2020 Ductile Mode Cutting of Brittle Materials
2020 Ductile Mode Cutting of Brittle Materials
2020 Ductile Mode Cutting of Brittle Materials
Kui Liu
Hao Wang
Xinquan Zhang
Ductile Mode
Cutting
of Brittle
Materials
Springer Series in Advanced Manufacturing
Series Editor
Duc Truong Pham, University of Birmingham, Birmingham, UK
The Springer Series in Advanced Manufacturing includes advanced textbooks,
research monographs, edited works and conference proceedings covering all major
subjects in the field of advanced manufacturing.
The following is a non-exclusive list of subjects relevant to the series:
1. Manufacturing processes and operations (material processing; assembly; test
and inspection; packaging and shipping).
2. Manufacturing product and process design (product design; product data
management; product development; manufacturing system planning).
3. Enterprise management (product life cycle management; production planning
and control; quality management).
Emphasis will be placed on novel material of topical interest (for example, books
on nanomanufacturing) as well as new treatments of more traditional areas.
As advanced manufacturing usually involves extensive use of information and
communication technology (ICT), books dealing with advanced ICT tools for
advanced manufacturing are also of interest to the Series.
Springer and Professor Pham welcome book ideas from authors. Potential
authors who wish to submit a book proposal should contact Anthony Doyle,
Executive Editor, Springer, e-mail: anthony.doyle@springer.com.
123
Kui Liu Hao Wang
Singapore Institute of Manufacturing Department of Mechanical Engineering
Technology National University of Singapore
Singapore, Singapore Singapore, Singapore
Xinquan Zhang
School of Mechanical Engineering
Shanghai Jiao Tong University
Shanghai, China
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface
Brittle material, in general, shows little ability to deform plastically and commonly
fracture at or very near to the elastic limit. Usually, it is considerably stronger in
compression than in tension. Nowadays, brittle material has been applied in numerous
important industries including aerospace, oil and gas, precision engineering, optics,
instruments, automotive, semiconductor, marine and micro-electromechanical sys-
tems. Meanwhile, there are rapid growing demands of its engineering applications
due to unique and non-replicable material properties. It has attracted great interests
from both engineers and academics for the sake of its excellent mechanical, electrical,
optical, physical and chemical properties. However, it is very difficult to machine
brittle material using conventional cutting technologies to obtain very smooth and
damage-free surfaces due to its high hardness, high wear resistance and high
toughness. In general, abrasive processes such as grinding, lapping and polishing are
commonly used for fabricating the brittle material components. The demerits asso-
ciated with those processes are poor machinability, subsurface damage, high manu-
facturing cost and time-consuming. Such that engineering applications of brittle
material are largely limited. Naturally, questions on how to overcome this problem are
surfacing up. Thereafter, a technology for efficiently cutting of brittle material is
urgently needed for the industry.
Ductile mode cutting of brittle material is a very promising and well-recognized
technology enabling to achieve high-quality and crack-free surface for the industry.
But what is ductile mode cutting and how to achieve ductile mode cutting of brittle
material? To answer those questions, this book is dedicated to an in-depth study and
understanding of material removal behaviour in cutting of brittle material, where
stock material is removed by plastic deformation rather than fracturing. Ductile
mode cutting of brittle material can be achieved under certain cutting conditions,
while crack-free and no subsurface-damage surfaces can be obtained simultane-
ously. The book intends to provide a comprehensive understanding to the research
community, including ductile mode cutting fundamentals such as mechanism,
characteristics, modelling and molecular dynamics simulation, ductile mode cutting
applications such as silicon, glass, tungsten carbide and calcium fluoride, as well as
hybrid ductile mode cutting like ultrasonic vibration and thermally assisted ductile
v
vi Preface
mode cutting of brittle material. The book details ductile mode cutting of brittle
material systematically in terms of fundamentals, engineering applications and
hybrid ductile mode cutting techniques, which is structured and organized as:
Chapter 1: Literature review and state of the art in terms of ductile nature and
plasticity of brittle material, ductile-to-brittle transition phenomena and mechanism,
dislocation dynamics, crack propagation behaviour, ductile regime grinding and
ductile mode cutting.
Chapter 2: Theoretical analysing of the mechanism of ductile mode cutting of brittle
material, such as the coexisting crack propagation and dislocation extension in chip
formation zone, large compressive stress and shear stress in cutting zone, fracture
mechanics, yield strength enhancement by dislocation hardening and strain gradi-
ent, and ductile mode cutting conditions as well.
Chapter 3: Experimental investigation on ductile mode cutting characteristics
including grooving surface morphology, material removal mechanism and material
removal mode, and tool wear mechanism.
Chapter 4: Mathematical modelling of ductile mode cutting of brittle material to
predict critical undeformed chip thickness based on work material’s properties, or
cutting tool geometry and cutting conditions used, as well experimental verification
in terms of grooved surface topography, chip morphology, critical depth of cut and
material removal ratio.
Chapter 5: Fundamentals and review of molecular dynamics including interatomic
potentials adopted for brittle material modelling, with examples of molecular
dynamics simulations performed on silicon and silicon carbide, as well as theo-
retical techniques to determine stress distributions during cutting of brittle material.
Chapter 6: Experimental studies on cutting characteristics of single crystal silicon
wafers including tool edge radius effect on critical undeformed chip thickness for
ductile mode cutting, upper bound for diamond tool edge radius achieving ductile
mode cutting of silicon, and ductile mode cutting performance.
Chapter 7: Experimental studies of glass cutting characteristics through grooving,
groove turning and ultrasonic vibration-assisted cutting of soda-lime glass in terms
of nanometric cutting mode, machined surface topography, surface roughness, chip
formation and tool wear.
Chapter 8: Material characteristics analysing of tungsten carbide and experimental
studies on cutting performance of tungsten carbide under normal and high cutting
speed at nanometric scale chip formation in terms of cutting force, machined sur-
face topography, surface roughness, chip formation and tool wear.
Chapter 9: Understanding of optical surface generation on calcium fluoride single
crystals, including cutting condition assessment and material constraints, surface
evaluation techniques and anisotropic machined surface morphology, theoretical
simulations, and advanced machining techniques.
Chapter 10: Analytical modelling of critical undeformed chip thickness prediction
in ultrasonic vibration-assisted cutting based on the variation of specific cutting
energy in both ductile mode cutting and brittle mode cutting of brittle material, as
well as experimental verification.
Preface vii
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Brittle Cutting Versus Ductile Cutting . . . . . . . . . . . . . . . . . . . 2
1.3 Ductile Nature and Plasticity of Brittle Material . . . . . . . . . . . . 3
1.4 Ductile-to-Brittle Transition . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.5 Ductile-to-Brittle Transition Mechanisms . . . . . . . . . . . . . . . . . 7
1.5.1 Dislocation Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.5.2 Dislocation Crack Behaviours . . . . . . . . . . . . . . . . . . . 8
1.6 Ductile Machining of Brittle Material . . . . . . . . . . . . . . . . . . . 9
1.6.1 Ductile Regime Grinding . . . . . . . . . . . . . . . . . . . . . . 10
1.6.2 Ductile Mode Cutting . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.7 Concluding Marks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
ix
x Contents
xv
xvi Nomenclature
f Dislocation fraction
Function of strain
Feed
Feed rate
Frequency
fab Work material removal ratio
fA Attractive pair potential
fc Cut-off function
fR Repulsive pair potential
f Wa Function of crack size
h Thickness of material to be removed
hab Strain hardening factor
h0 Self-hardening factor
k Material thermal conductivity
Interatomic spring constant
k 1, k 2 Scaling constant
kAB Shear flow stress along shear plane AB
kB Boltzmann constant
l Characteristic material length
ls Shear plane length
m Parameter of crystal material performance
Atom mass
Rate sensitivity factor
n Crack’s number within a finite plate
Spindle rotation speed
Refractive index
n Direction normal to the slip plane
nc Coefficient of heat conducting into tool
p Pressure during machining
pi Momentum of atom i
q Power density of the heat source
qi Atomic charge of atom i
r Tool cutting edge radius
rB Average radius of the Gaussian beam
rc Ratio of the undeformed chip thickness ac to chip thickness ao
re Effective distance between atoms i and j
ri Position of atom i
r0 Material-dependent constant in the Born–Mayer interatomic potential
Bond length in the Tersoff interatomic potential
Workpiece outer radius
s Slip direction
Elastic recovery
t Time
to Undeformed chip thickness
Nomenclature xix
xxiii
xxiv Acronyms
1.1 Background
Brittle material such as glass, silicon, tungsten carbide (WC), germanium and silicon
nitride have been widely employed in the industry including precision engineering,
optics, instruments, semiconductor and micro-electromechanical systems (MEMS)
due to its excellent mechanical, electrical, optical, physical and chemical properties.
Also, there are rapidly growing demands on manufacturing of brittle material achiev-
ing a good quality surface finish, stringent geometry accuracy and surface integrity
with less or free of subsurface damage. Meanwhile, to reduce the manufacturing
cost in the production of these components and devices made by brittle materials,
efficient machining of these brittle materials is very much demanded. Traditionally,
abrasive processes such as grinding, lapping and polishing have been widely used
for final surface finishing of brittle material. The demerits associated with those pro-
cesses include poor grindability, high manufacturing cost and subsurface damage
[1]. Furthermore, the abrasive processes will cause surface flatness deviation due
to its uncontrollable material removal resulting in the machined profile inaccuracy
[2]. Therefore, after grinding and lapping processes, chemical-mechanical polish-
ing (CMP) is essential to remove the subsurface damage layer caused by the hard
abrasive particles, which makes a very costly production [3]. Also, those abrasive
processes especially like CMP are extremely slow, meanwhile grinding and lapping
processes would impart subsurface damage led to a degraded surface integrity [4].
When cutting of brittle material using conventional cutting techniques like turning,
milling and drilling, usually the chip formation is a fracture process that will derive
to the machined surface being damaged and leads to an inacceptable surface quality.
In order to improve the machined surface integrity of brittle material, ductile mode
cutting (DMC), also called ductile regime cutting (DRC) or ductile cutting (DC), as
a promising technique, has been studied vigorously over the past decades. Common
understanding that ductile mode cutting is a material removal process where work
material is removed by plastic flow instead of brittle fracture deriving a damage-free
surface. Ductile mode cutting of brittle material can be achieved by having right
cutting conditions and tool geometry, where both machined surface finish and form
accuracy are better ensured. There is a ductile-to-brittle transition (DBT) in chip
formation when cutting of brittle material with a greatly reduced undeformed chip
thickness [5–13]. As a result, the subsequent polishing process could be no longer
necessary or the polishing time could be largely reduced because the crack-free sur-
faces can be directly produced by ductile mode cutting without subsurface damage or
the subsurface damage layer thickness being much smaller, which would significantly
reduce the manufacturing time and cost for brittle material. This advantage cannot
be under addressed because in machining even a minor improvement in productivity
would lead to a major impact in mass production.
A schematic comparison between ductile mode cutting and brittle mode cutting
(BMC) of brittle material helps to reveal the underlying mechanisms as shown
in Fig. 1.1: (a) ductile mode of material removal by eliminating a surface plas-
tic layer formed as a result of a high contact pressure in cutting zone; (b) brittle
fracture of work material leaving surface and subsurface damages, of which the
damage layer is as deep as 5–10 µm due to crack propagations in machining of
silicon [14]. The fundamental premise of ductile mode cutting states that all brit-
tle material will experience a transition from ductile mode cutting to brittle mode
cutting when cutting with a depth of cut (DOC) from zero to a large value regard-
less its hardness and brittleness. When cutting below the critical undeformed chip
thickness (UCT), the energy consumed for crack prorogation is larger than that
Fig. 1.1 Schematic diagrams of two cutting modes for brittle material: (1) abrasive grain, (2)
ductile removal of chips, (3) metal phase, (4) transformed amorphous layer, (5) brittle chips, and
(6) microcracks [14]
1.2 Brittle Cutting Versus Ductile Cutting 3
for plastic deformation, ductile mode cutting would be achieved in cutting of brittle
material successfully [15].
The idea of ductile mode machining and its concept appearing in the literature
was firstly reported in 1954 [16], where the material removal in abrasive wear of rock
salt occurred as a result of removing a plastically deformed layer rather than a brit-
tle fracture, although some cracking and fragmentation were still observed. Repro-
ducible results of diamond grinding of glass in a ductile mode was firstly reported in
1976 [17], which considerably improved the surface quality and machining accuracy.
Later, precision grinding of brittle material in a ductile mode had been extended to
others such as silicon and ceramics. Further improvement on ultra-precision machin-
ing technology in the 1990s marked the progression for ductile mode cutting to be
applied in more advanced brittle material such as different types of carbides [18–27].
Ductile mode cutting thus became an alternative way for finishing of brittle material
as it could produce crack-free mirror surface at a much higher efficiency and lower
cost than polishing processes owing to its high material removal rate.
verified by measuring the electrical conductivity of the material near the indenter tip
during the indentation process of brittle material. The measurement results revealed a
substantial increase of material conductivity below the indenter that can be plastically
deformed, which supports the theory of phase transition to a metallic state [34, 35].
One understanding of material removal mechanism can be illustrated by
indentation-sliding analysis [29, 36]. The material removal happens in six stages as
shown in Fig. 1.3: (a) Material under indenter started to subject an elastic deforma-
tion. This creates a small elastic deformation zone due to a high hydrostatic pressure
below the indenter; (b) A median vent formed on a plane at the elastic-plastic bound-
ary; (c) The median vent became stable when further increasing the loading; (d) The
median vent began to close once the loading is removed; (e) Lateral vents formed on
a plane nearly paralleling to the free surface when the indenter moved away. Residual
stresses are the main cause of lateral cracking; and (f) As the indenter is removed
completely, lateral vents extended to the free surface and eventually resulted in the
material removal by fracturing [36]. A possible material removal mechanism can
be classified into two modes, brittle mode or ductile mode, in micromachining and
micro-indentation of brittle material [31, 34]. One is due to plastic deformation in the
characteristic slip direction and another is due to brittle fracture on the characteristic
cleavage plane. When depth of cut becomes smaller, such as in sub-micrometre or
nanometre range, both stresses σ c and τ c increase to the same order as a perfect
material’s intrinsic strength. Thus, plastic deformation takes place before cleaving.
A basic hypothesis was postulated for ductile regime grinding: all materials, regard-
less of their hardness or brittleness, will undergo a transition from brittle machining
regime to ductile machining regime if the grinding infeed rate is made small enough.
Below this threshold of infeed rate, the energy required to propagate cracks is larger
than the energy required for plastic yielding, so plasticity becomes the predominant
grinding mechanism [7]. Figure 1.4 is a schematic diagram illustrated of ductile-to-
brittle transition in grooving of brittle material, where BDT is the abbreviation for
“brittle-to-ductile transition” [13, 37]. Section view B-B for the axial plane through
1.4 Ductile-to-Brittle Transition 5
the groove’s centre and perpendicular to work surface demonstrates that there is
a brittle-to-ductile transition in grooving of brittle material. When depth of cut is
smaller than the critical DOC, ductile mode cutting is achieved with smooth surface
in grooving of brittle material. When depth of cut is larger than the critical DOC,
some fracture occurs on the grooving surface which is call ductile-to-brittle transi-
tion. Further increasing the DOC, brittle mode cutting is achieved with fully fractured
surface.
Some studies of single grit abrasion grinding and micro cutting on myriad brittle
material including semiconductors, glass, crystals and advanced ceramics, demon-
strated similar transitions in the material-removal process as a function of grinding
force and/or depth of cut [8, 31, 38–42]. Evidence of the ductile-to-brittle transition
in grinding of glass appeared as both improvement in surface finish and changes in
6 1 Introduction
Ductile-to-brittle
transition zone
B B
Ductile mode cutting zone BDT zone Brittle mode cutting zone
B-B view
Critical depth of cut
the specific grinding energy [38]. Silicon carbide exhibits a transition from creep
brittleness to creep ductility. Basic theory for stress fields and creep rates around a
crack tip is related to ductile-to-brittle transition in silicon carbide [39]. Investiga-
tion on the initiation and distribution of dislocations and twins in the subsurface of
alumina subjected to single-point scratching indicated that the exist dislocation and
twin systems in the scratched alumina may probably cause microscopic plastic flow
or micro-cracking [43].
Moreover, micro cutting is a viable alternative to grinding and polishing tech-
niques in fabrication of high quality components made by brittle material. The
ductile-to-brittle transition phenomena were observed in micro-indentation and
micromachining on monocrystalline Si and LiNbO3 [31]. Single crystal germanium
wafers of 80 mm in diameter were machined using facing cuts on an ultra-precision
single point diamond turning (SPDT) machine. The chip topography showed a
ductile-to-brittle transition point, which is manifested by the frayed topology along
the thicker portion of the chip [8]. Using different diamond tools with rake angles of
zero degree and negative 25° at different cutting speeds, taper cutting experiments
were carried out with increasing depth of cut on silicon. The cutting groove surface
is changed from ductile mode to brittle mode as the depth of cut exceeds a critical
value [40]. These results suggest that any material, in spite of its ductility, could be
machined in ductile mode under the sufficiently small scale of machining.
