D. Raabe and R. Kirchheim, "Atomic-scale
mechanisms of deformation-induced cementite decomposition in pearlite," Acta Materialia, vol.
59, no. 10, pp. 3965-3977, 2011.
[149] R. Kirchheim, "Reducing grain boundary, dislocation line and vacancy formation energies by
solute segregation: II. Experimental evidence and consequences," Acta Materialia, vol. 55, no. 15,
pp. 5139-5148, 2007.
[150] Y. Ivanisenko, I. MacLaren, X. Sauvage, R. Valiev and H.-J. Fecht, "Shear-induced alpha to gamma
transformation in nanoscale Fe–C composite," Acta Materialia, vol. 54, pp. 1659-1669, 2006.
[151] W. Lojkowski, M. Djahanbakhsh, G. Buerkle, S. Gierlotka, W. Zielinski and H.-J. Fecht,
"Nanostructure formation on the surface of railway tracks," Materials Science and Engineering: A,
vol. 303, pp. 197-208, 2001.
(128)
[152] A. Korznikov, Y. Ivanisenko, D. Laptionok, I. Safarov, V. Pilyugin and R. Valiev, "Influence of severe
plastic deformation on structure and phase composition of carbon steel," Nanostructured
Materials, vol. 4, pp. 159-167, 1994.
[153] A. Goloborodko, O. Sitdikov, R. Kaibyshev, H. Miura and T. Sakai, "Effect of pressing temperature
on fine-grained structure formation in 7475 aluminum alloy during ECAP," Materials Science and
Engineering: A, vol. 381, no. 1-2, pp. 121-128, 2004.
[154] Y. Wang, P. Sun, P. Kao and C. Chang, "Effect of deformation temperature on the microstructure
developed in commercial purity aluminum processed by equal channel angular extrusion," Scripta
Materialia, vol. 50, no. 5, pp. 613-617, 2004.
[155] D. H. Shin, J.-J. Pak, Y. K. Kim, K.-T. Park and Y.-S. Kim, "Effect of pressing temperature on
microstructure and tensile behavior of low carbon steels processed by equal channel angular
pressing," Materials Science and Engineering: A, vol. 323, no. 1-2, pp. 409-415, 2002.
[156] R. J. Stokes and D. F. Evans, Fundamentals of interfacial engineering, New York: Wiley-VCH, 1997.
[157] T. Wang, Z. Li, B. Zhang, X. Zhang, J. Deng and F. Zhang, "High tensile ductility and high strength in
ultrafine-grained low-carbon steel," Materials Science and Engineering: A, vol. 527, no. 10-11, pp.
2798-2801, 2010.
[158] T. Lee, C. H. Park, D.-L. Lee and C. S. Lee, "Enhancing tensile properties of ultrafine-grained
medium-carbon steel utilizing fine carbides," Materials Science and Engineering: A, vol. 528, pp.
6558--6564, 2011.
[159] Y. Zhu and X. Liao, "Nanostructured materials retaining ductility," Nature Materials, vol. 3, pp.
351-352, 2004.
[160] Y. Zhao, Y. Zhu and E. J. Lavernia, "Strategies for Improving Tensile Ductility of Bulk
Nanostructured Materials," Advanced Engineering Materials, vol. 12, p. 769–778, 2010.
[161] R. KUZIAK, R. KAWALLA and S. WAENGLER, "Advanced high strength steels for automotive
industry," Archives of Civil and Mechanical Engineering, vol. 8, pp. 103-117, 2008.
[162] N. Zhong, X. Wang, L. Wang and Y. Rong, "Enhancement of the mechanical properties of a Nb-
microalloyed advanced high-strength steel treated by quenching–partitioning–tempering
process," Materials Science and Engineering: A, vol. 506, pp. 111--116, 2009.
[163] F. Wetscher, R. Stock and R. Pippan, "Changes in the mechanical properties of a pearlitic steel
due to large shear deformation," Materials Science and Engineering: A, Vols. 445-446, pp. 237--
243, 2007.
[164] A. Vorhauer, S. Kleber and R. Pippan, "Influence of processing temperature on microstructural
and mechanical properties of high-alloyed single-phase steels subjected to severe plastic
deformation," Materials Science and Engineering: A, Vols. 410-411, pp. 281-284, 2005.
(129)
[165] R. Armstrong, I. Codd, R. M. Douthwaite and N. J. Petch, "The plastic deformation of
polycrystalline aggregates," Philosophical Magazine, vol. 7, pp. 45-58, 1962.
[166] R. Valiev, N. Enikeev, M. Murashkin, V. Kazykhanov and X. Sauvage, "On the origin of the
extremely high strength of ultrafine-grained Al alloys produced by severe plastic deformation,"
Scripta Materialia, vol. 63, no. 9, pp. 949-952, 2010.
[167] M. Abramova, N. Enikeev, R. Valiev, A. Etienne, B. Radiguet, Y. Ivanisenko and X. Sauvage, "Grain
boundary segregation induced strengthening of an ultrafine-grained austenitic stainless steel,"
Materials Letters, vol. 136, pp. 349-352, 2014.
[168] C. Pande and K. Cooper, "Nanomechanics of Hall-Petch relationship in nanocrystalline materials,"
Progress in Materials Science, vol. 54, no. 6, pp. 689-706, 2009.
[169] R. Song, D. Ponge and D. Raabe, "Improvement of the work hardening rate of ultrafine grained
steels through second phase particles," Scripta Materialia, vol. 52, pp. 1075-1080, 2005.
[170] A. Nasser, A. Yadav, P. Pathak and T. Altan, "Determination of the flow stress of five AHSS sheet
materials (DP 600, DP 780, DP 780-CR, DP 780-HY and TRIP 780) using the uniaxial tensile and the
biaxial Viscous Pressure Bulge (VPB) tests," Journal of Materials Processing Technology, vol. 210,
pp. 429--436, 2010.
[171] B. Yan, Progress and challenges in forming AHSS, Ohio State University, Columbus, OH, 2009.
[172] T. Wierzbicki, Fracture of advanced high strength steel (AHSS), Massachusetts Institute of
Technology MIT, 2006.
[173] E. Bayraktar, F. A. Khalid and C. Levaillant, "Deformation and fracture behaviour of high
manganese austenitic steel," Journal of Materials Processing Technology, vol. 147, no. 2, pp. 145-
154, 2004.
[174] A. handbook, Failure analysis and prevention, Volume 11, 2002.
[175] G. Totten, K. Funatani and L. Xie, Handbook of metallurgical process design, Basel: Marcel Dekker,
Inc., 2004.
[176] O. Bouaziz, S. Allain, C. Scott, P. Cugy and D. Barbier, "High manganese austenitic twinning
induced plasticity steels: A review of the microstructure properties relationships," Current
Opinion in Solid State and Materials Science, vol. 15, no. 4, pp. 141-168, 2011.
[177] A. Hudgins, D. Matlock, J. Speer, J. Fekete and M. Walp, "The Susceptibility to Shear Fracture in
Bending of Advanced High Strength Sheet Steels," in Materials Science and Technology
Conference (MS&T), Detroit, Michigan, 2007.
[178] D. Broek, Elementary fracture mechanics, The Hague: Martinus Nijhoff Publishers, 1984.
(130)
[179] T. Pardoen, F. Hachez, B. Marchioni, P. Blyth and A. Atkins, "Mode I fracture of sheet metal,"
Journal of the Mechanics and Physics of Solids, vol. 52, no. 2, pp. 423-452, 2004.
[180] A. Needleman and V. Tvergaard, "Analyses of Plastic Flow Localization in Metals," Applied
Mechanics Reviews, vol. 45, no. 3S, pp. S3-S18, 1992.
[181] Y. Bao and T. Wierzbicki, "On fracture locus in the equivalent strain and stress triaxiality space,"
International Journal of Mechanical Sciences, vol. 46, no. 1, p. 81–98, 2004.
[182] Y. Bao and T. Wierzbicki, "On the cut-off value of negative triaxiality for fracture," Engineering
Fracture Mechanics, vol. 72, no. 7, p. 1049–1069, 2005.
[183] W. Garrison Jr and N. Moody, "Ductile fracture," Journal of Physics and Chemistry of Solids, vol.
48, no. 11, pp. 1035-1074, 1987.
[184] N. Fleck, J. Hutchinson and V. Tvergaard, "Softening by void nucleation and growth in tension and
shear," Journal of the Mechanics and Physics of Solids, vol. 37, no. 4, pp. 515-540, 1989.
[185] Y. Huang and J. Hutchinson, "A model study of the role of nonuniform defect distribution on
plastic shear localization," in The modelling in material behavior and its relation to design, 1990.
[186] J. Bandstra, D. Goto and D. Koss, "Ductile failure as a result of a void-sheet instability: experiment
and computational modeling," Materials Science and Engineering: A, vol. 249, no. 1-2, pp. 46-54,
1998.
[187] A. Needleman, "Continuum mechanics studies of plastic instabilities," Revue de Physique
Appliquée (Paris), vol. 23, no. 4, pp. 585 - 593, 1988.
[188] J. R. Rice, "The localization of plastic deformation," in 14th International Congress on Theoretical
and Applied Mechanics, Delft, 1976.
[189] J. Besson, "Continuum Models of Ductile Fracture: A Review," International Journal of Damage
Mechanics, vol. 19, no. 1, pp. 3-52, 2010.
[190] T. F. Morgeneyer, T. Taillandier-Thomas, L. Helfen, T. Baumbach, I. Sinclair, S. Roux and F. Hild, "In
situ 3-D observation of early strain localization during failure of thin Al alloy (2198) sheet," Acta
Materialia, vol. 69, pp. 78--91, 2014.
[191] Y. Mazaheri, N. Saeidi, A. Kermanpur and A. Najafizadeh, "Correlation of mechanical properties
with fracture surface features in a newly develope dual-phase steel," Journal of Materials
Engineering and Performance, vol. 24, no. 4, pp. 1573-1580, 2015.
[192] C. D. Calhoun and N. S. Stoloff, "The effects of particles on fracture processes in magnesium
alloys," Metallurgical Transactions, vol. 1, no. 4, pp. 997-1006, 1970.
[193] J. R. J. Low, R. H. Vanstone and R. H. Merchant, "An investigation of plastic fracture in aluminum
alloys," NASA-CR-131100, Technical report 2, Pittsburgh, PA, United States, 1972.
(131)
[194] T. B. Cox and J. R. J. Low, "Investigation of the plastic fracture of high strength steels," NASA-CR-
131102, Technical report 4, Pittsburgh, PA, United States, 1972.