1.5 Ductile-to-Brittle Transition Mechanisms 7
In the past decades, many experimental studies have been done on ductile-to-brittle
transition in machining of brittle material, but the nature of ductile-to-brittle transition
is not very clear. Many studies have been developing into understanding ductile-to-
brittle transition phenomenon in machining of brittle material and revealing their
mechanism. For indentation, pyramidal indenter is categorized as sharp-type indenter
and spherical indenter as blunt-type. If indentation-sliding is applied to simulate ultra-
precision cutting, grinding, or polishing, all these indenters do fall into the category
of sharp indenters as its edge radius or grit size is extremely small to be ignored.
Few material removal mechanisms have been proposed such as heavily extrusion
happened ahead of a large edge radius tool [44] and a large negative rake angle [45].
When cutting of brittle material at a depth of cut being sufficiently small, its tool
edge radius r normally in micron scale, will be at the same order as the used depth
of cut ao . As a result, the actual cutting edge will be the arc cutting edge, thereafter
the straight cutting edge will not be involved in cutting regardless its nominal rake
angle γ being positive or negative. In fact, its actual working rake angle γ ne is always
large negative, which is resulting in a large compressive stress in the cutting zone. In
this scenario, work material fracturing due to pre-existing defects will be suppressed
by the large cutting compressive stress undertook in the cutting region, meanwhile
plastic deformation will dominate the chip formation [22]. One view of ductile-to-
brittle transition is based on cleavage fracture due to pre-existing flaws. And a larger
depth of cut would definitely result in a larger undeformed chip thickness, which may
cause the material removal in ductile-to-brittle transition manner [46]. Dislocation
dynamics and dislocation-crack behaviours are the main concerned topics among
ductile-to-brittle transition mechanisms for machining of brittle material.
The dislocation dynamics of plastic flow was proposed as early as 1963, when it
was noticed that the knowledge of dislocations already can be used to calculate
stress-strain curves and other features of the mechanical behaviour under simple
loading conditions, such as the yield point, strain rate sensitivity, delay time and
fracture [47]. Dislocation nucleation controlled model is the one originally proposed
to explain ductile-to-brittle transition and subsequence extensions, in which brittle
or ductile behaviours are resulted from a competition between crack propagation and
spontaneously dislocation emission at the crack tip [48]. A necessary criterion for
brittle fracture in crystals was established in terms of the spontaneous emission of
dislocations from an atomically sharp cleavage crack. Contrary to previous expecta-
tions, an atomically sharp cleavage crack is stable in a wide range of crystal types,
but that in the face centred cubic metals investigated, blunting reactions occur spon-
taneously. Primary nucleation of dislocations in silicon takes place heterogeneously
8 1 Introduction
on defects along the crack tip. A source is easily activated at the intersection point of
the crack front and of an attracted dislocation. The rate of generation of new sources
along the crack front depends on the dislocation’s mobility [49]. Dislocation activity
in the vicinity of a crack tip and ductile-to-brittle transition were analysed using
discrete dislocation dynamics simulations [50].
The possibility of grinding brittle material in a ductile manner was proposed as early
as 1954, when it was noted that during frictional wear of rock salts, although there
was some cracking and surface fragmentation, the dominant material-removal pro-
cess was plastic deformation of the surface layers and not fracture [16]. By 1975,
improvement in precision diamond grinding mechanism allowed the first repro-
ducible evidence of grinding ductility in brittle glass material. Surface ground on
glass material using a silicon carbide wheel exhibited extensive plastic flow over the
surface, while surface ground with diamond wheels appeared to have been generated
by brittle fracture with some evidence of localized plastic flow [17, 62].
The first systematic studies of grinding ductility were performed using a single
grit grinding apparatus. The material-removal regime was shown to progress through
the three stages: plastic grooving, generation of median and lateral cracks, and finally
crushing [26]. It was demonstrated that the progression of material-removal mech-
anism was directly related to the force on the abrasion grain, with lower forces
corresponding to a decrease in the observed surface fracture.
The development of a research apparatus capable of ductile regime grinding was
described and an analytical and experimental investigation of the infeed rates neces-
sary for ductile regime grinding of brittle materials was presented [7]. Observation on
polishing and ultra-precision machining of semiconductor substrate materials indi-
cated that partial ductile grinding following by chemical-mechanical polishing has
many advantages [1]. Semi-ductile grinding and polishing of ophthalmic aspherics
and spherics were carried out to reduce polishing time without an intervening lapping
operation [41]. Grinding of brittle material under certain conditions that allow pre-
dominantly ductile material-removal is a new technology known as ductile-regime
grinding [7, 10]. When brittle material is ground through a ductile regime grind-
ing, surface finishes achieved is similar to those achieved in polishing or lapping
processes. But, grinding is a deterministic material-removal process to permit finely
controlled contour accuracy and complex shapes.
[1, 6–8, 39, 40, 64–66]. This suggests that the process of ductile chip formation may
be independent of material’s nature, e.g., brittle or ductile, hard or soft, crystalline or
amorphous, etc.
Ductile regime response during diamond turning of brittle germanium crystals was
evident from damage-free surfaces obtained. The chip topography provided insight
into the ductile regime machining of germanium that occurred along the tool nose
[8]. Distributed irreversible deformation in otherwise brittle ceramics, such as silicon
carbide and micaceous glass-ceramic, had been observed in Hertzian contacts. An
important manifestation of this deformation is an effective ductility in the indentation
stress-strain response [65]. A germanium surface and the chips produced from a sin-
gle point diamond turning process operated in the ductile regime had been analysed
by transmission electron microscopy and parallel electron-energy-loss spectroscopy.
Lacks of fracture damage on the finished surface and continue chip formation were
indicative of a ductile removal process [66]. The ductile mode machining of com-
mercial PZT (piezoelectric transition) ceramics indicated that the domain switching
is associated with the ductile machinability with PZT ceramics [6].
A technique for predicting the fracture damage zone in single point diamond turn-
ing of brittle material was carried out, and results from the finite element model were
compared with cutting tests on silicon using a commercially available single crystal
diamond (SCD) tool with a negative rake angle of −10°. The critical depth parameter
predicted by the model agreed with the measured fracture damage zone thickness for
a facing operation on silicon [67]. Therefore, experimental and theoretical results
from the above-mentioned literature indicate that brittle material can be machined
in ductile mode cutting.
References
26. Swain MV (1979) Microfracture about scratches in brittle solids. Proc Roy Soc London A,
Math Phy Sci 366:575–597
27. Dolev D (1983) A note on plasticity of glass. J Mater Sci L 2:703–704
28. Finnie I, Dolev D, Khatibloo M (1981) On the physical basis of Auerbach’s law. J Eng Mater
Tech 103:183–184
29. Lawn BR, Evans AG (1977) A model for crack initiation in elastic/plastic indentation fields. J
Mater Sci 12:2195–2199
30. Yan J, Yoshino M, Kuriagawa T et al (2001) On the ductile machining of silicon for micro
electro-mechanical systems (MEMS), optoelectronic and optical applications. Mater Sci Eng
A 297:230–234
31. Shimada S, Ikawa N, Inamura T et al (1995) Brittle-ductile transition phenomena in micro-
indentation and micromachining. CIRP Ann 44:523–526
32. Bridgman P, Šimon I (1953) Effects of very high pressures on glass. J App Phy 24:405–413
33. Sun YL, Zuo DW, Wang HY et al (2011) Mechanism of brittle-ductile transition of a glass-
ceramic rigid substrate. I J Min Metal Mater 18:229–233
34. Clarke DR, Kroll MC, Kirchner PD et al (1988) Amorphization and conductivity of silicon
and germanium induced by indentation. Phy R L 60:2156–2159
35. Gridneva IV, Milman YV, Trefilov VI (1972) Phase transition in diamond-structure crystals
during hardness measurements. Phy St Sol 14:177–182
36. Lawn BR, Wilshaw R (1975) Indentation fracture: principles and applications. J Mater Sci
10:1049–1081
37. Liu K, Li XP, Rahman M et al (2004) A study of the cutting modes in grooving of tungsten
carbide. I J Adv Manu Tech 24:321–326
38. Chandraseker S, Sathyanarayanan G (1987) An investigation into the mechanics of diamond
grinding of brittle materials. In: 15th NAMRC Proceeding, vol 2, pp 499–505
39. Campbell GH, Dalgleish BJ, Evans AG (1989) Brittle-to-ductile transition in silicon carbide.
J Amer Cer Soc 72:1402–1408
40. Fang FZ, Venkatesh VC (1998) Diamond cutting of silicon with nanometric finish. CIRP Ann
47:45–49
41. Moriwaki T, Shamoto E, Inoue K (1992) Ultra-precision ductile cutting of glass by applying
ultrasonic vibration. CIRP Ann 41:141–144
42. Zhong Z, Venkatesh VC (1995) Semi-ductile grinding and polishing of ophthalmic aspherics
and spherics. CIRP Ann 44:339–342
43. Zarudi I, Zhang L (1999) Initiation of dislocation systems in alumina under single-point scratch-
ing. J Mater Res 14:1430–1436
44. Shaw MC (1972) New theory of grinding. Mech Chem Eng T, Ins Eng Australia, pp 73–78
45. Komanduri R (1971) Some aspects of machining with negative rake tools simulating grinding.
I J Mach Too De Res 11:223–233
46. Nakasuji T, Kodera S, Hara S et al (1990) Diamond turning of brittle materials for optical
components. CIRP Ann 39:89–92
47. Hahn GT, Reid CN, Gilbert A (1963) The dislocation dynamics of plastic flow. Proc I Prod
Eng Res Conf Pittsburgh, USA, pp 293–301
48. Rice JR, Thomsom R (1974) Ductile versus brittle behaviour of crystals. Phil Mag 29:73–97
49. Michot G, de Oliveira MAL, Champier G (1999) A model of dislocation multiplication at a
crack tip influencing on the brittle to ductile transition. Mater Sci Eng A 272:83–89
50. Hartmaier A, Gumbsch P (1999) The brittle-to-ductile transition and dislocation activity at
crack tips. J Comp-Ai Mater Des 6:145–155
51. Thomsom RM, Sinclair JE (1982) Mechanics of cracks screened by dislocation. Act Meta
30:1325–1334
52. Ohr SM (1985) An electron microscope study of crack tip deformation and its impact on the
dislocation theory of fracture. Mater Sci Eng 72:1–35
53. Ferney BD, Hsia KJ (1999) The influence of multiple slip systems on the brittle-ductile tran-
sition in silicon. Mater Sci Eng A 272:422–430
14 1 Introduction
54. Lin IH, Thomsom R (1986) Cleavage, dislocation emission, and shielding for cracks under
general loading. Ac Meta 34:187–206
55. Samuels J, Roberts SG, Hirsch PB (1988) The brittle-to-ductile transition in silicon. Mater Sci
Eng A 105(106):39–46
56. Ebrahimi F, Shrivastava S (1997) Crack initiation and propagation in brittle-to-ductile transition
regime of NiAl single crystals. Mater Sci Eng A 239–240:386–392
57. Kimura Y, Pope DP (1998) Ductility and toughness in intermetallics. Inetrmetallics 6:567–571
58. Imayev VM, Imayev RM, Salishchev GA (2000) On two stages of brittle-to-ductile transition
in TiAl intermetallic. Intermetallics 8:1–6
59. John CST (1975) The brittle-to-ductile transition in pre-cleaved silicon single crystals. Phil
Mag 30:1193–1212
60. Rouxel T, Sangleboeuf JC (2000) The brittle to ductile transition in a soda-lime-silica glass.
J N-Cry Sol 271:224–235
61. Chen W, Ravichandran G (2000) Failure mode transition in ceramics under dynamic multiaxial
compression. I J Frac 101:141–159
62. Huerta M, Malkin S (1976) Grinding of glass: surface strength and fracture strength. ASME T
J Eng Ind 98:468–473
63. Toh SB, McPherson R (1986) Fine scale abrasive wear of ceramics by a plastic cutting process.
In: Brookes CA, Warren R, Almond EA (eds) Science of hard materials, Adam Hilger, Boston,
pp 865–871
64. Venkatesh VC, Awaluddin MS, Ariffin AR (1999) The tool life, mechanics, and economics in
conventional and ultra-precision machining. ASME I Mech Eng Con Ex 10:847–854
65. Lawn BR, Padture NP, Cai H et al (1994) Making ceramics ductile. Science 263:1114–1116
66. Morris JC, Callahan DL, Kulik J et al (1995) Origins of the ductile regime in single-point
diamond turning of semiconductors. J Ame Cer Soc 78:2015–2020
67. Strenkowski JS, Hiatt GD (1990) A technique for predicting the ductile regime in single point
diamond turning of brittle materials. Funda Iss Mach: Ame Soc Mech Eng 43:67–80
Part I
Ductile Mode Cutting Fundamentals
Chapter 2
Ductile Mode Cutting Mechanism
2.1 Introduction
Over the past half century, many studies have been conducted to understand
machining fundamentals and behaviours of brittle material. It has been reported
that there is a brittle-to-ductile transition (BDT) in cutting of brittle material when
the undeformed chip thickness is largely reduced from the conventional range [1–4].
Dislocation dynamics and dislocation-crack behaviours have been the main topics
among these subjects. The dislocation dynamics of plastic flow was proposed as early
as 1963 [5]. The dislocation nucleation controlled model is one proposed to explain
brittle-to-ductile transition, where brittle or ductile behaviours are resulted from a
competition between crack propagation and spontaneously dislocation emission at
the crack tip [6]. Contrary to the previous expectation, an atomically sharp cleavage
crack is stable in a wide range of crystal types, but in face cantered cubic metals,
blunting reactions occur spontaneously. The rate of new source generation along the
crack front depends on dislocation’s mobility [7]. Dislocation activity in vicinity of
a crack tip and brittle-to-ductile transition were analysed using discrete dislocation
dynamics simulations [8].
The mechanics of cracks screened by dislocation showed that crack opening dis-
placement is given by the total screening Burgers vector of dislocation cloud [9].
Dislocations were emitted from crack tip during the early stages of crack propaga-
tion and were driven out of the crack tip area, leaving behind a dislocation-free zone
[10]. As cracks moved into the thicker part of the specimen, they often propagated
in a zigzag manner by emitted dislocation on two alternative slip planes. Multiple
slip systems increase crack tip shielding by increasing near tip dislocation density
[11]. Brittle-to-ductile transition in silicon was controlled by the processes with the
same activation energy as that of dislocation motion, and the ductility was a result
of the shielding of cracks by dislocations emitted from a few dislocation sources
at favourable sites along the crack fronts [12]. Crack initiation occurred by the for-
mation of stable microcracks in localized slip bands that were formed in vicinity of
In metal cutting, the rake angle of cutting tool used can be positive or negative,
depending on the cutting processes and work materials. Figure 2.1 shows a shear-
plane model of continuous chip formation with a large negative rake angle, of which
chips are formed in the cutting zone by plastic deformation on tool rake face running
away from tool cutting edge [1]. The cutting tool tip is considered perfect sharp as
a point. Here, AB is shear plane, ls is length of shear plane, v is cutting velocity, vc
is chip flow velocity, vs is shear velocity, ϕ is shear angle, γ is rake angle and β is
mean friction angle on tool rake face. The cutting force relationship is also shown in
Fig. 2.1, where F r is resultant tool force, F c is cutting force, F t is thrust force, F s is
2.2 Cutting Force and Stress in Cutting Zone 19
ls
A
B Fs
ϕ
Ff Fns Fc
Workpiece Fr
Ft
Fn
β O
vs
vc
φ v
shear force on shear plane, F ns is normal force acting on shear plane, F f is frictional
force on tool rake face and F n is normal force acting on tool rake face.
Normally ductile mode cutting of brittle material can be achieved when the unde-
formed chip thickness used is the same order as the cutting edge radius (or called
cutting edge sharpness). In this case, cutting tool tip cannot be considered perfect
sharp, meanwhile an arc cutting edge must be considered involving in the cutting.
Figure 2.2 is a schematic diagram showing an orthogonal view of chip formation in
ductile mode cutting of brittle material with a large negative rake angle and an arc
cutting edge, where DE is tool rake face, BD is arc cutting edge, BC is tool flank
face, O is the centre of arc cutting edge BD, γ is tool rake angle, r is tool cutting edge
radius (or called tool cutting edge sharpness), and ac is undeformed chip thickness
(UCT) [1].
r
D A
C ac
B
Workpiece
20 2 Ductile Mode Cutting Mechanism
Figure 2.3 shows a schematic diagram of cutting forces in ductile mode cutting of
brittle material. Setting the centre O of arc cutting edge BD as the origin of Cartesian
and Polar coordinate systems, X-axis and Y-axis represent horizontal and vertical
directions, respectively [1].
In Fig. 2.3, AB represents the shear zone in ductile mode cutting of brittle material,
which is a curved face rather than a plane for the most cases, K is a random point on
the arc cutting edge BD, ao is chip thickness, dk is an infinitesimally small cutting
edge at the point K within the cutting edge BDE, and α k is the angle between cutting
direction and tangent direction towards cutting movement at point K and is equal
to π /2 + γ k , where γ k is the local rake angle at the point K on the cutting edge.