[195] D. BroeK, A study on ductile fracture (PhD thesis), Delft: Delft University of Technology, 1971.
[196] A. Belyakov, Y. Kimura and K. Tsuzaki, "On Structural Mechanism of Continuous Recrystallization
in Ferritic Stainless Steel after Large Strain Processing," Materials Science Forum, Vols. 503 - 504,
pp. 323-328, 2006.
[197] A. Belyakov, T. Sakai, H. Miura, R. Kaibyshev and K. Tsuzaki, "Continuous recrystallization in
austenitic stainless steel after large strain deformation," Acta Materialia, vol. 50, pp. 1547-1557,
2002.
[198] Z. Horita and T. G. Langdon, "Achieving exceptional superplasticity in a bulk aluminum alloy
processed by high-pressure torsion," Scripta Materialia, vol. 58, no. 11, pp. 1029-1032, 2008.
[199] T. G. Langdon, "An evaluation of the strain contributed by grain boundary sliding in
superplasticity," Materials Science and Engineering: A, vol. 174, no. 2, pp. 225-230, 1994.
[200] M. Furukawa, Z. Horita and T. G. Langdon, "Factors influencing microstructural development in
equal-channel angular pressing," Metals and Materials International, vol. 9, no. 2, pp. 141-149,
2003.
[201] T. Langdon, "A unified approach to grain boundary sliding in creep and superplasticity," Acta
Metallurgica et Materialia, vol. 42, no. 7, pp. 2437-2443, 1994.
[202] T. G. Langdon, "The mechanical properties of superplastic materials," Metallurgical Transactions
A, vol. 13, no. 5, pp. 689-701, 1982.
[203] R. Mishra, T. Bieler and A. Mukherjee, "Superplasticity in powder metallurgy aluminum alloys and
composites," Acta Metallurgica et Materialia, vol. 43, no. 3, pp. 877-891, 1995.
[204] J. R. Rice and J.-S. Wang, "Embrittlement of interfaces by solute segregation," Materials Science
and Engineering: A, vol. 107, pp. 23-40, 1989.
[205] R. Z. Valiev, Y. Estrin, Z. Horita, T. G. Langdon, M. J. Zehetbauer and Y. Zhu, "Producing Bulk
Ultrafine-Grained Materials by Severe Plastic Deformation: Ten Years Later," JOM, vol. 68, no. 4,
pp. 1216-1226, 2016.
(132)
Publications
[1] M. Haddad, Y. Ivanisenko, E. Courtois-Manara, A. Kobler and H.-J. Fecht, "SEM in-situ tensile tests for
C45 steel processed by different warm high pressure torsion procedures.," To be submitted to Acta
materialia, 2017.
[2] A. Sommer, M. Haddad and H. Fecht, "Light Effect on Water Viscosity: Implication for ATP
Biosynthesis," Nature scientific reports, vol. 5, p. 12029, 2015.
[3] M. Haddad, Y. Ivanisenko and H.-J. Fecht, "Behavior of C45 steel Subjected to Different High Pressure
Torsion (HPT) Procedures," International Journal of Applied Science and Technology, vol. 5, no. 1, pp.
144-153, 2015.
[4] M. Haddad, Y. Ivanisenko, E. Courtois-Manara and H.-J. Fecht, "In-situ tensile test of high strength
nanocrystalline bainitic steel," Materials Science and Engineering: A, vol. 620, pp. 30-35, 2014.
[5] U. Herr, B. Riedmueller, M. Haddad, K. AbuShgair, M. Madel, J. Benjamin Jakob and K. Thonke,
"Investigation of Si nanowires from VLS growth," Physica status solidi (c), vol. 9, no. 11-12, p. 1912–
1915, 2012.
[6] A. P. Sommer, M. K. Haddad and H.-J. Fecht, "It is Time for a Change: Petri Dishes Weaken Cells,"
Journal of Bionic Engineering, vol. 9, no. 3, pp. 353-357, 2012.
(133)
Conference contributions
[1] A. P. Sommer, M. K. Haddad and H.-J. Fecht, "Tuning the Wheel of Life with Light (Invited lecture)," in
International Conferences on Laser Applications in Life Sciences, Ulm, 2014.
[2] M. Haddad, Y. Ivanisenko and H.-J. Fecht, "Deformation and fracture behavior of high strength
nanocrystalline medium carbon steel investigated by in-situ SEM tensile tests," in E-MRS Acta
Materialia Gold Medal Symposium, Warsow, 2012.
[3] Y. Ivanisenko, J. Ning, M. Haddad, E. Courtois-Manara, X. Sauvage, R. Valiev and H.-J. Fecht,
"Processing of high strength carbon steels by severe plastic deformation at elevated temperatures
(Invited lecture)," in XI International conference on nanostructured materials (Nano2012), Rhodes,
2012.
[4] M. Haddad, J. Ning, U. Hörmann, M. Murashkin, Y. Ivanisenko and H. Fecht, "Continuous
recrystallization and mechanical properties of a C45 steel after high pressure torsion," in Deutsche
Physikalische Gesellschaft conference, Dresden, 2011.
(134)
Acknowledgement
First of all, I would like to express my very great appreciation to my supervisor, Professor Hans-
Jörg Fecht for his great help, valuable guidance and endless support during planning,
development, discussion and reporting of this research, which developed and improved my
technical skills and scientific knowledge.
I would also like to express my deep gratitude to Dr. Julia Ivanisenko from Karlsruhe Institute of
Technology (KIT) for her patient guidance, enthusiastic encouragement and useful critiques and
advices during the past six years. She always gave me her time and knowledge very generously
in order to help me to finish my Ph.D.
My grateful thanks are extended to Professor Ulrich Herr and Professor Carl III Krill for their
constructive discussions and suggestions during my research work. I would also like to thank Mr.
Alexander Minkow, Mr. Günter Nusser and Mr. Jürgen Ankele for their technical assistance.
Furthermore, I wish to thank various friends and colleagues for their support during my work at
the Institute of Micro and Nanomaterials/Ulm University; Dr. Mattias Wiora, Dr. Kai Brühne, Mr.
Jules Dake, Mrs. Helga Faißt, Dr. Rainer Wunderlich, Mr. Markus Mohr, Dr. Andrei Sommer, Mr.
Benjamin Riedmüller, Mr. Michael Mertens and Mrs. Neda Wiora.
The generous financial support from the Deutsche Forschungsgemeinschaft (project FE 313/13-
1, IV98/2-1) is gratefully acknowledged.
Finally, I would like to thank my beloved parents, brother, and sister for their patience, support,
and love in spite of the distance.
(135)
List of Acronyms
3D-APT 3D atom-probe tomography
ACOM Automated phase and crystal orientation mapping
AFM Atomic force microscopy
AHSS Advanced high strength steel
ARB Accumulative roll bonding
BF TEM Bright field transmission electron microscopy
C2S2 Continuous confined strip shearing
CEN European Standardization Committee
DCAP Dissimilar-channel angular pressing
DF TEM Dark field transmission electron microscopy
DP steel Duel phase steel
EBSD Electron backscatter diffraction
EBSP Electron backscattered diffraction pattern
ECAP Equal channel angular pressing
FE-SEM Field emission scanning electron microscope
FIB Focused ion beam
GND Geometrically necessary dislocations
HPT High pressure torsion
HRTEM High resolution transmission electron microscopy
ISO International Organization for Standardization
IT diagram Isothermal transformation diagram
MEMS Micro-electro-mechanical systems
PED Pulse electrodeposition
RT HPT Room temperature high pressure torsion
SEM Scanning electron microscope
SPD Severe plastic deformation
STEM Scanning transmission electron microscope
TEM Transmission electron microscopy
TRIP steel Transformation induced plasticity steel
TTT diagram Time-temperature-transformation
TWIP steel Twinning induced plasticity steel
UFG materials Ultrafine-grained materials
WHPT Warm high pressure torsion
XRD X-ray diffraction
(136)
List of Figures
FIGURE 1: (A) A SCHEMATIC DRAWING ILLUSTRATING THE PROCESSING PROCEDURE OF THE HIGH PRESSURE TORSION
TECHNIQUE; (B) A SCHEMATIC DRAWING SHOWING THE DIFFERENT TYPES OF HIGH PRESSURE TORSION
PROCESSING. .................................................................................................................................... 13
FIGURE 2: AN ILLUSTRATION FOR THE HIGH PRESSURE TORSION SAMPLE SHOWING THE PARAMETERS USED TO ESTIMATE THE
IMPOSED STRAIN. REPRINTED FROM REFERENCE [28], WITH PERMISSION FROM ELSEVIER. ............................ 14
FIGURE 3: A DRAWING ILLUSTRATING THE ECAP PRINCIPLE AND SHOWING THE DIE’S ANGLES AND . ........................ 18
FIGURE 4: A GRAPH ILLUSTRATING THE ECAP FUNDAMENTAL ROUTES WITH THE ASSOCIATED SLIP SYSTEMS FOR THE
SUCCESSIVE PASSES. REPUBLISHED WITH PERMISSION OF ELSEVIER AND KOREAN INSTITUTE OF METALS AND
MATERIALS, FROM REFERENCES [48] AND [49] RESPECTIVELY; PERMISSIONS CONVEYED THROUGH COPYRIGHTS
CLEARANCE CENTER, INC. ................................................................................................................... 19
FIGURE 5: SCHEMATIC DRAWING EXPLAINING THE ACCUMULATIVE ROLL BONDING ARB PROCEDURE. REPRODUCED WITH
PERMISSION OF PERGAMON, FROM REFERENCE [26]; PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE
CENTER, INC. ................................................................................................................................... 23