It can be seen that both γ and γ k are in large negative values. K is a point on the
curved shear face AB corresponding to the point K on the arc cutting edge BD, ϕ k is
the shear angle corresponding to the infinitesimally small random cutting edge dk.
dFf K dFc Ff K Fc
dFr Fr
dFt Ft
dFn Fn
Therefore, the cutting edge BDE can be expressed in both Polar coordinate system
and Cartesian coordinate system as the following [1, 17]:
In Polar coordinate system, for the cutting edge BD:
x = r cos θ π
− ≤θ ≤γ <0 (2.1)
y = r sin θ 2
Hence, the first order derivative of the above cutting edge expressions could be
written as the following. For the cutting edge BD, it is given by:
∂y x
=√ = cot(−θ ) − r ≤ y ≤ r sin γ (2.5)
∂x r − x2
2
∂y
= − cot γ y > r sin γ (2.6)
∂x
Based on the above expressions, for an infinitesimally small random cutting edge
dk at the random point K within the cutting edge BDE, the angle α k between its
tangent direction and X-axis direction is given by [1, 17]:
⎧
⎨ αk = arctan ∂ y = arctan √ x = π2 + θ ac ≤ r (1 + sin γ )
∂ x r −x
2 2
Therefore, the rake angle γ k of the infinitesimally small cutting edge dk is given by:
π
γk = αk − (2.8)
2
22 2 Ductile Mode Cutting Mechanism
i.e.
γk = θ ac ≤ r (1 + sin γ )
(2.9)
γk = γ ac > r (1 + sin γ )
As shown in Fig. 2.3 based on the cutting geometry, the shear angle ϕ k corre-
sponding to the infinitesimally small cutting edge dk with the rake angle γ k within
the cutting edge BDE is given by:
rc cos γk
tan ϕk = (2.10)
1 − rc sin γk
i.e.
rc cos γk
ϕk = arctan (2.11)
1 − rc sin γk
k AB dt1 k AB dk
d Fr = = (2.12)
sin ϕk cos(ϕk + β − γk ) cos(ϕk + β − γk )
l
k AB w
Fr = d Fr k = dk (2.13)
cos(ϕk + β − γk )
0
For different undeformed chip thickness, the resultant cutting forces F r are given
by [17, 18]:
⎧
⎪
⎪ γ ac +r sin γ
⎪
⎪ F = k AB r
dγk + k AB
dy ac > r (1 + sin γ )
⎨ r
−π
cos(ϕk +β−γk )
0
cos(ϕ+β−γ ) cos γ
2
arcsin( arc −1) (2.14)
⎪
⎪
⎪
⎩ Fr =
⎪ ac ≤ r (1 + sin γ )
k AB r
cos(ϕk +β−γk )
dγk
− π2
2.2 Cutting Force and Stress in Cutting Zone 23
i.e.
⎧
⎪ γ
⎪
⎪ F = k AB r
dγk + k AB (ac +r sin γ )
ac > r (1 + sin γ )
⎪
⎨ r −π cos(ϕk +β−γk ) cos(ϕ+β−γ ) cos γ
2
arcsin( arc −1) (2.15)
⎪
⎪
⎪
⎩ Fr = ac ≤ r (1 + sin γ )
k AB r
⎪ cos(ϕk +β−γk )
dγk
− π2
Substituting Eq. (2.11) into Eq. (2.15), the resultant cutting forces F r for different
undeformed chip thickness are given by:
⎧
⎪ γ
⎪
⎪ F = k AB r dγk + k AB (ac +r sin γ )
ac > r (1 + sin γ )
⎪
⎨ r r cos γ
cos arctan 1−rc c sin kγ +β−γk cos(ϕ+β−γ ) cos γ
−π 2 k
(2.16)
Since the shear zone AB in ductile mode cutting of brittle material is a curved face,
it is very difficult to express the cutting forces by using the tool rake angle, because
the tool geometry is represented by the arc cutting edge together with tool rake face.
As a result, the actual rake angle along the cutting edge BD varies from −π /2 to
γ monotonically. Therefore, an equivalent tool rake angle is used to simplify the
mathematical expressions for ductile mode cutting model. Figure 2.4 schematically
shows the equivalent rake angle of cutting edge when the arc cutting edge BD per-
forms cutting operation. Here using the centre O of the arc cutting edge BD as the
origin of Cartesian coordinates system, horizontal and vertical directions are set as
X axis and Y axis, respectively [1].
When undeformed chip thickness is ac ≤ r (1 + sin γ ) as shown in Fig. 2.4a, for
simplicity, the arc cutting edge BG is substituted by the chord BG so as to calculate the
equivalent rake angle. That is, the angle between chord BG and line BO is considered
as the equivalent rake angle γ e for this situation. Here, γ e is given by:
π ac
γe = − + arctan √ (2.17)
2 (2r − ac )ac
When undeformed chip thickness is ac > r (1 + sin γ ) as shown in Fig. 2.4b, for
simplicity, the arc cutting edge BD is substituted by the chord BD, and the straight
cutting edge DG is regarded as no variation for calculating of the equivalent rake
−→
angle. That is, the angle between vector BG and Y-axis is considered as the equivalent
24 2 Ductile Mode Cutting Mechanism
O D X
γe G
C
ac
Workpiece B
Y E
Tool γ
O
G X
C γe
ac
D
Workpiece B
−→ −→
rake angle γ e . From the Cartesian coordinate system, vectors B D and DG are given
by:
−→ −
→ −
→
B D = r cos γ j + (r sin γ + r ) k (2.18)
and
−→ −
→ −
→
DG = [r (1 + sin γ ) − ac ] tan γ j + [ac − r (1 + sin γ )] k (2.19)
respectively.
−→
By adding the two vectors together, vector BG is given by:
−→ −→ −→ −
→ −
→
BG = B D + DG = {[r (1 + sin γ ) − ac ] tan γ + r cos γ } j + ac k (2.20)
−→
Thus, the angle between vector BG and Y-axis is considered as the equivalent rake
angle γ e for this scenario. Here, γ e is given by:
π ac
γe = − + arctan (2.21)
2 [r (1 + sin γ ) − ac ] tan γ + r cos γ
From the cutting geometry as shown in Fig. 2.3, the equivalent shear angle ϕ e
corresponding to the equivalent rake angle γ e can be obtained as:
2.2 Cutting Force and Stress in Cutting Zone 25
rc cos γe
tan ϕe = (2.22)
1 − rc sin γe
i.e.
rc cos γe
ϕe = arctan (2.23)
1 − rc sin γe
The mean normal stress acted on the curved shear face AB, σ s , can be derived as [1,
17]:
where F c is cutting force along cutting direction and F t is thrust force normal to
cutting direction as shown in Fig. 2.3.
Thus, the apparent shear stress τ s on the curved shear face AB in ductile mode
cutting can be obtained as:
where ϕ e is the equivalent shear angle corresponding to the equivalent rake angle
γ e which is used to simplify the mathematical expressions for the actual rake angle
along the arc cutting edge BD, and varied from –π /2 to γ monotonically.
Comparing Eq. (2.24) with Eq. (2.25), the following expression can be obtained:
Usually, in cutting of brittle material as shown in Fig. 2.3, since the cutting tool
rake angle is large negative,
and the undeformed
chip thickness is smaller than the
tool cutting edge radius Ft = − tan γe , the thrust force F t is larger or much larger
Fc 1
than the cutting force F c , and the equivalent shear angle ϕ e is extremely small.
Therefore, according to Eq. (2.26), the mean normal stress σ s is much larger than
the shear stress τ s on the curved shear face.
26 2 Ductile Mode Cutting Mechanism
Many studies have been done on fracture mechanisms and fracture toughness of
brittle material [19–22]. Griffith was an early pioneer who successfully analysed the
fracture-dominant problem and considered the propagation of brittle cracks in glass
in 1920. In the middle of 1950s, Irwin modified the Griffith theory, according to
which fracture occurs when a critical stress distribution ahead of crack tip is reached
[23]. Then, cracking behaviours and fracture mechanics of brittle material have been
widely investigated during the past decades. It is well agreed that the stress intensity
factor and fracture toughness largely affect crack propagations for brittle material
[24–30].
Due to different fabrication processes, brittle material more or less contains pre-
existing flaws, such as point defect, dislocation, crack, pore, inclusion, segregation
and centre, which give rise to incompatible deformation, etc. It has been shown that
the two most important defects affecting the material failure are crack and dislocation
[20]. Especially, for brittle material such as tungsten carbide, glass, quartz, silicon
and calcium fluoride, crack is the most important defect affecting its failure during
cutting operation. Furthermore, those cracks create a localized stress concentration
around them. Under the action of a crack driving force, also called stress intensity
factor K I , these flaws would essentially cause fracture through flaw extension, such
as crack propagation, when the undertaking tensile stresses exceed a limit.
Crack propagation has been well documented as the research subject in the field
of linear elastic fracture mechanics and fatigue of work material. Usually, there are
three possible modes of crack propagation generally identified by the subscripts I,
II and III as shown in Fig. 2.5: (a) mode I also called opening mode, (b) mode II
usually called sliding mode, and (c) mode III also referred to tearing mode [20].
Y σy
τxy
dy σx
b
dx X
θ
2a
In practice, among these three crack propagation modes, the most applicable mode
is the opening mode (mode I).
In linear elastic fracture mechanics, the stress conditions concerning the opening
mode (mode I) of crack propagation of length 2a in an infinite plate, can be described
in terms of a stress intensity factor K I about the crack tip [20]. As shown in Fig. 2.6,
the plate is subjected to a tensile stress σ at infinity. An element dxdy of the plate at
a distance b from the crack tip and at an angle θ between X-axis and the line from
crack tip to the element with respect to the crack plane, experiences normal stresses
σ x and σ y in X and Y directions and a shear stress τ xy .
By considering the stress intensity factor for the opening mode (mode I) K I in the
derivation of the following stress field equations [23]:
⎛ ⎞
σx ⎛ ⎞
⎜ ⎟ 1 − sin θ2 sin 3θ2
⎜ ⎟ K θ
⎜ σy ⎟ = √ I
cos ⎝ 1 + sin θ2 sin 3θ2 ⎠ (2.27)
⎜ ⎟ 2
⎝ ⎠ 2π b
sin θ2 cos 3θ2
τx y
σz = ν(σx + σ y ) for plane strain
σz = 0 for plane stress
where σ z is normal stress in Z direction of the element, and ν is Poisson’s ratio. Here,
for ductile mode cutting of brittle material as shown in Fig. 2.3, the normal tress σ z
in Z direction of the element is zero.
As shown in Fig. 2.7a, an internal through crack of length 2a is situated in a
plate of finite width W and is subjected to a tensile stress σ at the boundary, where
the crack is perpendicular to the tensile stress σ . The stress intensity factor K I is
expressed as [17, 31]:
√ a
K I = Y σ πa f (2.28)
W
28 2 Ductile Mode Cutting Mechanism
W W Ȧ ș
ș
2a
2a
ı ı
(a) Non-inclined crack (b) Inclined crack
where Y is a geometric factor, and f Wa is a function of crack size which is expressed
by a trigonometrical function as [31]:
a
W 1
πa 2
f = tan (2.29)
W πa W
a π 2a2 2π 4 a 4
21
f = 1+ + +··· (2.30)
W 3W 2 15W 4
Therefore, in an infinite body the stress intensity factor for the opening mode (mode
I) K I is given by
21
√ π 2a2 2π 4 a 4
K I = Y σ πa 1 + + +··· (2.31)
3W 2 15W 4
In practice, the crack may be at an angle to the applied stress direction as shown
in Fig. 2.7b, where ω represents crack angle. Then, for an infinite body the stress
intensity factor for the opening mode (mode I) of crack propagation K I is given by
[31]:
√ a
K I = Y σ πa f sin2 ω (2.32)
W
Substituting Eq. (2.30) into Eq. (2.32) gives:
21
√ π 2a2 2π 4 a 4
K I = Y σ πa 1 + + + · · · sin2 ω (2.33)
3W 2 15W 4
2.3 Material Fracture in Cutting Zone 29
The above mathematic formulae indicate that the stress intensity factor for the
opening mode (mode I) K I is dependent on the loading configuration and the geom-
etry of the crack system.
The fracture toughness of brittle material in chip formation zone is based upon
dislocation activated by normal stress and shear stress in cutting zone. For enhancing
2a
σs
30 2 Ductile Mode Cutting Mechanism
In chip formation zone during cutting of brittle material, there coexists the probability
of crack propagation and dislocation extension. Modern treatments of yielding and
plastic flow are based upon the elementary atomic displacements responsible for
permanent set. Permanent deformation of crystalline material involves the growth of
slipped regions under the action of shear stress τ s . Plastic deformation, which can
be described by using the growth of a slip-line field, is thereby accomplished with
the movement of dislocation. The yield flow strength is determined by the stress
required to generate dislocation and move them through the crystal. The permanent
set or plastic strain constituted by dislocation can be expressed mathematically as
[5]:
ε p = 0.5bρ d̄ (2.34)
ε̇ p = 0.5bρ v̄ (2.35)
ρ = ρo + f Cε p (2.36)
where ρo represents the density of mobile dislocation occurred in the undeformed
crystal, C is a constant depended upon the crystal material, and f is a fraction.
Normally, the average dislocation velocity v̄ is related to the applied tensile stress
σ in the form of:
m
σ
v̄ = (2.37)
σo
2.4 Fracture Toughness Enhancement 31
In 1934, Taylor produced the first detailed theory of work hardening related to dis-
location. According to the basic idea of Taylor work hardening theory, the yield
strength depends on the internal stresses opposing to the movement of dislocation.
It is supposed that most dislocations do not pass completely through a crystal, but
through elastic interaction with other dislocation and through obstruction provided
by mosaic boundaries, they become stuck inside. These stuck dislocations cause
internal stresses, which derives to the raising of yield strength.
The Taylor work hardening model in dislocation theory provides a simple mean-
field description of the dislocation interaction processes at the micro-scale [32, 33].
The work hardening dislocation theory indicates that the Peach-Koehler shear stress
τ pair due to the interaction of a pair of dislocations at a distance L is proportional:
μb
τ pair ∝ (2.38)
L
where μ is shear modulus, b is Burgers vector and L is mean dislocation spacing.
This sets a critical applied stress to break or untangle the interactive pair dislocations
so that slip can occur even if one of the dislocations is pinned by an obstacle. Slip
begins at random points in the crystal and occurs by the separation of one positive
and one negative dislocation at each of these points. These dislocations move apart
by the average distance L and then become stuck, so that when their dislocation
density reaches ρT , they have produced a plastic strain γ given by:
γ = ρT bL (2.39)
The shear stress for a single-slip system of a single crystal depends upon the total
dislocation density ρT ,
1
ρT = (2.40)
L2
Hence, the plastic strain γ is given by:
32 2 Ductile Mode Cutting Mechanism
b
γ = (2.41)
L
Therefore, the Taylor relationship between shear stress τ on the slip plane and dis-
location density is given by:
αt μb √
τ= = αt μb ρT (2.42)
L
where α t is an empirical material coefficient on the order of one and usually ranges
from 0.2 to 0.5.
The total dislocation density ρT is the sum of the densities of statistically stored
dislocations ρ S , and geometrically necessary dislocations ρG , i.e.
ρT = ρ S + ρG (2.43)
Substituting Eq. (2.43) into Eq. (2.42), the relationship between shear stress and total
dislocation density is given by:
√
τ = αt μb ρ S + ρG (2.44)
If the von Mises rule of distortion energy is used, the tensile flow stress can be written
as:
√ √ √
σ = 3τ = 3αt μb ρ S + ρG (2.45)
σ = σr e f f (ε) (2.48)
2.4 Fracture Toughness Enhancement 33
where σ ref is a reference stress in uniaxial tension and f is a function of strains. For
most ductile materials, the function f can be written as a power-law relation:
f (ε) = ε N (2.49)
where
2
ε= εi j εi j (2.50)
3
is the effective strain and N is plastic work hardening exponent (0 ≤ N < 1).
For the hardening resulted from the statistically stored dislocation alone, and in
the absence of the strain gradient term (see Eq. (2.46)), the uniaxial stress-strain law
(see Eq. (2.45)) becomes:
√ √
σ = 3αt μb ρ S = σY f (ε) (2.51)
1
εi j = (u i, j + u j,i ) (2.52)
2
and the second gradient of displacement is given by:
where ∂¯i is forward gradient operator. Strain tensor εij and strain gradient tensor ηi jk
have the symmetry εi j = ε ji and ηi jk = η jik . The deviatoric strain tensor εi j and
deviatoric strain gradient tensor ηi jk are:
34 2 Ductile Mode Cutting Mechanism
1
εi j = εi j − εkk δi j (2.54)
3
and
respectively, where the hydrostatic part of strain gradient tensor ηiHjk is given by:
1
ηiHjk = (δik η j pp + δ jk η j pp ) (2.56)
4
where the characteristic material length l for strain gradient plasticity in the above
equation can be determined from shear modulus μ and Burgers vector b, using the
equation [35, 36]:
2
μ
l= 3αt2 b (2.59)
σY
Therefore, when the chip formation zone in cutting of brittle material is at mesoscale,
due to the large strain gradient, the work material normal flow stress σ will be
increased such that the fracture toughness of the material is increased.
In summary, in cutting of brittle material with the undeformed chip thickness
being sufficiently small [38] and the ratio of tool cutting edge radius to undeformed
chip thickness being larger than 1, ductile mode cutting can be achieved to form
continuous chips and generate smooth surface. This is because chip formation will
be dominated by dislocation rather than fracturing due to the following three effects.
• First, extremely small undeformed chip thickness formed by the cutting process
geometry, and with the ratio of tool cutting edge radius to undeformed chip thick-
ness being larger than 1, work material in cutting zone undertakes extremely large
compressive stress and shear stress, with the compressive stress being much larger
than the shear stress. This stress status produces a largely reduced stress intensity
factor K I and activates dislocation emission in the material.
2.4 Fracture Toughness Enhancement 35
References
1. Liu K (2002) Ductile cutting for rapid prototyping of tungsten carbide tools. NUS Ph.D. thesis,
Singapore
2. Ngoi BKA, Sreejith PS (2000) Ductile regime finish machining – a review. Int J Adv Manuf
Technol 16:547–550
3. Neo KW, Kumar AS, Rahman M (2012) A review on the current research trends in ductile
regime machining. Int J Adv Manuf Technol 63:465–480
4. Antwi EK, Liu K, Wang H (2018) A review on ductile mode cutting of brittle materials. Front
Mech Eng 13:251–263
5. Hahn GT, Reid CN, Gilbert A (1963) The dislocation dynamics of plastic flow. In: Proceedings
of the international production engineering research conference, Pittsburgh, USA, pp. 293–301
6. Rice JR, Thomsom R (1974) Ductile versus brittle behaviour of crystals. Philos Mag 29:73–97
7. Michot G, de Oliveira MAL, Champier G (1999) A model of dislocation multiplication at a
crack tip influencing on the brittle to ductile transition. Mater Sci Eng A 272:83–89
8. Hartmaier A, Gumbsch P (1999) The brittle-to-ductile transition and dislocation activity at
crack tips. J Comput-Aided Mater Des 6:145–155
9. Thomsom RM, Sinclair JE (1982) Mechanics of cracks screened by dislocation. Acta Metall
30:1325–1334
10. Ohr SM (1985) An electron microscope study of crack tip deformation and its impact on the
dislocation theory of fracture. Mater Sci Eng 72:1–35
11. Ferney BD, Hsia KJ (1999) The influence of multiple slip systems on the brittle-ductile tran-
sition in silicon. Mater Sci Eng A 272:422–430
12. Samuels J, Roberts SG, Hirsch PB (1988) The brittle-to-ductile transition in silicon. Mater Sci
Eng A 105(106):39–46
13. Ebrahimi F, Shrivastava S (1997) Crack initiation and propagation in brittle-to-ductile transition
regime of NiAl single crystals. Mater Sci Eng A 239–240:386–392
14. Imayev VM, Imayev RM, Salishchev GA (2000) On two stages of brittle-to-ductile transition
in TiAl intermetallic. Intermet 8:1–6
15. Chen W, Ravichandran G (2000) Failure mode transition in ceramics under dynamic multiaxial
compression. Int J Fract 101:141–159
16. Blackley WS, Scattergood RO (1994) Chip topography for ductile-regime machining of ger-
manium. ASME Trans J Eng Ind 116:263–266
17. Liu K, Li XP (2001) Modelling of ductile cutting of tungsten carbide. Trans NAMRI/SME
29:251–258
18. Liu K, Li XP (2001) Ductile cutting of tungsten carbide. J Mater Process Technol 113:348–354
19. Broek D (1984) Elementary engineering fracture mechanics. Martinus Nijihoff Publishers,
Springer, Netherlands, The Hague
20. Ewalds HL, Wanhill RJH (1989) Fracture mechanics. Edward Arnold, London
21. Jayatilaka A de S (1979) Fracture of engineering brittle materials. Appl Sci Lond:19–115
22. Meyers MA (1994) Dynamic behaviour of materials. Wiley, New York, pp 488–566
23. Irwin GR (1957) Analysis of stress and strain near the end of a crack traversing a plate. ASME
Trans J Appl Mech 24:361–364
24. Kendall K (1976) Interfacial cracking of a composite. J Mater Sci 11:1267–1269
25. Pisarenko GS, Krasowsky AY, Vainshtock VA et al (1987) The combined micro- and macro-
fracture mechanics approach to engineering problems of strength. Eng Fract Mech 28:539–554
26. Weertman J (1978) Fracture mechanics: a unified view for Griffith-Irwin-Orowan cracks. Acta
Metall 26:1731–1738
27. Pook LP (1985) The fatigue crack direction and threshold behavior of mild steel under mixed
mode I and III loading. Int J Fatigue 7:21–30
28. Topper TH, Yu MT (1985) The effect of overloads on threshold and crack closure. Int J Fatigue
7:159–164
29. Strenkowski JS, Hiatt GD (1990) A technique for predicting the ductile regime in single point
diamond turning of brittle materials. Fundam Issues Mach: Am Soc Mech Eng 43:67–80
References 37
30. Smith A, Nurse A, Graham G et al (1996) Ultrasonic cutting – a fracture mechanics model.
Ultrasonics 34:197–203
31. Liu K, Li XP, Liang SY (2007) The mechanism of ductile chip formation in cutting of brittle
materials. Int J Adv Manuf Technol 33:875–884
32. Cottrell AH (1953) Dislocations and plastic flow in crystals. The Clarendon Press, Oxford
University
33. Kovacs I, Zsoldos L (1973) Dislocations and plastic deformation. Pergamon Press, Oxford, pp
252–283
34. Fleck NA, Hutchinson JW (1997) Strain gradient plasticity. In: Hutchinson JW, Wu TY (eds)
Advances in applied mechanics, vol 33. Academic Press, New York, pp 295–236
35. Gao H, Huang Y, Nix WD et al (1999) Mechanism-based strain gradient plasticity – I. Theory.
J Mech Phys Solids 47:1239–1263
36. Huang Y, Gao H, Nix WD et al (2000) Mechanism-based strain gradient plasticity – II Analysis.
J Mech Phys Solids 48:99–128
37. Shi MX, Huang Y, Hwang KC (2000) Plastic flow localization in mechanism-based strain
gradient plasticity. Int J Mech Sci 42:2115–2131
38. Bifano T, Bierden PA (1997) Fixed-abrasive grinding of brittle hard disk substrates. Int J Mach
Tools Manuf 37:935–946
Chapter 3
Ductile Mode Cutting Characteristics
3.1 Introduction
In cutting of brittle material such as silicon, quartz, glass and ceramics with
conventional machining conditions, chip formation is usually a fracture process that
damages its machined workpiece surface and leads to an unacceptable surface qual-
ity. However, it has been found that ductile mode chip formation can be achieved by
having the right cutting conditions and tool geometry, where both machined surface
finish and form accuracy are better ensured. It has been reported that there is a brittle-
to-ductile transition (BDT) in cutting of tungsten carbide material when undeformed
chip thickness is largely reduced from the conventional range [1–6]. Ductile mode
cutting of brittle material is depended on its stress state in cutting region: whether
or not its shear stress in chip formation region is larger than critical shear stress for
chip formation (τsli p > τc ), and whether or not fracture toughness of work material
is larger than its stress intensity factor (K I < K C ). When τsli p < τc and K I > K C ,
crack propagation dominates its chip formation. Thus, the cutting mode obtained is
a brittle mode [7].
In this chapter, ductile mode cutting characteristics and material removal
behaviour of brittle material are investigated systematically through grooving and
milling tests using tungsten carbide as an example work material. Cutting mode
transitions in grooving of tungsten carbide, machined surfaces and chip formations,
cutting forces in ductile mode cutting and tool wear mechanism are presented and
discussed.
12.7mm
v
Cutting tool
10μm
Chuck
Fz α Crack
Commercial tungsten carbide inserts are used as workpiece, which is made by powder
metallurgy technology using sub-micron powders such as tungsten carbide, cobalt
and etc. Its grade of this commercial tungsten carbide insert is A30, or its ISO
application code is P30. Tungsten carbide insert is a standard square shape. Its insert
dimension is 12.7 mm × 12.7 mm × 4.76 mm. And its nominal chemical composition
of the tungsten carbide insert is listed in Table 3.1. Table 3.2 provides typical physical,
thermal and mechanical properties of tungsten carbide insert used [9, 10].
Cubic boron nitride (CBN), hardest material next to diamond, is used as cutting tools
in this test, since its room temperature Vickers hardness is approximately twice that
42 3 Ductile Mode Cutting Characteristics
of most hard tool materials. While its hardness (~1800 HV) at 1000 °C is similar to
that of tungsten carbide at room temperature, and CBN has a higher wear resistance
than ceramics, good thermal conductivity, and good chemical and thermal stability
up to about 800 °C. A MB730 grade CBN material is used as the cutting tool because
of its capability for rough and finish machining of difficult-to-machining materials
even under high cutting speed. It contains 80% CBN by volume with a TiC ceramic
binder phase in this grade. It is made by bonding a 0.5–1 mm layer of CBN onto a car-
bide substrate by sintering under pressure. While carbide provides shock resistance,
CBN layer provides very high wear resistance and cutting edge strength. Table 3.3
summarizes the CBN tool material properties at room temperature [10].
Grooving Direction
A
300μm
3.3 Grooving Surface Morphology 43
presents grooving direction [1, 4]. As the grooving starts from depth of cut being
zero then increased, machined workpiece surface is smooth at the beginning, and
then changed to be rough in the region near section A-A, with cracks propagating
into workpiece. Based upon experimental result as shown in Fig. 3.3, brittle-to-
ductile (BDT) transition in grooving of brittle material is schematically illustrated in
Fig. 1.4. Cross-section B-B is the axial plane through the centre of machined groove
and perpendicular to the workpiece surface. Three type’s surfaces are obtained on
machined groove surface indicating that three cutting regimes are generated including
ductile mode cutting zone, transition cutting zone and brittle mode cutting zone. This
clearly demonstrates that there is a transition from ductile mode cutting to brittle mode
cutting in grooving of tungsten carbide as the depth of cut being increased from zero.
When depth of cut is below a critical value, chip formation happens under ductile
mode cutting. As depth of cut exceeds a critical value, chip formation occurs under
brittle mode cutting.
Surface characteristics and damage morphology of tungsten carbide workpiece in
the grooving tests are shown in Fig. 3.4: (a) ductile mode grooved surface, (b) semi-
brittle fractured surface and (c) brittle fractured surface, where grooving direction is
from left to right [1, 4]. Surface characteristics and damage types of tungsten carbide
workpiece produced by the CBN tool consist of deformed and displaced material at
the grooving edge, cracking within the groove, cracking extending outward from the
groove edge and material pull-out or fracture failure at the groove end. As shown
in Fig. 3.4, lateral cracking, radial cracking and median cracking from groove edge
at the groove end, and overloads acting on the groove, lead to large lump material
removal in grooving of tungsten carbide.
A theoretical model has been proposed to predict the ductile-to-brittle transition
in grooving of tungsten carbide material based upon analyses of cutting forces, tool
geometry and cutting process geometry, temperature depended hardness and fracture
mechanics [2, 3], which indicates that there is a critical depth of cut in grooving
of tungsten carbide work material. The above experimental results well verify the
proposed prediction model for the ductile-to-brittle transition in cutting of tungsten
carbide.
The measured critical values of depth of cut of the eight grooves vary within
a range. This is attributable to the unavoidable flaws and defects within tungsten
carbide workpiece. OMIS and scanning electron microscope (SEM) observations of
some macro-cracks and micro-cracks within tungsten carbide workpiece are shown in
Fig. 3.5 [1, 2]. Figure 3.6 are SEM micrographs of tungsten carbide sample surfaces,
showing some pores and segregations within tungsten carbide workpiece [1, 2].
As shown in Fig. 3.5a, the length of the macro-crack is in a millimeter scale, but
as shown in Fig. 3.5b, the lengths of micro-cracks are in a micrometer scale. This
indicates that the crack length within tungsten carbide workpiece varies in a wide
range. Within one tungsten carbide workpiece, number of cracks n, length of each
crack 2a and crack angle ω are random and uncertain. For a finite plate, different
number of cracks cause the different width of infinite body W. Therefore, according
to Eq. (2.33), for tungsten carbide workpiece even under the same stress field, the
values of stress intensity factor K I could be different, as well as the critical depth
of cut obtained in grooving. That is, critical depth of cut varies in a range, which
depends upon flaws and defects within tungsten carbide workpiece.
(a) Pores within the sample (b) Segregations within the sample
Material removal mechanisms in cutting of tungsten carbide are classified into two
modes. One is the process due to dislocation forming plastic deformation on the char-
acteristic slip plane and the other is due to crack propagation along the characteristic
cleavage plane. When the resolved shear stress τ slip in the easy slip direction exceeds
a certain critical value τ c inherent to workpiece material, and the stress intensity
factor K I is less than material’s fracture toughness K C so as to ensure a cleavage
would not take place, a plastic deformation takes place in a small stressed field within
workpiece material with a specified scale. The plastic slip traces on machined groove
surface are well shown in Fig. 3.4a. On the other hand, a cleavage may take place by
crack propagation when the stress intensity factor K I exceeds the material fracture
toughness K C preceding a plastic deformation. SEM micrographs of semi-brittle and
brittle fractured surfaces of the machined groove as shown in Fig. 3.4b and c, indicate
that brittle fracture could occur in cutting of tungsten carbide under certain situations.
Interpretations of the plausible reasons for crack growth acceleration due to
compressive overloads have centered on crack closure concepts. The compressive
overloads is postulated to lead to flattening of the fracture surface asperities [4].
Observations of fracture surface abrasion induced by compressive overloads have
been reported for some alloy [11]. Figure 3.7 is the SEM micrographs of frac-
tured surfaces of the machined tungsten carbide [1, 4]. The SEM micrographs show
compression-induced abrasion marks (marked in Fig. 3.7b by the letters ‘A’ and
‘B’) in an area immediately behind one crack tip. In this case, the crack is prop-
agated ahead of a stress concentration under fully compressive loads until crack
arrest occurred. The abrasion marks correspond to the crack growth region where
compressive overloads are applied.
Note that the abrasion slide lines on the two different marked sections ‘A’ and
‘B’ on the fracture surface should be parallel, but actually they are not parallel, as
shown in Fig. 3.7b. In fact, the marked sections ‘A’ and ‘B’ are in a whole within
46 3 Ductile Mode Cutting Characteristics
A B
a micro-crack before grooving deformation, which can be testified from the right
side configuration of section ‘A’ and the left side configuration of section ‘B’ as
shown in Fig. 3.7b. In the grooving, two sections undertake compressive overloads
and the compressive-included abrasion flattening occurs; meanwhile the micro-crack
extends and the section ‘B’ is stripped off the section ‘A’ (grooving direction from
left to right) so that the abrasion slide lines on the two sections should not be parallel.
The influence of compressive overloads on the growth of cracks is strongly depen-
dent on the micro mechanism of crack growth, in particular on prior to development
of crack surface morphology and roughness-induced crack closure during the groov-
ing of tungsten carbide work material. Thus, compressive overloads are likely to have
a more dominant effect on crack propagations to produce fracture failure in cutting
of tungsten carbide.
The chip formation mode in grooving of tungsten carbide depends on the stress
state in the chip formation region under certain cutting conditions, i.e. whether or
not τsli p > τc and K I < K C . When cutting conditions being varied, especially
for undeformed chip thickness and cutting tool geometry, as well as cutting region
temperature, cutting speed and feed rate, the critical values of K I and τ c would be
changed.
Milling tests are conducted on tungsten carbide at different undeformed chip thick-
nesses to investigate their ductile mode cutting characteristics as well. Experimental
results show that there is a transition from brittle mode cutting to ductile mode cutting
when reducing the maximum undeformed chip thickness d max from few micrometres
to few hundred nanometres. Brittle mode cutting and ductile mode cutting are iden-
tified according to chips formed and machined surface integrity. SEM photograph
of the formed chips and machined surface under the cutting conditions of cutting
3.5 Material Removal Mode 47
speed of 296.6 m/min (4000 rpm), feed rate of 0.02 mm/rev and depth of cut of 2 µm
(calculated maximum undeformed chip thickness d max is 1.164 µm), are shown in
Fig. 3.8a and b, respectively [1, 7]. Chips are formed in a shape of particles by crack
propagation causing fracture in cutting zone. Machined surface is covered by cracks
and fracture. These indicate that its chip formation is under brittle mode cutting.