FIGURE 6: SCHEMATIC DRAWING ILLUSTRATES THE TWIST EXTRUSION PROCESSING TECHNIQUE AND SHOWS THE SAMPLE
BEFORE, DURING, AND AFTER THE TWIST EXTRUSION. REPRODUCED WITH PERMISSION OF SPRINGER-VERLAG
FRANCE, FROM REFERENCE [70]. ......................................................................................................... 25
FIGURE 7: AN ILLUSTRATION OF THE GRAIN REFINEMENT DURING SEVERE PLASTIC DEFORMATION (SPD) SHOWING THE
DIFFERENT SEQUENCE PROCESSES; (A) EARLY DISLOCATIONS DEVELOPMENT, (B) ELONGATED STRUCTURE
FORMATION, (C) SUBGRAIN BOUNDARIES FORMATION, (D) BLOCKING DISLOCATION PROPAGATION BY SUBGRAIN
BOUNDARIES, (E) FINE STRUCTURE FORMATION. THIS FIGURE IS REPRODUCED WITH PERMISSION OF ELSEVIER S.A,
FROM REFERENCE [78]. PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. ................ 27
FIGURE 8: SEM IMAGES SHOWING THAT THE MICROSTRUCTURE CONSISTS OF PEARLITE AND FERRITE; (A) FOR AS DELIVERED
ST45 STEEL; (B) FOR AS DELIVERED C45 STEEL. ...................................................................................... 38
FIGURE 9: (A) THE TIME-TEMPERATURE-TRANSFORMATION (TTT) DIAGRAM FOR THE C45 STEEL; (B) SEM MICROGRAPH
SHOWING THE BAINITIC MICROSTRUCTURE OF THE C45 STEEL SAMPLE AFTER THE PATENTING TREATMENT. ...... 39
FIGURE 10: (A) STANDARD COMPRESSION-TORSION TESTING MACHINE SCHENCK WITH A CAPACITY OF 400 KN; (B) SPECIALLY
DESIGNED BRIDGMEN TYPE ANVILS EQUIPPED WITH THE MACHINE. THIS FIGURE IS REPUBLISHED FROM REFERENCE
[123]. ............................................................................................................................................ 40
FIGURE 11: (A) SCHEMATIC DRAWING FOR THE EBSD SYSTEM; (B) EBSD SAMPLE EXTRACTION THROUGH FIB. .............. 42
FIGURE 12: (A) A MICROGRAPH USING THE ELECTRON BEAM SHOWING THE TEM SAMPLE EXTRACTION USING THE FIB; (B)
SEM MICROGRAPH SHOWING THE LOCATION OF THE TEM SPECIMEN FROM THE TENSILE SAMPLE AFTER IN-SITU
(137)
TENSILE TEST. FIGURE 12(B) IS REPUBLISHED WITH PERMISSION OF ELSEVIER S.A, FROM REFERENCE [109];
PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. ................................................. 44
FIGURE 13: (A) SCANNING THE AREA OF THE SPECIMEN WITH 1 NM SPOTS WITH A SCAN STEP SIZE OF TENS OF NM; (B)
ELECTRON DIFFRACTION SPOT PATTERNS; (C) ACOM SETUP FOR FEI TECNAI F20 ST ELECTRON MICROSCOPE.
REPUBLISHED WITH PERMISSION OF OLDENBOURG WISSENSCHAFTSVERLAG GMBH, FROM REFERENCE [131];
PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. ................................................. 45
FIGURE 14: (A) A DRAWING SHOWS THE SHAPE OF THE SQUARE-BASED DIAMOND PYRAMID USED IN VICKERS HARDNESS TEST;
(B) A DRAWING SHOWS THE SHAPE OF THE RESULTING INDENT AND D1 AND D2 ARE THE DIAGONALS. ............. 46
FIGURE 15: (A) SEM IMAGE SHOWING A BERKOVICH TIP (FROM UNIVERSITY OF NEBRASKA-LINCOLN WEBSITE, ACCESSED 3RD
APRIL 2017, HTTP://BM3.UNL.EDU/GRAPHICS/NANOINDENTER_BERKOVICH_TIP_01.JPG); (B) A SCHEMATIC
REPRESENTATION OF LOAD VERSUS INDENTER DISPLACEMENT INTO THE SURFACE OF THE SAMPLE FOR
NANOINDENTATION EXPERIMENT SHOWING BOTH CURVES OF LOADING AND UNLOADING WHILE S IS THE INITIAL
UNLOADING STIFFNESS. ...................................................................................................................... 48
FIGURE 16: (A) SCHEMATIC DRAWING FOR A STANDARD TENSILE TEST SETUP; (B) SCHEMATIC REPRESENTATION SHOWING
CUTTING POSITION OF TENSILE SPECIMENS FROM HPT DISKS. ................................................................... 49
FIGURE 17: (A) A DIAGRAM SHOWING THE MICROHARDNESS OF THE CYCLIC HPT PROCESSED ST45 STEEL SAMPLES THROUGH
THEIR DIAMETER AFTER 1, 4, AND 6 CYCLES; (B) A DIAGRAM SHOWING THE MICROHARDNESS OF THE CYCLIC HPT
PROCESSED C45 STEEL SAMPLES THROUGH THEIR DIAMETER AFTER 4, AND 6 CYCLES; (C) BAR CHART COMPARING
THE MICROHARDNESS OF BOTH C45 AND ST45 STEELS BEFORE AND AFTER THE HPT PROCESSING FOR 4 AND 6