When cutting with the maximum undeformed chip thickness d max being reduced to
920 nm (feed rate of 0.015 mm/rev and other cutting conditions unchanged), chips
formed become continuous. Since continuous chips are formed by dislocation, this
indicated that their chip formation is under ductile mode cutting. Figure 3.9 shows
a SEM photograph of one chip formed in ductile mode cutting. It can be seen that
the chip is formed in the same way as that for cutting of metals, where materials are
removed from workpiece by dislocations generating chips in layers contacted to each
other [1, 7]. When cutting with the maximum undeformed chip thickness d max being
further reduced to 338 nm (feed rate of 0.005 mm/rev and other cutting conditions
unchanged), continuous chips are also generated. SEM photographs for these chips
are shown in Fig. 3.10: (a) serrated continuous chips, (b) continuous chip, and (c)
close-up view of continuous chip [1, 7]. Machined surface is shown in Fig. 3.11,
which is a smooth surface without any crack [1, 7]. Both continuous chips formed
and smooth surface generated indicate that ductile mode cutting is obtained.
Experimental results indicate that in cutting of tungsten carbide there is a critical
value for undeformed chip thickness, at or below which chips are formed in ductile
mode cutting—which generates continuous chips by dislocation in cutting zone.
This agrees well with the theoretical analysis discussed in Chap. 2. Here, the first
condition for ductile mode cutting of brittle material is to have such a small value of
undeformed chip thickness so that compressive stress in cutting zone is large enough
to suppress its stress intensity factor K I to be smaller than fracture toughness K C [1,
7]. In above milling tests for cutting of tungsten carbide, to achieve ductile mode
cutting, the maximum undeformed chip thickness used has to be 920 nm or smaller.
The results also indicate that in cutting of tungsten carbide, ductile mode cutting
is achieved when undeformed chip thickness is smaller than tool cutting edge radius,
which is the second condition as described in Chap. 2. In above milling tests, the ratio
of tool cutting edge radius to undeformed chip thickness for achieving ductile mode
3.5 Material Removal Mode 49
cutting is 6.5, showing that the undeformed chip thickness used is much smaller than
tool cutting edge radius.
Cutting forces for milling of tungsten carbide are also examined using a dynamome-
ter. When cutting at the speed of 296.6 m/min (4000 rpm), feed rate of 0.015 mm/rev
and depth of cut of 2 µm (calculated maximum undeformed chip thickness d max
is 920 nm), three cutting force components during two revolutions are recorded as
shown in Fig. 3.12a [1, 7]. Maximum force components are: F x = 10 N (cutting force
F c ), F y = 83 N and F z = 143 N (thrust force F t ). Here, thrust force F t is much larger
than cutting force F c . When cutting with the maximum undeformed chip thickness
d max being reduced to 338 nm (feed rate is 0.005 mm/rev and other cutting conditions
unchanged), three cutting force components during two revolutions are also recorded
as shown in Fig. 3.12b [1, 7]. Maximum force components are: F x = 6 N (cutting
force F c ), F y = 58 N and F z = 101 N (thrust force F t ). Again, thrust force F t is
much larger than cutting force F c . These indicate that in ductile mode cutting thrust
force is much larger than cutting force, so that large compressive stress is generated
in cutting zone, which shields the growth of pre-existing flaws in work material by
suppressing the stress intensity factor K I . This agrees well with theoretical analysis
as described in Chap. 2.
50 3 Ductile Mode Cutting Characteristics
Examination of the CBN tool wear after cutting of tungsten carbide material is
carried out using a SEM and an energy dispersive X-ray spectrometer (EDS). SEM
close-up view of CBN tool flank face is shown in Fig. 3.13 [1, 6]. It is revealed that
some grooves with abrasive traces are formed on CBN tool flank face, indicating
a typical abrasion wear. The cause of the appearance may be attributed that the
soft binder of CBN tool is abraded by hard carbide particles of tungsten carbide
workpiece. Meanwhile, grooving traces are observed on tool flank face by using
SEM investigation.
Figure 3.14 shows SEM photograph and EDS spectrum of CBN tool flank face
after cutting of tungsten carbide under cutting speed of 593.1 m/min: (a) SEM pho-
tograph of tool flank face, and (b) EDS spectra at site A and (c) EDS spectra at site
B [6]. SEM examination of tool flank face indicates that there is a layer formed on
the cutting edge as shown in Figs. 3.13 and 3.14a. It is found that the layer is likely
to be a solid solution of work material and cutting tool material. EDS examination at
the white site A on tool wear surface reveals that its elements are mainly W, Co, Ti
and C. Its element percentage analysis shows that W is 52.58%, Co is 17.17%, Ti is
3.7 Tool Wear Mechanisms 51
Rake Face
Flank Face
(a) Tool wear on frank face (b) Close-up view of frank wear
9.76% and C is 20.49%. It seems that the layer is more likely to be a piece of tungsten
carbide adhered to CBN tool surface. This wear behavior is a typical adhesion wear.
EDS examination at the black site B on tool flank face reveals that its elements
are mainly B, N, W, Co, Ti and C. However, the main elements are B and N. It should
be noted that elements B, C and N are not easy to be detected due to the absorption
of low-energy X-rays by the windows of the EDS detector. Only when they present
in substantial amounts could they be detected [12]. Elements from work material
tungsten carbide have been detected on CBN tool wear surface by the EDS analysis,
which indicates that the elements from work material have been diffused into cutting
tool material during the cutting processes. This wear behavior is a typical diffusion
wear. SEM and EDS examinations on tool wear surfaces suggest that during cutting
of tungsten carbide using CBN tools, tool wear mainly occurs on flank face. And
tool wear mechanisms are dominated by diffusion, adhesion and abrasion.
Ductile mode cutting characteristics and material removal mechanism of brittle mate-
rials are investigated through grooving and milling tests using tungsten carbide as
an example work material in this chapter. Experimental results demonstrate that in
cutting of tungsten carbide, there is a transition from ductile mode cutting to brittle
mode cutting when depth of cut being increased from zero to a certain value. Similar
experimental results for grooving of soda-lime glass [13–16], ZKN7 glass [17], fused
silica glass [18], BK7 glass [19] and single crystalline silicon [19, 20] are achieved as
well. SEM observations on groove surfaces show that there are three cutting modes
generated in grooving of tungsten carbide when depth of cut being increased, said
ductile mode cutting (DMC), semi-brittle mode cutting (SMC) and brittle mode cut-
ting (BMC). In general, smooth surface generated and continuous chips formed are
used to verify its material removal mode in cutting of brittle material being DMC.
52 3 Ductile Mode Cutting Characteristics
A
B
While fractured surface generated and particle chips formed are used to verify its
material removal in cutting of brittle material being BMC.
And in ductile mode cutting of tungsten carbide, thrust force F t is much larger
than cutting force F c , which results in a large compressive stress in cutting zone [21].
Large compressive stress and shear stress could shield the growth of pre-existing
flaws in work material by suppressing its stress intensity factor K I , such that K I < K C
making work material is able to undertake a large cutting stress without fracturing to
3.8 Concluding Remarks 53
achieve ductile mode cutting. SEM and EDS examinations on cutting tools indicate
that tool wear mainly occurs on flank face and tool wear mechanisms are dominated
by diffusion, adhesion and abrasion in cutting of tungsten carbide.
References
1. Liu K (2002) Ductile cutting for rapid prototyping of tungsten carbide tools. NUS Ph.D. thesis,
Singapore
2. Liu K, Li XP (2001) Modelling of ductile cutting of tungsten carbide. Trans NAMRI/SME
29:251–258
3. Liu K, Li XP (2001) Ductile cutting of tungsten carbide. J Mater Process Technol 113:348–354
4. Liu K, Li XP, Rahman M et al (2004) A study of the cutting modes in grooving of tungsten
carbide. Int J Adv Manuf Technol 24:321–326
5. Liu K, Li XP, Liang SY (2004) Nanometer scale ductile cutting of tungsten carbide. J Manuf
Process 6:187–195
6. Liu K, Li XP, Rahman M et al (2003) CBN tool wear in ductile cutting of tungsten carbide.
Wear 255:1344–1351
7. Liu K, Li XP, Liang SY (2007) The mechanism of ductile chip formation in cutting of brittle
materials. Int J Adv Manuf Technol 33:875–884
8. Li XP, Rahman M, Liu K et al (2003) Nano-precision measurement of diamond tool edge radius
for wafer fabrication. J Mater Process Technol 140:358–362
9. Upadhyaya GS (1996) Nature and properties of refractory carbides. Nova Science Publishers,
New York, pp 213–292
10. Pierson HO (1996) Handbook of refractory carbides and nitrides: properties, characteristics,
processing and applications. Noyes Publications, New Jersey, pp 100–116
11. Topper TH, Yu MT (1985) The effect of overloads on threshold and crack closure. Int J Fatigue
7:159–164
12. Barry J, Byrne G (2001) Cutting tool wear in the machining of hardened steels, part II: cubic
boron nitride cutting tool wear. Wear 247:152–160
13. Liu K, Li XP, Liang SY et al (2005) Nanometer scale ductile mode cutting of soda-lime glass.
J Manuf Process 7:95–101
14. Moriwaki T, Shmoto E, Inoue K (1992) Ultraprecision ductile cutting of glass by applying
ultrasonic vibration. CIRP Ann 41:141–144
15. Antwi EK, Liu K, Wang H (2018) A review on ductile mode cutting of brittle materials. Front
Mech Eng 13:251–263
16. Liu K, Li XP, Liang SY et al (2004) Nanometer scale ductile mode cutting of soda-lime glass.
Trans NAMRI SME 32:39–45
17. Fang FZ, Chen LJ (2000) Ultra-precision cutting for ZKN7 glass. CIRP Ann 49:17–20
18. Zhou M, Wang XJ, Ngoi BK et al (2002) Brittle ductile transition in the diamond cutting of
glasses with the aid of ultrasonic vibration. J Mater Process Technol 121:243–251
19. Fang FZ, Venkatesh VC (1998) Diamond cutting of silicon with nanometric finish. CIRP Ann
47:45–49
20. Zhang JG, Zhang JJ, Cui T et al (2017) Sculpturing of single crystal silicon microstructures by
elliptical vibration cutting. J Manuf Process 29:389–398
21. Liu K, Li XP, Rahman M (2003) Characteristics of high speed micro cutting of tungsten carbide.
J Mater Process Technol 140:352–357
Chapter 4
Modelling of Ductile Mode Cutting
4.1 Introduction
A theoretical prediction model of critical undeformed chip thickness for ductile mode
cutting of brittle material is presented and described in detail in this chapter, in which
the critical value of undeformed chip thickness for ductile mode cutting of brittle
material can be predicted from the work material properties, cutting tool geometry
and cutting conditions.
The nature of ductile-to-brittle transition in cutting of brittle material implies that
their physical characteristics would vary when the surface energy is varied. Therefore,
the predictive model for ductile mode cutting of brittle material is developed based
upon the following aspects: cutting forces in chip formation zone, heat generation
and temperature rise in the cutting region, relationship between work material hard-
ness and temperature, and relationship between work material hardness and fracture
toughness.
A frame chart of the theoretical prediction model for ductile mode cutting of brit-
tle material is shown in Fig. 4.1 [1]. Firstly, cutting forces and mean stresses in the
cutting region are predicted from work material properties and cutting conditions
used. Secondly, temperature-dependent hardness of work material is obtained from
cutting forces and work material properties. Furthermore, the fracture toughness of
work material is predicted from mean stresses in the cutting region and work material
characteristics. Finally, the undeformed chip thickness in cutting of brittle material at
the brittle-to-ductile transition region is predicted from hardness, mechanical prop-
erties and characteristics of work material.
Mean Stress in
Cutting Condition Cutting Region
Equivalent Tool Angle
Cutting Tool Geometry Cutting Force Critical Undeformed
Chip Thickness or
Critical Depth of Cut
Work Material
Hardness
Fig. 4.1 Frame chart of theoretical modelling of ductile mode cutting of brittle material
Figure 4.2 schematically shows an orthogonal view of ductile mode cutting of brittle
material with a large negative rake angle and an arc cutting edge [1]. From Fig. 4.2,
cutting force can be predicted from brittle material properties, cutting tool geometry
and cutting conditions. For different undeformed chip thickness, resultant cutting
Tool Chip
γ
O
X
r
D A
C K
ac
B Ff Fc
Fr
Workpiece Ft
Fn
β
4.2 Cutting Force and Mean Stress 57
forces in ductile mode cutting of brittle material can be determined from Eq. (2.16).
Therefore, the mean normal stress σ s and the apparent shear stress τ s acted on the
curved shear plane produced by cutting forces during the cutting of brittle material,
could be determined from Eqs. (2.24) and (2.25), respectively. According to the
analysis derived from Eq. (2.26), brittle material undertakes very large stresses around
the contacted region between work material and cutting tool in the cutting process.
It is well known that there is a heat generation and temperature rising in the cutting
zone due to energy consumption during machining [2]. As shown in Fig. 4.2 in
the cutting zone, the rate of energy consumption in a cutting process Pm can be
determined from cutting force and cutting velocity, i.e.
Pm = Fc v (4.1)
Pm = Ps + P f (4.2)
where Ps is heat generation rate in chip formation zone (shear zone heat rate) and Pf
is heat generation rate at tool-chip interface (frictional heat rate).
The rate Pf generated by friction between the chip and the tool is given by:
P f = F f vc = F f vrc (4.3)
Pf
θ f = (1 − n c ) (4.4)
ρcvac aw
58 4 Modelling of Ductile Mode Cutting
θm = Aθ f (4.5)
Substituting Eq. (4.3) into the above equations, following equation is obtained:
AF f rc
θm = (1 − n c ) (4.6)
ρcac aw
where Γ is the proportion of heat conducted from chip formation zone into work
material.
Therefore, the temperature in chip formation zone θ e can be determined as:
θe = θm + θs + θo (4.8)
i.e.
(1 − )Fc + ( − 1 + A − n c A)F f rc
θe = + θo (4.9)
ρcac aw
Here, tungsten carbide is used as an example of brittle material to estimate its hard-
ness at elevated temperature. The hot hardness of single crystal WC on all major
crystallographic orientations evaluated decreases rapidly when increasing tempera-
ture, and the single crystal hardness on its hardest orientation is only about half of the
polycrystalline material depending on the test temperature [6]. The WC-Co compos-
ite suffers a strength loss due to oxidization and microstructure defects when raising
temperature, as well as a change from an essentially brittle failure mode at a lower
temperature to increasingly more ductile failure at a higher temperature occurred for
brittle material [11].
Within the temperature range from 20 to 1000 °C, the hardness of tungsten carbide
decreases with increasing WC grain size as a result of increasing the temperature [9,
10], and follows the relationship:
H = Ho + K y d −1/2 (4.10)
where H is hardness of work material, d is mean WC grain size, and H o and K y are
functions of the properties of individual phases, i.e. composition and microstructural
parameters of work material, given by
and
1 − Vwc
B= (4.13)
Vwc (1 − C)
Fracture toughness is a material property of the ability to resist the growth of a pre-
existing crack or flaw. Scientists have done experimental investigations and tests on
fracture toughness of brittle material widely [13–17]. Investigation of the elevated
temperature properties of WC-Co material indicated that the critical toughness value
of brittle material K C can be calculated from a microstructural model as a function
of physical properties of brittle material [18–24]:
0.6
E E −0.6 −1.5
KC = 1.55μ2.5 bCo εc
0.5 0.6
1 + 0.012 H (4.15)
H H
where μ is shear modulus for cobalt, bCo is Burgers vector of dislocation for cobalt, E
is elastic modulus, and εc is a critical value of composite strain at fracture. The plas-
ticity index (E/H )0.6 varies from 8.5 to 11.5 slowly for the most WC-Co composites
[18].
For WC-Co composites, when the hardness H > 10000 MPa, the expression
0.5
1.55μ2.5 bCo takes the numerical value of 2.67 × 107 , and εc = 0.015 [19], then
Eq. (4.15) becomes:
0.6 −0.6
E E
K C = 2.15 × 106 1 + 0.012 H −1.5 (4.16)
H H
As analysed in Chap. 2, cracks will propagate into work material and then brittle
fractures will occur once the stress intensity factor for the opening mode K I exceeds
a critical value of fracture toughness K C in cutting of WC work material.
ER
dc = (4.17)
H2
where d c is critical undeformed chip thickness or critical indentation depth, H is mate-
rial hardness and R is material fracture energy. One approach to define the fracture
energy at small scales is to replace it with a dimensionally analogous measurement
of the energy needed to propagate cracks, namely:
4.6 Critical Undeformed Chip Thickness 61
K C2
R∼ (4.18)
H
for the undeformed chip thickness in cutting or depth of cut in grooving at the
ductile-to-brittle transition region in cutting of brittle material. Expression (4.19)
can be written as
2
E KC
dc = u (4.20)
H H
Ze
Center Damage transition line
conditions are achieved. Cross-feed f determines the position of depth d c along the
tool nose. Larger values of feed f derives the depth d c moving closer to the tool
central line [29].
Based on the cutting geometry shown in Fig. 4.3, the following equation is derived
to determine the critical undeformed chip thickness d c and damage depth yc with the
given cutting tool nose radius R, tool feed f and brittle-to-ductile transition location
Z e:
Z e2 − f 2 dc2 dc + yc
= − 2 (4.21)
R2 f2 R
Grooving direction
Radial crack
Ductile mode
Semi-brittle mode
Brittle mode
Fig. 4.4 Illustration of various damage types associated with the grooving
4.6 Critical Undeformed Chip Thickness 63
groove, cracking extending outwards from the groove edge, and material pull-out
or fracture failure at the groove end. As shown in Figs. 3.3 and 4.4, a large lump
material is removed at the end of the machined groove due to lateral cracking, radial
cracking and median cracking and overloads acting on the groove during the brittle
material grooving process.
Firstly, to determine the critical undeformed chip thickness, at or below which the
chip is formed under ductile mode cutting. Secondly, to examine the ratio of cutting
force F c to thrust force F t in cutting zone when undeformed chip thickness is smaller
than tool cutting edge radius.
Figure 4.5 shows a schematic diagram of maximum undeformed chip thickness
in cutting process [1, 32]. Here, O1 and O2 are the centres of two adjacent arc cutting
edges, and the distance between O1 and O2 is feed rate f used. The undeformed chip
thickness is achieved by arranging combinations of cutting
tool nose radius R, depth
of cut ao and feed rate f, as shown in Fig. 4.5. When 2Rao − ao2 ≤ f as shown in
Fig. 4.5a, the maximum undeformed chip thickness d max can be simplified using the
equation:
dmax = ao (4.22)
When 2Rao − ao2 > f as shown in Fig. 4.5b, the maximum undeformed chip
thickness d max can be determined using the equation [32]:
O1 O2 O1 O2
f
R R
f
dmax
ao
dmax ao
(a) Large feed rate 2 Ra o − a o2 ≤ f (b) Small feed rate 2 Ra o − a o2 > f
The groove formation develops a plastically deformed zone as side ridges around
the groove. Therefore, the apparent groove depth does not simply mean the volume
actually removed. Typical profile of a groove is shown in Fig. 4.6 [1, 32, 33], where
AV is the area of a machined groove measured in the cross-section of the groove and
(A1 + A2 ) is the area of the work material pushed by plastic deformation onto the
groove edges.
The amount of work material removal produced by the tool AW , is calculated by
subtracting the cross-section areas of the ridge A1 and A2 from the cross-section area
of the groove AV , as sketched in Fig. 4.6. The formula is
A W = A V − (A1 + A2 ) (4.24)
AW A V − (A1 + A2 )
f ab = = (4.25)
AV AV
when
f ab = 0 Ideal micro-ploughing (the surface is deformed without any material
removal), which is also called rubbing or plowing. In this scenario, A1
+ A2 = AV . That is, the material of area AV is fully pushed away from the
groove to form two side ridges with the areas of A1 and A2 .
f ab = 1 Ideal micro-cutting with the material removal volume directly proportional
to the cross-sectional area of the machined groove. In this scenario, A1 +
A2 = 0. That is, the material of area AV is totally removed by the cutting
tool.
f ab > 1 Micro-cracking for brittle material only. In this scenario, A1 + A2 < 0. That
is, more material than the area AV is removed by fracturing.
The machined tungsten carbide sample surfaces are examined firstly using an optical
measurement inspection system (OMIS) and a scanning electron microscope (SEM).
The OMIS photographs of the grooved tungsten carbide sample surface are shown
in Fig. 4.7. SEM micrographs on full view of one machined groove of tungsten
carbide work material are shown in Fig. 4.8 [1]. Figure 4.9 are SEM micrographs of
the groove surfaces machined in ductile mode near the brittle-to-ductile transition
region [1, 3, 4].
From Figs. 4.7 and 4.8, it can be seen that as the cutting starts from depth of
cut of zero, then the depth of cut increases continuously. The machined tungsten
carbide surface is smooth at the beginning but then changed in the region near to
section A-A to be rougher with cracks propagating into the workpiece. This indicates
that there is a transition from ductile cutting mode to brittle cutting mode during the
grooving of tungsten carbide material as shown in Fig. 4.9. More details can be seen
from the SEM observations of the tungsten carbide workpiece in the brittle-to-ductile
transition region as shown in Fig. 4.9: (a) ductile cutting mode, and (b) brittle cutting
mode.
Cutting Direction
100μm
Ductile Mode Brittle mode
(a) Overview of the machined groove (b) Close-up view of transition region
Fig. 4.7 OMIS photographs of the grooved tungsten carbide sample surface
66
Fig. 4.8 SEM micrographs on full view of one machined groove of tungsten carbide work material [1]
4 Modelling of Ductile Mode Cutting
4.9 Grooving Verification 67
Fig. 4.9 SEM micrographs of cutting modes on the grooved tungsten carbide sample surface
Figure 4.10 shows different types of chips formed in grooving of tungsten carbide
work material: (a) continuous chips formed in ductile mode cutting, and (b) fragment
chips formed in brittle mode cutting. The cutting conditions for chips formed as
shown in Fig. 4.10 are: (a) cutting speed of 144 m/min, depth of cut of 4µm and dry
grooving; (b) cutting speed of 144 m/min, depth of cut of 8µm and dry grooving,
respectively [1, 3]. During the grooving, when the depth of cut is below a critical
value, chips are formed under ductile mode cutting. More details can be found from
the OMIS and SEM photographs of the machined tungsten carbide sample surface
in the transition region as shown in Figs. 4.7, 4.8 and 4.9a, and as well as continuous
chips formed under ductile mode cutting in the grooving as shown in Fig. 4.10a.
When the depth of cut is larger than the critical value, chips are formed under brit-
tle mode cutting. More details can be found from the OMIS and SEM photographs
of the machined tungsten carbide sample surface in the transition region as shown
Fig. 4.10 Different type chips formed in grooving of tungsten carbide work material
68 4 Modelling of Ductile Mode Cutting
in Figs. 4.7, 4.8, and 4.9b, and as well as fragment chips formed under brittle mode
cutting in the grooving as shown in Fig. 4.10b. All those indicate that there is a tran-
sition from ductile mode cutting (DMC) to brittle mode cutting (BMC) in grooving
of tungsten carbide.
The critical depth of cut for the transition from ductile to brittle chip formation in
grooving of tungsten carbide is determined by measuring the cross-section profile
of the machined groove at the brittle-to-ductile transition region A-A using a surface
stylus profiler.
The brittle-to-ductile transition region A-A is identified by examining the
machined groove surface as shown in the OMIS photograph and SEM micrographs
(see Figs. 4.7, 4.8 and 4.9). The criterion for identifying the location of transition
section A-A is set as in between the end of smooth surface and the start of rough
surface on the machined groove. The surface profiles at the cross-section A-A in
the transition regions for two machined grooves are shown in Fig. 4.11: (a) the 4th
machined groove and (b) the 6th machined groove [1, 3, 4]. Viewing from Fig. 4.11,
the critical depths of cut for the 4th and 6th grooves are 4.082 µm and 4.772 µm,
respectively. Eight such critical values of those machined grooves are obtained in
the grooving experiment. These critical values of depth of cut are listed in Table 4.1
[1, 3, 4]. The average critical depth of cut is 4.761 µm. That is, in general when
grooving tungsten carbide, the cutting is transited from ductile mode to brittle mode
as the depth of cut exceeds the critical value of 4.761 µm, which is smaller than the
tool cutting edge radius of 5.8 µm [37].
For the same cutting process, the theoretical model as described in Chap. 2 is used
to predict the value of the critical depth of cut. The fracture toughness of tungsten
carbide at room temperature is 12.8 MPa m1/2 [38]. Substituted the material properties
of tungsten carbide such as elastic modulus, hardness and critical fracture toughness
listed in Table 3.2 into Eq. (4.20), together with the value of the brittle-to-ductile
transition factor u of 0.15 for ceramics [25], the theoretical value for critical depth
of cut in grooving of tungsten carbide is determined as 4.55 µm [1, 4].
In comparison with the experimental results as listed in Table 4.1, the theoret-
ical value for critical depth of cut is 5% higher. It should be pointed out that the
experimentally measured values of the critical depth of cut could vary in a range,
depending on the identified location of section A-A at the transition region. Since the
start of rough surface with fracture marks could not be clearly identified, the location
of section A-A could more or less vary in a region. On the other hand, a factor that
can have contributed the error to the theoretical value of critical depth of cut is the
brittle-to-ductile transition factor u. Due to the value u for tungsten carbide work
material not available, a value u for ceramics is used in the calculation. As tungsten
carbide is less brittle compared to ceramics, it can be expected that if the properly
4.9 Grooving Verification 69
determined value u for tungsten carbide work material is used, the predicted result
would agree better with the experimental result.
The surface profile of the machined 2nd groove of tungsten carbide workpiece at
the transition region from ductile mode to brittle mode (section A-A) is measured
using the surface stylus profiler, as shown in Fig. 4.12 [1]. Width w and depth h of
the machined groove at the section A-A are also obtained from the measured groove
profile being as 127.5 µm and 4.995 µm, respectively. The amount of work material
removal AW and the ratio of work material removal f ab can be calculated from
Eqs. (4.24) and (4.25), respectively. Here, the ratio of the volume of work material
removal to the volume of the machined groove for the given groove as shown in
Fig. 4.12, f ab , is obtained as 0.940.
In ductile chip formation, the percentage of cutting against ploughing has been
measured using the material removal ratio f ab —the average ratio of the volume of
work material removed to the volume of the machined groove. The ratios of work
material removal for the eight grooves are listed in Table 4.2. The average ratio
at transition region from ductile mode to brittle mode is f ab = 0.938. Clearly, the
grooving of tungsten carbide work material at the brittle-to-ductile transition section
A-A is almost under an ideal-cutting mode. It can be concluded that tungsten carbide
work material could be machined in ductile mode with certain cutting conditions.
4.10 Concluding Remarks 71
References
1. Liu K (2002) Ductile cutting for rapid prototyping of tungsten carbide tools. NUS Ph.D. thesis.
Singapore
2. Boothroyd G, Knight WA (1989) Fundamentals of machining and machine tools. Marcel
Dekker, New York
72 4 Modelling of Ductile Mode Cutting
3. Liu K, Li XP (2001) Ductile cutting of tungsten carbide. J Mater Proc Tech 113:348–354
4. Liu K, Li XP (2001) Modelling of ductile cutting of tungsten carbide. T NAMRI/SME
29:251–258
5. Kuhn H, Medlin D (1988) ASM handbook V 8. ASM International Materials Park, Novelty
6. Lee M (1983) High temperature hardness of tungsten carbide. Metall Trans A Phys Meta Mater
Sci 14:1625–1629
7. Schaller R, Ammann JJ, Bonjour C (1988) Internal friction in WC-Co hard metals. Mater Sci
Eng A105(106):313–321
8. Raghunathan S, Caron R, Freiderichs J et al (1996) Tungsten carbide technologies. Adv Mater
Proc 149:21–23
9. Milman YV, Chugunova S, Goncharuck V et al (1997) Low and high temperature hardness of
WC-6 wt%Co alloys. Int J Refract Metal Hard Mater 15:97–101
10. Milman YV, Luyckx S, Northrop IT (1999) Influence of temperature, grain size and cobalt
content on the hardness of WC-Co alloys. Int J Refract Metal Hard Mater 17:39–44
11. Acchar W, Gomes UU, Kaysser WA (1999) Strength degradation of a tungsten carbide-cobalt
composite at elevated temperatures. Mater Charac 43:27–32
12. Uygur ME (1997) Modelling tungsten carbide/cobalt composites. Adv Mater Proc 151:35–36
13. Bolton JD, Keely RJ (1983) Fracture toughness (Kic) of cemented carbides. Fib Sci Tech
19:37–56
14. Shetty DK, Wright IG, Mincer PN et al (1985) Indentation fracture of WC-Co cermets. J Mater
Sci 20:1873–1882
15. Han D, Mecholsky JJ (1990) Fracture analysis of cobalt-bonded tungsten carbide composites.
J Mater Sci 25:4949–4956
16. James MN, Human AM, Luyckx S (1990) Fracture toughness testing of hard metals using
compression-compression precracking. J Mater Sci 25:4810–4814
17. Schubert WD, Neumeister H, Kinger G et al (1998) Hardness to toughness relationship of
fine-grained WC-Co hardmetals. Int J Refract Metal Hard Mater 16:133–142
18. Laugier MT (1987) Palmqvist toughness in WC-Co composites viewed as a ductile/brittle
transition. J Mater Sci L 6:768–770
19. Laugier MT (1987) Comparison of toughness in WC-Co determination by a compact tensile
technique with model predictions. J Mater Sci Lett 6:779–780
20. Laugier MT (1987) Hertzian indentation of ultra-fine grain size WC-Co composites. J Mater
Sci Lett 6:841–843
21. Laugier MT (1987) Palmqvist indentation toughness in WC-Co composites. J Mater Sci Lett
6:897–900
22. Laugier MT (1988) Elevated temperature properties of WC-Co cemented carbides. Mater Sci
Eng A 105(106):363–367
23. Laugier MT (1989) Validation of the Palmqvist indentation approach to toughness determina-
tion in WC-Co composites. Cera I 15:121–125
24. Laugier MT (1989) Toughness determination in ceramics using sharp and blunt indentation
techniques. Cera I 15:323–325
25. Bifano TG, Dow TA, Scattergood RO (1991) Ductile-regime grinding: a new technology for
machining brittle materials. ASME T J Eng Ind 113:184–189
26. Venkatesh VC, Inasaki I, Toenshof HK et al (1995) Observations on polishing and ultraprecision
machining of semiconductor substrate materials. CIRP Ann 44:611–618
27. Beltrao PA, Gee AE, Corbett J, Whatmore RW (1999) Ductile mode machining of commercial
PZT ceramics. CIRP Ann 48:437–440
28. Blackley WS, Scattergood RO (1994) Chip topography for ductile-regime machining of ger-
manium. ASME T J Eng I 116:263–266
29. Ngoi BKA, Sreejith PS (2000) Ductile regime finish machining—a review. Int J Adv Manu
Tech 16:547–550
30. Venkatesh VC, Awaluddin MS, Ariffin AR (1999) The tool life, mechanics, and economics in
conventional and ultra-precision machining. ASME I Mech Eng Con Ex 10:847–854
References 73
31. Ruff AW, Shin H, Evans CJ (1995) Damage process in ceramics resulting from diamond tool
indentation and scratching in various environments. Wear 181–183:551–562
32. Liu K, Li XP, Rahman M (2003) Characteristics of high speed micro cutting of tungsten carbide.
J Mater Proc Tech 140:352–357
33. Liu K, Li XP, Rahman M et al (2004) A study of the cutting modes in grooving of tungsten
carbide. Int J Adv Manu Tech 24:321–326
34. Liu K, Li XP, Liang YS (2004) Nanometer-scale ductile cutting of tungsten carbide. J Manu
Proc 6:187–195
35. Liu K, Li XP, Rahman M et al (2004) Study of ductile mode cutting in grooving of tungsten
carbide with and without ultrasonic vibration assistance. Int J Adv Manu Tech 24:389–394
36. Zum Gahr KH (1987) Microstructure and wear of materials. Elsevier, Amsterdam, pp 115–146
37. Li XP, Rahman M, Liu K et al (2003) Nano-precision measurement of diamond tool edge radius
for wafer fabrication. J Mater Proc Tech 140:358–362
38. Meyers MA (1994) Dynamic behaviour of materials. Wiley, New York, pp 488–566
Chapter 5
Molecular Dynamics Simulation
of Ductile Mode Cutting
5.1 Introduction
Fig. 5.1 Multiscale modelling methods of simulation and comparison with experimental and ana-
lytical methods
d2 ri d(mvi ) d pi
m 2
= = = Fi (5.1)
dt dt dt
5.1 Introduction 77
F i = −∇i U (r 1 , r 2 , r 3 , . . . , r N ) (5.2)
r i = xi i + yi j + z i k (5.3)
∂ ∂ ∂
∇i = i+ j+ k (5.4)
∂xi ∂ yi ∂z i
E= U1 (ri )
i
+ U2 r i , r j
i< j
+ U3 r i , r j , r k
i< j<k
+ ...
+ U N ri , r j , rk , . . . , r N (5.5)
i< j<k<...<N
the useful simulation results. Each class of materials demands a suitable interatomic
potential that can accurately describe the interacting forces. Several commonly used
potentials refer to Table 5.1.
The fundamental approach to simulating the nanometric cutting process of brittle
materials is to align a group of atoms in the shape of a cutting tool as modelled by
Komanduri and Raff [1] where the cutting tool comprised Newtonian, thermostat and
boundary atoms. Each type of atoms serves a particular purpose for the analysis. For
example, the boundary atoms were defined to be rigid and served as the bulk material
of the cutting tool which was unaffected in an actual machining process. Moving the
layer of boundary atoms toward the workpiece atoms enables the actual cutting
process. In Fig. 5.3, the number of atoms appears to be small but the modelling of an
actual nanometric cutting process incorporates multiple layers of atomic cells that
can prove to be computationally expensive. Like most FEM studies, the tool may
be modelled as a rigid body where tool wear and tool-workpiece interactions are
negligible. From this aspect, the number of atoms in the model may be significantly
reduced to save computational time. However, tool-workpiece interactive potential
functions must be defined to model the effects of tool wear.