CYCLES UNDER 4.5 GPA OF PRESSURE; (D) BAR CHART SHOWING THE HARDNESS OF ST45 SAMPLES AFTER CYCLIC
HPT PROCESSING AND CONTINUOUS ROTATIONS HPT PROCESSING; (E) BAR CHART SHOWING THE
MICROHARDNESS OF C45 SAMPLES PROCESSED BY CONTINUOUS ROTATIONS HPT UNDER 6 GPA OF PRESSURE FOR
0, 3, AND 5 ROTATIONS. .................................................................................................................... 57
FIGURE 18: (A) AND (B) SEM IMAGES SHOWING THE SURFACE STRUCTURE OF ST45 STEEL AFTER 4 AND 6 CYCLES OF HPT
PROCESSING AT 4.5 GPA OF PRESSURE; (C) AND (D) SEM IMAGES SHOWING THE SURFACE STRUCTURE OF C45
STEEL AFTER 4 AND 6 CYCLES OF HPT PROCESSING AT 4.5 GPA OF PRESSURE; (E) SEM IMAGE SHOWING ETCHING
RESISTANT AREA (SURROUNDED BY THE ELLIPSE) ON THE SURFACE OF THE C45 SAMPLE PROCESSED BY HPT FOR 6
CYCLES UNDER 4.5 GPA OF PRESSURE; (F) SEM IMAGE SHOWING THE SURFACE STRUCTURE OF THE C45 SAMPLE
PROCESSED BY HPT FOR 5 ROTATIONS UNDER 4.5 GPA OF PRESSURE; (G) SEM IMAGE SHOWING THE SURFACE
STRUCTURE OF THE C45 SAMPLE PROCESSED BY HPT FOR 5 ROTATIONS UNDER 6 GPA OF PRESSURE.............. 58
FIGURE 19: MICROHARDNESS OF RT HPT PROCESSED SAMPLES MEASURED ALONG THE HPT SAMPLE DIAMETER. OBTAINED
FROM DR. JIANG-LI NING, KIT. ........................................................................................................... 59
FIGURE 20: SEM OBSERVATIONS OF THE CEMENTITE MORPHOLOGY WITH HPT DEFORMATION; (A) 0 (HPT SAMPLE
CENTER); (B) 190; (C) 310. ..................................................................................................... 59
FIGURE 21: TEM IMAGES SHOWING THE STRUCTURE OF FERRITE PHASE AFTER HPT PROCESSING TO SHEAR STRAIN 190:
(A) THE CELLULAR STRUCTURE WITH SAED PATTERN; (B) THE BANDED NANO-GRANULAR STRUCTURE WITH SAED
PATTERN; (C) THE CORRESPONDING DARK FIELD IMAGE IN THE SAME AREA WITH (B) CREATED USING A PART OF
110Α AND 200Α RINGS. REPUBLISHED WITH PERMISSION OF TRANS TECH PUBLICATIONS, FROM REFERENCE [136];
PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. ................................................. 60
(138)
FIGURE 22: TEM IMAGES IN AN AREA WITH BANDED-STRUCTURE AFTER HPT PROCESSING TO SHEAR STRAIN 190; (A) THE
BRIGHT FIELD IMAGE; (B) THE DARK FIELD IMAGE OF THE SAME AREA CREATED USING SOME OF THE CEMENTITE
REFLECTIONS WITHIN THE 110Α RING. REPUBLISHED WITH PERMISSION OF TRANS TECH PUBLICATIONS, FROM
REFERENCE [136]; PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. ....................... 61
FIGURE 23: TEM IMAGES OF THE C45 STEEL AFTER HPT PROCESSING TO SHEAR STRAIN 310; (A) THE BRIGHT FIELD TEM
IMAGE AND THE SAED PATTERN; (B) THE DARK TEM FIELD IMAGE OF FERRITE IN THE SAME AREA IN (A) CREATED
USING A PART OF THE REFLECTIONS OF 110Α AND 200Α RINGS; (C) THE DARK FIELD TEM IMAGE OF CEMENTITE IN
THE SAME AREA WITH (A) CREATED USING SOME OF THE CEMENTITE REFLECTIONS WITHIN THE 110Α RING.
REPUBLISHED WITH PERMISSION OF TRANS TECH PUBLICATIONS, FROM REFERENCE [136]; PERMISSION CONVEYED
THROUGH COPYRIGHTS CLEARANCE CENTER, INC. .................................................................................. 62
FIGURE 24: SIZE DISTRIBUTION OF THE REMAINING CEMENTITE PARTICLES IN THE SAMPLES, (A) 190; (B) 310.
REPUBLISHED WITH PERMISSION OF TRANS TECH PUBLICATIONS, FROM REFERENCE [136]; PERMISSION CONVEYED
THROUGH COPYRIGHTS CLEARANCE CENTER, INC. .................................................................................. 63
FIGURE 25: TENSILE SPECIMENS: (A,B) SCHEMATIC REPRESENTATION SHOWING CUTTING POSITION OF TENSILE SPECIMENS
FROM HPT DISKS; (C) VIEW OF THE SPECIMENS BEFORE THE TENSILE TEST, SPECIAL COATING WITH TIO2 IS APPLIED
ON THE UPPER SURFACE OF THE SPECIMEN. FIGURE 25(A) IS REPUBLISHED WITH PERMISSION OF ELSEVIER S.A,
FROM REFERENCE [109]; PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. .............. 63
FIGURE 26: CRACKS RESULTED FROM RT HPT IN TENSILE SPECIMENS OF C45 STEEL. .................................................. 64