Xiao et al. [12] performed molecular dynamics simulations on the cutting process
of silicon carbide, using a diamond tool with a defined cutting edge radius as shown
in Fig. 5.3. It is necessary to study the edge radius effect in the nanometric cutting
of silicon carbide. In other scenarios where the uncut chip thickness ranges above
200 nm, an infinitely sharp diamond cutting tool edge may be assumed. Wang et al.
[13] modelled the cutting tool as an ‘L’ shape in an orthogonal nanometric cutting
process of silicon as shown in Fig. 5.4. The use of an ‘L’-shaped cutting tool neglects
the tool edge radius effect that cannot be observed until the edge radius approaches
the uncut chip thickness.
With respect to brittle materials that often manifest as covalent and ionic crystals,
several key potential functions have to be adopted for accurate representation of
ductile-mode cutting. Silicon has been widely studied with the Tersoff potential that
describes the system energy as:
5.2 Potential Functions for Brittle Materials 81
Fig. 5.4 Molecular dynamics simulation of an orthogonal cutting process of silicon using an ‘L’-
shaped diamond tool. Reprinted by permission from Springer [13] © 2013
1
E= Ei = Vi j (5.6)
i
2 i= j
where the interatomic bond energies V ij are defined by the repulsive pair potential
f R , attractive pair potential f A and the cut-off function f C .
Vi j = f C ri j f R ri j + bi j f A ri j (5.7)
f R ri j = Ai j exp λri j (5.8)
f A ri j = −Bi j exp −μri j (5.9)
⎧
⎪ 1, ri j < R
⎨1 1
f C ri j = 2 + 2 cos (S−R
π ri j −R )
, R < ri j < S (5.10)
⎪
⎩
0, ri j > S
The bond order term is defined in Eq. 5.11, where ζ ij denotes the number of extra
bonds linked to atom i excluding the ij bond, θ ijk represents the bond angles between
ij and ik, and the remaining parameters A, B, R, S, λ, μ, β, n, c, d, and h are material
constants used in the potential.
−1/2n
bi j = 1 + βin ζinj (5.11)
82 5 Molecular Dynamics Simulation of Ductile Mode Cutting
ζi j = f C (rik )ωik g θi jk (5.12)
k=i, j
2
g θi jk = 1 + c2 /d 2 − c2 / d 2 + h − cos θi jk (5.13)
The typical parameters used for modelling of the covalent structure are described
in Table 5.2.
The Vashishta potential is commonly used in modelling of silicon carbide [12, 16–18].
Vashishta et al. [19] developed the Vashishta potential function that accounts for
two- and three-body interactions, where the total potential energy of the system is
calculated by Eqs. 5.14 and 5.15 [16]:
(3)
U= Ui(2)
j ri j + U jik ri j , r jk , rik (5.14)
i< j i, j<k
Hi j Zi Z j − r Di j r Wi j
Ui(2)
j ri j = ηi j + e λ − 4 e− ξ − 6 (5.15)
r r 2r r
where U (2)
ij is the two-body term, H ij represents the steric repulsion strength, Z i and
Z j are the effective electronic charges, Dij is the charge-dipole attraction strength,
W ij is the van der Waals interaction strength, ηij is the steric repulsion exponent
term, and λ and ξ are the screening lengths for Coulomb and charge–dipole terms,
5.2 Potential Functions for Brittle Materials 83
respectively. The three-body form describes the structural changes under pressure
and the melting behaviour of the material, and is written in terms of the spatial and
angular functions [16]:
Ui(3)
jk ri j , rik = R
(3)
ri j , rik P (3) θ jik (5.16)
γ γ
R (3) ri j , rik = B jik exp + r0 − ri j (r0 − rik ) (5.17)
ri j − r0 rik − r0
2
(3)
cos θ jik − cos θ̄ jik
P θ jik = 2 (5.18)
1 + C jik cos θ jik − cos θ̄ jik
where Bjik is the interaction strength, θ jik is the angle between rij and rik , C jik and θ̄
are constants, and (r 0 − r ik ) is the step function. To determine the atomic forces
acting on a particular atom, α can be defined by the summation of the two-body and
three-body components:
∂E
Fα = − = F (2) + F (3) (5.19)
∂rα
∂ (2)
F (2) =− V ri j (5.20)
∂rα i< j i j
∂ (3)
F (3) = − V jik ri j , rik , cos θ jik (5.21)
∂rα i, j<k
The so-called Buckingham potential has been relatively successful at modelling ionic
crystals where the atomic charges play a major role in the interatomic reactions.
The potential energy function in Eq. 5.22 typically consists of three portions—the
Coulombic interactions, a Born-Mayer term, and a subset of the Lennard-Jones poten-
tial. Coulombic interactions (first term in Eq. 5.22) occur where ions i and j have
designated atomic charges qi and qj such that the like-charges will result in repulsion
and opposite charges in attraction, and ε0 is the permittivity of vacuum. These long-
range terms define the ionic interactions between the two ions of different charge
signs such that one is positive, and the other is negative. The second term represents
the Born-Mayer term that considers the repulsive interaction due to the overlap in
84 5 Molecular Dynamics Simulation of Ductile Mode Cutting
electron clouds over shorter interatomic distances. It describes the decay in repul-
sive forces with the increase in interatomic distances. The last term addresses the
short-range van der Waals forces, which takes the form of a Lennard-Jones potential.
1 qi q j ri j Ci j
Ui j = + Ai j exp − − 6 (5.22)
4π ε0 ri j ρi j ri j
Silicon single crystals have been widely studied with molecular dynamics simulation
using the Tersoff potential energy function that is capable of supporting the diamond-
like tetrahedral lattice structure. Cai et al. [25] modelled cutting temperatures (T )
during nanoscale cutting of silicon single crystals by an energy conversion from the
kinetic energy of atoms:
5.3 Simulation for Nano-cutting of Brittle Materials 85
1 3
m i vi2 = N k B T (5.23)
2 i 2
where N represents the number of atoms in the system with individual velocities
vi , and k B is the Boltzmann constant (equal to 1.3806503 × 10−23 J/K). Cutting
temperatures were reported to range from 550–725 K, 320–512 K, and ~315 K, at
Point A in the primary deformation zone, at Point B the tertiary deformation zone, and
at Point C of the bulk crystal, respectively, as shown in Fig. 5.5. These temperature
readings were indicative of the high wear rates of diamond cutting tools during
machining of silicon. It was suggested that the high cutting temperatures promoted
oxidative wear on the diamond that severely reduced the hardness of the tool [25].
Machining forces acting on the workpiece can also be simulated by integrating
the potential energy and summing the interactive forces between a particular atom
and its surrounding atoms. Cai et al. [26] observed the following features during
nanometric machining of silicon:
• Thrust forces were higher than cutting forces
• Tool edge radius had negligible effect on cutting forces but significant impact on
thrust force.
The simulated machining conditions and results are shown in Table 5.4, where d s
is the spring-back thickness of the machined surface by elastic recovery, and d e is
the thickness of damage in the subsurface regions.
Fig. 5.5 Schematic of nanometric chip formation process and subsurface stress distributions [25]
Table 5.4 Simulated forces, elastic recovery and subsurface damage [26]
Rc (nm) F c (10−9 N) F t (10−9 N) d s (nm) d e (nm)
2.5 75.8 156.3 0.6198 1.4394
3.0 83.1 188.8 0.6268 1.7990
4.0 102.4 226.1 0.5643 1.9773
5.0 100.1 260.5 0.5940 2.2899
86 5 Molecular Dynamics Simulation of Ductile Mode Cutting
Dai et al. [27] modelled the vibration assisted cutting of silicon single crystals
at 1.5 nm cutting depths with a vibrational frequency of 60 GHz that follows sinu-
soidal motions. In the elliptical vibration model, the rigid cutting tool vibrated with
amplitudes of 1 and 0.5 nm along the cutting direction and the normal to the cutting
plane, respectively. One-dimensional vibration was also simulated with a vibrational
amplitude of 1 nm along the cutting plane. Their simulations revealed extremely
high temperatures in the produced chip due to the incapacity for heat dissipation at
the nanoscale. However, machined surface temperatures with the implementation of
vibration-assisted machining technology were lower (850–1300 K) as compared to
normal cutting (1200–1600 K).
The hydrostatic stresses were also evaluated, and it was found that the elliptical
vibration-assisted cutting significantly reduced the average hydrostatic stresses from
1.32 GPa and 1.21 GPa, under normal cutting conditions and with 1D vibrations, to
0.36 GPa, respectively. Figure 5.5 displays the simulation results during nanometric
cutting under different vibrational conditions and the resultant hydrostatic stresses.
The successful modelling of the hydrostatic stress distribution deepens our under-
standing on the effects of the widely implemented vibration-assisted machining tech-
nique on the high-pressure-phase-transition (HPPT) which is commonly involved in
machining silicon single crystals [28–30].
The occurrence of phase transformation during nanoscale cutting can be identified
by the coordination number (CN) technique. The β-silicon corresponds to CN = 6
and α-silicon corresponds to the diamond atomic structure CN = 4 [31]. With this
technique, Zhang et al. [31] were able to observe the transformation from α-silicon
to a mixture of β-silicon and amorphous silicon, followed by the recollection of
the original structure upon relief of stresses on the machined surface. The changes
in CN coincide with the change in interatomic distances between silicon atoms as
shown in Fig. 5.6 where the α-silicon has a bond length of 2.35 Å, the fivefold
coordinated system of bct5-Si has a bond length of 2.44 Å, and the bond length of
the six-coordinated β-silicon system measures 2.44–2.58 Å.
Dai et al. [32] also applied the same technique to evaluate the effects of tool
geometry on the subsurface damage. They found that a positive rake angle had lower
Fig. 5.6 Schematic of a the diamond cubic structure, b the fivefold coordinate structure, and c the
six-coordinated tetragonal structure. Reprinted by permission from Springer [32]
5.3 Simulation for Nano-cutting of Brittle Materials 87
Fig. 5.7 Simulation results of nanometric cutting on silicon and the respective phase transforma-
tions in accordance to the coordination number (CN) with different rake angles: a 50°, b 30°, c 0°,
d −30°, and e −50°. Reprinted by permission from Springer [32]
88 5 Molecular Dynamics Simulation of Ductile Mode Cutting
λi + λ j
λi j = (5.24)
2
μi + μ j
μi j = (5.25)
2
Ai j = Ai A j (5.26)
Bi j = Bi B j (5.27)
Ri j = Ri R j (5.28)
Si j = Si S j (5.29)
Wang et al. [34] showcased the capabilities of the Tersoff potential with a wide
variation in cutting conditions of 4H–SiC, such as the change in cutting directions
relative to the crystal orientation, cutting depths and the cutting speed. The increase
in cutting depths promoted the subsurface damage depth (i.e. thickness of the amor-
phous phase beneath the machined surface) and the machining forces. The increase in
cutting forces F c is naturally translated to the increase in the workpiece temperatures
under adiabatic conditions per time increment according to Eq. 5.30 [34]:
(Fc Vc − P)t
T = (5.30)
3/2N k B
bi j + b ji
E= f c ri j V R ri j − V A ri j (5.31)
i> j
2
Fig. 5.9 Modelled wear of the diamond cutting tool edge with sp3 –sp2 transition [33]
5.3 Simulation for Nano-cutting of Brittle Materials 91
where the repulsive and attractive contributions are defined by the dimer energy D0
and the bond length r 0 as:
D0 √
V R ri j = exp −β 2S(r − r0 ) (5.32)
S−1
S D0
V A ri j = exp −β 2/S(r − r0 ) (5.33)
S−1
The cutoff function f c (r), bond order bij , and angular functions g(θ ) are deter-
mined by Eqs. 5.34, 5.35 and 5.37, respectively. Table 5.6 summarizes the parameters
compatible with this potential.
⎧
⎨1
f c ri j = 0 (5.34)
⎩1 π r −R
2
− sin 2
1
2 d
−1/2
bi j = 1 + χi j (5.35)
χi j = f c (rik ) exp 2μ ri j − rik g θi jk (5.36)
k(=i, j)
c2 c2
g θi jk = λ 1 + − 2 (5.37)
d2 d + (h + cos θ)2
High cutting temperatures measured during the simulation are direct indicators of
the occurrence of tool wear, especially in terms of the tool mechanical hardness and
resistance to phase transitions at higher temperatures. Abrasive wear is suggested
to occur due to the deterioration in mechanical hardness at elevated temperatures
Table 5.6 Potential parameters used to model SiC with the ABOP potential [38]
Si–I C–C Si–C Si–II
D0 (eV) 3.24 6.00 4.36 3.24
r 0 (eV) 2.232 1.4276 1.79 2.2222
S 1.842 2.167 1.847 1.57
β (Å−1 ) 1.4761 2.0099 1.6991 1.4760
γ 0.114354 0.11233 0.011877 0.09253
c 2.00494 181.910 273987 1.13681
d 0.81472 6.28433 180.314 0.63397
h 0.259 0.5556 0.68 0.335
2μ (Å−1 ) 0.0 0.0 0.0 0.0
R (Å) 2.82 2.00 2.40 2.90
D (Å) 0.14 0.15 0.20 0.15
92 5 Molecular Dynamics Simulation of Ductile Mode Cutting
Fig. 5.10 Diamond cutting tool: a before cutting, and b after cutting 10.9 nm of SiC. Reprinted
from [41] with permission from Elsevier
Fig. 5.11 a Measured temperature profile during nanometric cutting and b temperature distribution
during orthogonal cutting of SiC with a diamond cutting tool. Reprinted from [41] with permission
from Elsevier
5.3 Simulation for Nano-cutting of Brittle Materials 93
Fig. 5.12 Simulated orthogonal cutting of SiC at very cutting depths: a 40 nm, b 35 nm, c 30 nm,
d 25 nm, e 20 nm, and f 10 nm. Reprinted from [12] with permission from Elsevier
through chip formation and cracks that propagated relatively parallel to the cutting
plane, which indicated shallower surface cracks on the final machined surface.
The cutting edge radius of a single crystal diamond cutting tool (usually ~20 nm)
can be of the same magnitude as the depth of cut during nanometric chip formation
processes (ac ). Naturally, the size effect associated with cutting edge radius occurs
at this length scale and results in the large compressive stresses imposed on the work
material from the effectively negative rake angle. These compressive stresses grad-
ually decrease in magnitude further into the subsurface regions. Figure 5.13 depicts
a typical stress distribution during nanometric cutting, where the mesh represents
individual atoms. The compressive stress is essential for ductile-mode machining
where the stress intensity factors K I at subsurface defects can be reduced and a crack
shielding zone is formed within the compressively stressed region [43]. Therefore,
plasticity by dislocation movement will take precedence within the crack shielding
zone.
The stress distribution in the tool-workpiece interface can be determined through
the analysis of the cutting force distribution. Cutting forces can be derived based on
the effects of the interatomic attraction and repulsive forces on the collection of atoms
and are typically very small and characterized by size effects where the thrust forces
are significantly higher than the cutting forces (along the cutting direction) during
machining at small cutting depths. Luo et al. [44] suggested that the cutting force
acting on the workpiece (denoted by w) can be determined using an example of the
interatomic forces between two atoms with the Morse potential function in Eq. 5.38.
94 5 Molecular Dynamics Simulation of Ductile Mode Cutting
Fig. 5.13 Schematic of nanometric chip formation process and subsurface stress distribution [43]
dUi j
F ri j = − = 2D exp −2α ri j − re − exp −α ri j − re (5.38)
dri j
Nt
Nw Nt
du rwti j
Nw
du rwwi j
Fwi = Fwti j + Fwwi j = − + − (5.39)
j=i j=i j=i
drwti j j=i
drwwi j
Nt
Nw
Nt
du rtti j
Nw
du rtwi j
Fti = Ftti j + Ftwi j = − + − (5.40)
j=i j=i j=i
drtti j j=i
drtwi j
where N w and N t represent the number of atoms in the workpiece and the tool,
respectively.
One may also express the shear and normal stresses at the tool-workpiece inter-
face in terms of the tool geometry and the effective tool-workpiece contact area.
Figure 5.14 shows a cross-sectional view of an orthogonal nanoscale cutting process
where EDB is the cutting tool edge and BC denotes the flank face. The effective
tool-workpiece contact area can be determined by Eq. 5.41, where Rc is the tool edge
radius, ac is the uncut chip thickness, and f is the cutting speed. G represents one of
the multitude of atoms with radius r 0 located at the tool-workpiece interface. Thus,
the totally number of atoms j at the interface can be defined by Eq. 5.42.
5.4 Stress Distribution During Nanoscale Cutting 95
O FT,j
θ FC,j
D G
α
Rc
G C
ac
τ B
Workpiece σ
π R c − ac
A= − arcsin · Rc · f (5.41)
2 ac
j = A/πr02 (5.42)
The cutting forces F C and thrust forces F T can be assumed as evenly distributed
amongst each atom at the tool-workpiece interface as F C,j and F T,j . Hence, the normal
forces F N and shear forces F S can be derived by:
where α is the angle determined by the ratio between the lattice constant (α 0 ) divided
by the tool edge radius (i.e. α = a0 /Rc ). And finally, the normal stresses σ and shear
stresses τ can be calculated as:
σ = FN /a02 (5.45)
τ = FS /a02 (5.46)
To predict the thickness of the crack shielding zone, where material removal is
dominated by dislocation movement, a 2D atomic model was proposed by He et al.