FIGURE 27: STRESS/STRAIN DIAGRAMS OF RT HPT PROCESSED C45 STEEL SAMPLES. ................................................. 64
FIGURE 28: MICROSTRUCTURE OF C45 STEEL AFTER RT HPT AND SUBSEQUENT ANNEALING FOR TWO HOURS; (A, B) AT
400°C, BRIGHT FIELD TEM IMAGES, (C) AT 400°C, BRIGHT FIELD (LEFT) AND DARK FIELD (RIGHT) TEM IMAGES
SHOWING ARRANGEMENTS OF DISLOCATIONS, (D) HR TEM IMAGE OF C45 STEEL AFTER RT HPT AND SUBSEQUENT
ANNEALING FOR TWO HOURS AT 400°C, (E, F) BRIGHT FIELD TEM IMAGES AFTER ANNEALING AT 450°C FOR 2
HOURS. ........................................................................................................................................... 65
FIGURE 29: STRESS/STRAIN DIAGRAMS OF RT HPT PROCESSED AND ANNEALED C45 STEEL SAMPLES. ........................... 67
FIGURE 30: (A, B, C, D) THE MICROSTRUCTURE OF C45 STEEL AFTER WHPT PROCESSING AT 350°C FOR 5 ROTATIONS UNDER
6 GPA OF PRESSURE; (A) BF TEM IMAGE TAKEN IN TANGENTIAL DIRECTION, AN ARROW INDICATES SHEAR
DIRECTION AT WHPT; (B) BF TEM IMAGE SHOWING TRACES OF CEMENTITE LAMELLAS ALONG THE INTERFACES;
(C) ACOM 2-PHASES IMAGE, RED COLOR INDICATES FERRITE WHILE GREEN COLOR INDICATES CEMENTITE; (D)
ORIENTATION MAPS CREATED IN RADIAL AND TANGENTIAL DIRECTIONS; (E) GRAIN SIZE DISTRIBUTION OBTAINED
FROM A POPULATION OF 260 GRAINS. REPUBLISHED WITH PERMISSION OF ELSEVIER S.A, FROM REFERENCE [109];
PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. ................................................. 68
FIGURE 31: THE ENGINEERING AND THE TRUE STRESS-STRAIN CURVE FOR THE WHPT PROCESSED SAMPLE AT 350°C.
REPUBLISHED WITH PERMISSION OF ELSEVIER S.A, FROM REFERENCE [109]; PERMISSION CONVEYED THROUGH
COPYRIGHTS CLEARANCE CENTER, INC. ................................................................................................. 69
FIGURE 32: SEM IMAGES TAKEN DURING THE IN-SITU TENSILE TEST OF WHPT PROCESSED C45 STEEL; (A) TENSILE SPECIMEN
IMMEDIATELY BEFORE FRACTURE, NO NECKING IS RECOGNIZED; (B) SURFACE FLAWS ONE OF WHICH (SHOWN WITH
LARGE WHITE ARROW) GAVE AN ORIGIN OF THE CRACK; (C) SHEAR BANDS NEAR SURFACE DEFECTS; (D) ENLARGED
SEM IMAGE OF THE SURFACE FLAW TAKEN IMMEDIATELY BEFORE FRACTURE SHOWING DEVELOPMENT OF TWO
(139)
CRACKS SPREADING AT AN ANGLE OF 35° TO THE SAMPLE EDGE (SHOWN WITH ARROWS); (E) THE SAMPLE AFTER
FAILURE SHOWING A SLANT THROUGH THE THICKNESS FRACTURE SURFACE. REPUBLISHED WITH PERMISSION OF
ELSEVIER S.A, FROM REFERENCE [109]; PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC.
...................................................................................................................................................... 70
FIGURE 33: SEM IMAGES OF WHPT-PROCESSED C45 SAMPLE TAKEN AT DIFFERENT MAGNIFICATIONS SHOWING DIFFERENT
FEATURES OF THE FRACTURE SURFACE; (A) OVERVIEW IMAGE OF DIMPLES; (B) ENLARGED VIEW OF AN AREA IN (A);
(C) LARGE OVAL SHAPED DIMPLE WITH AN ARRAY OF INCLUSIONS ON THE BOTTOM; (D) DEFORMATION MARKS ON
THE WALL OF A LARGE DIMPLE RESULTING FROM SHEAR BANDS. REPUBLISHED WITH PERMISSION OF ELSEVIER S.A,
FROM REFERENCE [109]; PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. .............. 71
FIGURE 34: (A) BF TEM IMAGE SHOWING THAT THE MICROSTRUCTURE OF THE C45 STEEL AFTER THE IN-SITU TENSILE TEST
HAD NOT CHANGED, AN ARROW INDICATES SHEAR DIRECTION AT WHPT AND TENSILE DIRECTION; (B) STEM IMAGE
OF DISLOCATIONS IN THE FERRITE AFTER TENSILE TEST. REPUBLISHED WITH PERMISSION OF ELSEVIER S.A, FROM
REFERENCE [109]; PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. ....................... 74
FIGURE 35: (A) IMAGE SHOWING THE IN-SITU TENSILE TEST SETUP. (B) IMAGE SHOWING THE DUMMY SAMPLE WITH THE
THERMOCOUPLE ATTACHED TO ITS SURFACE. ......................................................................................... 75
FIGURE 36: (A) STRESS-STRAIN DIAGRAM OBTAINED THROUGH IN-SITU TENSILE TEST; (B) SEM IMAGE AT A STRAIN LEVEL OF
25% SHOWING NECKING AS WELL AS SHEAR BANDS; (C) SEM IMAGE AT A STRAIN LEVEL OF 27% AND THE ARROWS
IS INDICATING TWO MICRO-CRACKS; (D) SEM IMAGE SHOWING THE PROPAGATION OF THE CRACK DURING THE IN
SITU TENSILE TEST EXPERIMENT............................................................................................................ 75
FIGURE 37: (A) STRESS-STRAIN CURVE OBTAINED FROM THE IN-SITU TENSILE EXPERIMENT AT 370°C USING A STRAIN RATE OF
0.02 µM; (B) STRESS-STRAIN CURVE OBTAINED FROM THE IN-SITU TENSILE EXPERIMENT AT 440°C USING A STRAIN
RATE OF 0.02 µM; (C) SEM IMAGE OBTAINED FROM THE 370°C TENSILE EXPERIMENT SHOWING NECKING AND
SHEAR BANDS; (D) SEM IMAGE OBTAINED FROM THE 440°C TENSILE EXPERIMENT SHOWING NECKING AND SHEAR
BANDS; (E) SEM IMAGE SHOWING THE CRACK PROPAGATION DURING TENSILE TESTING THE SAMPLE AT 370°C; (F)
SEM IMAGE SHOWING THE FINAL FRACTURE AFTER OPERATING THE TENSILE TEST AT 440°C. ........................ 76
FIGURE 38: (A, B, C) MICROSTRUCTURE FOR THE WHPT SAMPLE PROCESSED AT 380°C FOR 10 ROTATIONS; (A) BF TEM
IMAGE; (B) ACOM ORIENTATION MAP CREATED IN A TANGENTIAL DIRECTION TO THE SHEAR PLANE; (C)
CORRESPONDING ACOM 2-PHASES IMAGE, RED COLOR REPRESENTS FERRITE WHILE GREEN COLOR REPRESENTS
CEMENTITE; D) GRAIN SIZE DISTRIBUTION OBTAINED THROUGH EBSD FROM A POPULATION OF 505 GRAINS. .. 77
FIGURE 39: THE ENGINEERING AND THE TRUE STRESS-STRAIN CURVE FOR THE WHPT PROCESSED SAMPLE AT 380°C. ..... 79
FIGURE 40: SEM MICROGRAPHS SHOWING; (A) TENSILE SPECIMEN IMMEDIATELY BEFORE FRACTURE, NO NECKING IS
RECOGNIZED; (B) SURFACE FLAWS ONE OF WHICH GAVE AN ORIGIN OF THE CRACK; (C) SURFACE DEFECT WITH NO
SHEAR BANDS RECOGNIZED; (D) THE SAMPLE AFTER FAILURE SHOWING A SLANT THROUGH THE THICKNESS
FRACTURE SURFACE. .......................................................................................................................... 81
FIGURE 41: (A) SEM MICROGRAPH SHOWING THAT THE FRACTURE SURFACE IS COVERED WITH VERY FINE DIMPLES; (B)
HIGHER MAGNIFICATION SEM MICROGRAPHS OF THE FRACTURE SURFACE SHOWING THAT THE DIMPLES HAVE
EQUIAXED SHAPE............................................................................................................................... 82
FIGURE 42: (A) APT 3D RECONSTRUCTED VOLUME SHOWING THE HETEROGENEOUS DISTRIBUTION OF CARBON ATOMS (RED
COLOR). (B) CARBON CONCENTRATION PROFILE ACROSS A CARBIDE/MATRIX INTERFACE (DIRECTION LABELED A ON
(140)
IMAGE (A), SAMPLING VOLUME THICKNESS 2 NM). (C) CARBON CONCENTRATION PROFILE ACROSS A SEGREGATION
ALONG A BOUNDARY (DIRECTION LABELED B ON IMAGE (A), SAMPLING VOLUME THICKNESS 2 NM). ............... 91
FIGURE 43: CURVES FOR THE NATURAL LOGARITHM OF THE TRUE STRESS VS. THE NATURAL LOGARITHM OF THE TRUE STRAIN
FOR: A) THE 350°C WHPT PROCESSED SAMPLE. B) THE 380°C WHPT PROCESSED SAMPLE. ....................... 98
FIGURE 44: FROM REFERENCE [23]; (A) IMAGE SHOWING THE 3-D VON MISES EQUIVALENT STRAIN FIELDS AND FINAL CRACK
PATTERN IN THE MATERIAL AFTER LOAD STEP 4, (B) LAMINOGRAPHY IMAGE SHOWING THE 3-D RENDERING OF
DAMAGE/VOID EVOLUTION AND THE SLANT FRACTURE. REPUBLISHED WITH PERMISSION OF ELSEVIER S.A, FROM
REFERENCE [190]; PERMISSION CONVEYED THROUGH COPYRIGHTS CLEARANCE CENTER, INC. ..................... 101
FIGURE 45: FLOWCHART SUMMARIZES THE MICROSTRUCTURE EVOLUTION DURING RT HPT PROCESS. ........................ 111
FIGURE 46: FLOWCHART SUMMARIZES THE MICROSTRUCTURE EVOLUTION DURING WHPT PROCESS. ......................... 112
(141)
Curriculum vitae
The curriculum vitae was removed due to reasons of data protection.
(142)
Parts of my dissertation are already published in the following journal articles:
[1] M. Haddad, Y. Ivanisenko and H.-J. Fecht, "Behavior of C45 steel Subjected to Different High Pressure Torsion (HPT) Procedures," International Journal of Applied Science and Technology, vol. 5, no. 1, pp. 144-153, 2015.
[2] M. Haddad, Y. Ivanisenko, E. Courtois-Manara and H.-J. Fecht, "In-situ tensile test of high strength nanocrystalline bainitic steel," Materials Science and Engineering: A, vol. 620, pp. 30-35, 2014.