96 5 Molecular Dynamics Simulation of Ductile Mode Cutting
[43] to account for the critical resolved shear stress τ c required for plasticity and the
cleavage stresses σ cleave for brittle failure. The critical tensile stress prior to brittle
failure is given by Eq. 5.47. The critical tensile stress for plastic deformation is given
by Eq. 5.48, which accounts for the Schmid factor m and the Peierls stress τ p for
dislocation motion, and also correlates with the cleavage and compressive stresses
at crack nucleation sites, σ cleave and σ comp , respectively:
2Eγs πρt
σf = (5.47)
π c 8a0
where E is the Young’s modulus, γ s is the surface energy, and ρ t is the radius of
curvature at the tip of a crack of length c. Therefore, the critical forces at the crack
shielding zone boundary can be determined by the minimum compressive stress as:
Considering a chain of atoms and the transmittance of force from atom mi-1 to
mi to mi+1 , each atom is separated according to its interatomic spacing r 0 and held
in equilibrium by a force F proportional to a spring constant and displacement (i.e.
F = kx). Figure 5.15 depicts the 2D atomic model of the (001) plane of silicon by
which the spring constant k can be derived as a function of the interatomic attractive
and repulsive forces in Eq. 5.50. Though the mass-spring model is considered to be
an overly simplified description, it can sufficiently elucidate the crack propagation
behavior during nanoscale cutting [45].
Fig. 5.15 Projection view of the atoms and bonds denoted by a spring constant k, and the relative
local forces [43]
5.4 Stress Distribution During Nanoscale Cutting 97
where
E = σ/ε = d F/r02 /(dr/r0 ) = (1/r0 ) · d F/r (5.51)
The cutting force acting on an atom i, F i , that has been transmitted from the tool-
workpiece interface can thus be calculated by Eq. 5.52. The thickness of the crack
shielding zone L can be determined by Eq. 5.53 through identifying the atom where
the local force F i is equal to the critical force F crit :
Fi = Fi−1 − Ea0 3 − 2/ 1 + (xi /a0 ) (5.52)
L = na0 Fi = Fi−1 − Ea0 3 − 2/ 1 + (xi /a0 ) (5.53)
References
24. Brass A (1989) Molecular dynamics study of the defect behaviour in fluorite structure crystals
close to the superionic transition. Philos Mag A Phys Condens Matter, Struct Defects Mech
Prop 59:843–859
25. Cai MB, Li XP, Rahman M (2007) Study of the temperature and stress in nanoscale duc-
tile mode cutting of silicon using molecular dynamics simulation. J Mater Process Technol
192(193):607–612
26. Cai MB, Li XP, Rahman M (2005) Molecular dynamics simulation of the effect of tool edge
radius on cutting forces and cutting region in nanoscale ductile cutting of silicon. Int J Manu
Technol Manag 7:455
27. Dai H, Chen J, Liu G (2019) A numerical study on subsurface quality and material removal
during ultrasonic vibration assisted cutting of monocrystalline silicon by molecular dynamics
simulation. Mater Res Express 6
28. Yan J, Asami T, Harada H, Kuriyagawa T (2009) Fundamental investigation of subsurface
damage in single crystalline silicon caused by diamond machining. Precis Eng 33:378–386
29. Yan J, Asami T, Harada H, Kuriyagawa T (2012) Crystallographic effect on subsurface damage
formation in silicon microcutting. CIRP Ann—Manuf Technol 61:131–134
30. Yan J, Zhao H, Kuriyagawa T (2009) Effects of tool edge radius on ductile machining of silicon:
an investigation by FEM. Semicond Sci Technol 24
31. Zhang P, Zhao H, Zhang L, Shi C, Huang H (2014) A study on material removal caused by phase
transformation of monocrystalline silicon during nanocutting process via molecular dynamics
simulation. J Comput Theor Nanosci 11:291–296
32. Dai H, Chen G, Fang Q, Yin J (2016) The effect of tool geometry on subsurface damage
and material removal in nanometric cutting single-crystal silicon by a molecular dynamics
simulation. Appl Phys A Mater Sci Process 122:1–16
33. Goel S, Luo X, Reuben RL, Bin Rashid W (2011) Atomistic aspects of ductile responses of
cubic silicon carbide during nanometric cutting. Nanoscale Res Lett 6:589
34. Wang M, Zhu F, Xu Y, Liu S (2018) Investigation of nanocutting characteristics of off-axis
4H-SiC substrate by molecular dynamics. Appl Sci 8:2380
35. Cai MB, Li XP, Rahman M (2007) Study of the mechanism of nanoscale ductile mode cutting
of silicon using molecular dynamics simulation. Int J Mach Tools Manuf 47:75–80
36. Zhang L, Zhao H, Ma Z, Huang H, Shi C, Zhang W (2012) A study on phase transformation
of monocrystalline silicon due to ultra-precision polishing by molecular dynamics simulation.
AIP Adv 2:42116
37. Zhang L, Zhao H, Yang Y, Huang H, Ma Z, Shao M (2014) Evaluation of repeated single-
point diamond turning on the deformation behavior of monocrystalline silicon via molecular
dynamic simulations. Appl Phys A 116:141–150
38. Erhart P, Albe K (2005) Analytical potential for atomistic simulations of silicon, carbon, and
silicon carbide. Phys Rev B 71:35211
39. Brenner DW (1989) Relationship between the embedded-atom method and Tersoff potentials.
Phys Rev Lett 63:1022
40. Albe K, Nordlund K, Nord J, Kuronen A (2002) Modeling of compound semiconductors:
analytical bond-order potential for Ga, As, and GaAs. Phys Rev B 66:35205
41. Goel S, Luo X, Reuben RL (2012) Molecular dynamics simulation model for the quantitative
assessment of tool wear during single point diamond turning of cubic silicon carbide. Comput
Mater Sci 51:402–408
42. Shimada S, Tanaka H, Higuchi M, Yamaguchi T, Honda S, Obata K (2004) Thermo-chemical
wear mechanism of diamond tool in machining of ferrous metals. CIRP Ann—Manuf Technol
53:57–60
43. He T, Liu K, Li XP, Rahman M (2004) A method of atomic dynamic modeling for nanometric
ductile cutting of silicon wafer material. In: International conference on precision engineering,
Singapore, pp 409–417
44. Luo X, Cheng K, Guo X, Holt R (2003) An investigation on the mechanics of nanometric
cutting and the development of its test-bed. Int J Prod Res 41:1449–1465
45. Lawn B (2003) Fracture of brittle solids. Cambridge, New York
Part II
Ductile Mode Cutting Applications
Chapter 6
Ductile Mode Cutting of Silicon
6.1 Introduction
Silicon wafer is a fundamental material being widely used for semiconductor device
manufacturing due to its unique electronic and mechanical properties. It is fabricated
from the single crystal silicon ingot through several fabrication processes, including
slicing, edge grinding, finishing, lapping and polishing. After integrity circuit (IC)
fabrication, back thinning and dicing of wafers into dies are the subsequent fabrication
processes. Currently in wafer fabrication, those machining processes such as wafer
slicing, edge grinding, finishing, lapping, dicing and back thinning, are all based on
grinding or/and abrasive process, which easily generate micro cracks and subsurface
damage [1–4]. The average wafer subsurface damage layer thickness is more than
6 µm even under the fine grinding process [1]. It is obviously too costly and time-
consuming to remove the subsurface damaged layer by heave chemical-mechanical
polishing (CMP) process. Therefore, ductile mode cutting of silicon wafers with free
and/or less subsurface damage is urgently needed for the semiconductor industry.
Smooth surfaces and continuous chips could be achieved when the undeformed
chip thickness was less than 58 nm in cutting of silicon using a single crystal diamond
(SCD) tool with an extremely small cutting edge radius (50 nm) [5]. Transmission
electron microscopy (TEM) observation of nanomachined silicon crystals indicated
that the surface damage achieved in diamond turning appeared more homogenous
than that in precision grinding [6]. The nanometric cutting of silicon wafers showed
that a brittle-ductile transition occurred when the undeformed chip thickness was
increased [7]. The cutting tool edge geometry has a significant effect on the nano-
metric ductile cutting of semiconductor materials [7, 8]. The same brittle-ductile
transition of chip formation was observed in cutting of silicon and single-crystal
germanium wafers [9, 10]. Ultra-precision ductile mode cutting of soda-lime glass
was firstly conducted using a single crystal diamond tool in 1992 [11]. It was found
that in both the conventional cutting and ultrasonic vibration assisted cutting, the
machined grooves were formed from ductile mode to brittle mode when the depth of
cut exceeded a critical value and ultrasonic vibration assisted cutting would improve
the ductile cutting performance of glass. The profile of the tool edge was perfectly
transferred to the cutting groove formed by the ultrasonic vibration cutting compared
with the conventional cutting. It was reported that the crystallographic orientation
dependence of the surface features changed with the depth of cut due to the differ-
ence in material removal mechanism between plastic deformation and brittle fracture
[12]. The ductile regime machining was achieved when the machining of silicon was
along the ‹110› directions and the maximum chip thickness was less than 0.5 µm
[13]. Molecular dynamics simulation of the nanometric cutting of silicon indicated
that the material removal mechanism was a typical pressure-induced phase trans-
formation of silicon material [14]. These results suggest that silicon wafers could
be machined in ductile mode under certain cutting conditions. Diamond tool wear
behaviour during cutting of single crystal silicon was reported that the predominant
tool wear mechanism in terms of trailing micro-grooves depended on undeformed
chip thickness [15]. The other one is the generation of nano or micro-grooves at
the tool flank, which forms sub-cutting edges of much smaller radii on the main
cutting edge [16]. Experimental study on diamond cutting of silicon wafer demon-
strated that silicon carbide and diamond-like carbon particles were generated during
cutting, which caused scratch and plough on the tool flank face so as to form the
groove marks [17]. Crack initiation in machining monocrystalline silicon was sim-
ulated using renormalization group molecular dynamics which indicated the crack
propagation occurred after depth of cut being larger than 1 µm [18]. Nanometric
cutting of silicon wafer experimental results indicated that there is an upper bound
value of diamond tool edge radius, above which the chip formation mode is changed
from ductile to brittle even through the undeformed chip thickness remains to be
smaller than the tool edge radius, which was found to be between 700 and 800 nm
[19].
Although many interesting studies have been conducted on diamond cutting of
silicon wafer, the most work is focused on diamond tool wear behaviour, diamond tool
edge radius effect on ductile mode cutting, ductile-to-brittle transition, simulation on
crack initiation, as well as other semiconductor materials. In this chapter, machined
work surface characteristics of silicon wafer in ductile mode cutting are studied
comprehensively. Theoretical analysis indicates that in ductile mode cutting of silicon
wafers, machined workpiece surface is free of cracks and its surface roughness is
in nanometer scale. Nanometric ductile cutting experiments of silicon wafers using
single crystal diamond tools are carried out on an ultra-precision machining lathe to
verify the theoretical prediction.
from the tool arc cutting edge is penetrated into workpiece material and its value is
gradually decreased by one atom each time until the stress is reduced to zero. Thus,
there is a compressive stressed zone in the chip formation zone as shown in Fig. 5.13,
in which the compressive stress reaches the maximum value at the tool-workpiece
interface and decreases to zero at the boundary of the zone. The stress decreasing
gradient depends on the work material properties. Within the compressive stressed
zone, crack propagation can be blocked when the compressive stress is large enough,
because the stress intensity factor K I at the crack tip is reduced by the large com-
pressive stress [21]. Therefore, there is a shielding zone for crack propagation within
the compressive stressed zone, where crack propagation is screened by the large
compressive stress. At the boundary of the shielding zone, the dislocation emission
and crack propagation are in an equilibrium condition. Actually, in cutting of brittle
materials, dislocation emission and crack propagation are coexisted within the com-
pressive stressed zone. But within the shielding zone, the chip formation process is
dominated by dislocation rather than crack propagation. Therefore, in ductile cut-
ting the chip formation is a dislocation process, and the surface generated is free of
fracture or cracks.
According to cutting geometry as shown in Fig. 2.3, resultant cutting force F r can
be expressed by Eq. (2.16). In cutting of brittle materials, such as silicon, ductile mode
chip formation may be achieved when the undeformed chip thickness is smaller than
the tool edge radius. Thus, for ductile mode cutting of brittle material, the resultant
cutting force F r is given by [19]:
where rc = ac /a0 is cutting ratio, and k AB is shear flow stress along the curved shear
surface AB.
At the tool-workpiece interface, the mean normal compressive stress acting on
the arc cutting edge σ c can be derived as:
Fn Fr cos β
σc = = (6.2)
Ac Ac
where Ac = f (r, ac , R, γ ) is contact area between the cutting tool and silicon wafer
workpiece. For a given single crystal diamond cutter, the tool nose radius R and rake
angle γ are constant, with the ratio of maximum undeformed chip thickness d max to
tool edge radius r fixed as a constant value of less than 1, the tool-workpiece contact
area Ac is a function of the tool edge radius r, i.e.,
Ac = f (r ) (6.3)
106 6 Ductile Mode Cutting of Silicon
σc
0 R
0.0321 f 2
Ra = (6.4)
R
which shows that the ideal surface roughness is mainly determined by cutting tool
geometry and feed rate used. Therefore, according to Eq. (6.4), in nanometric cutting
of silicon wafer with a micro scale feed rate and a millimeter scale tool nose radius,
the surface roughness must be in a nanometer scale.
As analyzed in Chap. 2, the key condition for the ductile mode chip formation in
cutting of brittle material is the extremely large compressive stress, which is so large
that crack propagation of material’s pre-existing defects is screened and the chip
formation is dominated by dislocation emission. Therefore, since the compressive
stress decreases with the tool cutting edge radius, as shown in Fig. 6.1, for ductile
mode chip formation, the tool edge radius should be the smaller the better, and there
must be an upper bound of the tool edge radius for each workpiece material, above
which the cutting is not in ductile mode.
6.3 Experimental Details 107
Silicon has 14 atoms arranged into a diamond structure. Silicon atom diameter is
1.175 Å, silicon interatomic distance is 2.3 Å, and lattice content is 5.4307 Å. The
two most common orientations used for silicon wafers are the ‹100› and the ‹111›
planes. These two orientations satisfy almost all requirements. A ‹100› wafer will
break into rectangular pieces if it is shattered. The physical properties of silicon is
listed in Table 6.1 [23].
Single point diamond turning (SPDT) of silicon wafers is carried out on an ultra-
precision lathe with 10 nm positioning resolution using a single crystal diamond
tool. Figure 6.2 displays the ultra-precision CNC lathe (Toshiba ULG-100) used in
the experiments and Fig. 6.3 displays the experimental setup of SPDT. Silicon (111)
wafers with 76.2 mm in diameter, 0.5 mm in thickness and having a lapped finish
are used as the workpiece. Silicon wafers are bonded on aluminum blanks using a
heat-softened glue and then vacuum chucked on the machine spindle. As the layer
of glue may not be evenly spread out, pre-trimming is performed before the start
of the experiments to ensure that the silicon wafer surface is extremely flat. After
pre-trimming and before the experiments, the machined wafer surface is examined
using a surface profiler to ensure its flatness. Its flatness is achieved about 0.25 µm
after trimming, where the measurement is done on the silicon wafer bounded on the
20 mm thick aluminum blank.
Tool geometry of the single crystal diamond cutter is listed in Table 6.2 [21]. The
tool edge radius of diamond cutters is measured through an indentation method
using an atomic force microscope (AFM) [24]. The initial tool edge radius of the
fresh diamond cutter is 23 nm. In order to obtain different tool edge radius of diamond
cutters, a special lapping process is designed as shown in Fig. 6.4 [19]. Diamond
cutters are lapped using a diamond slurry attached to a grooved aluminum alloy bar
to round the cutting edge, of which the Al bar is held by a CNC lathe chuck and
rotating around its central axis. Different tool edge radii of diamond cutters can be
achieved by controlling the lapping time. And late SPDT is conducted using the
pre-prepared single crystal diamond cutters with different cutting edge radii.
Nanometric cutting is conducted using different tool edge radii (sharpness) firstly to
investigate the critical conditions and tool edge radius effect on ductile mode cutting
of single crystalline silicon wafer material. Three single crystal diamond cutters with
Diamond Slurry
Diamond Tool
110 6 Ductile Mode Cutting of Silicon
different tool edge radii are pre-prepared using the special designed lapping method
as shown in Fig. 6.4. Cutting conditions for the tests of tool cutting edge radius
effect on ductile mode cutting of single crystalline silicon wafer material are listed
in Table 6.3 [7]. Maximum undeformed chip thickness d max can be calculated from
Eqs. (4.22) and (4.23). Here, the values of undeformed chip thickness achieved vary
in the range of 10 nm to 66 nm for diamond cutter A, in the range of 118–337 nm
for diamond cutter B, and in the range of 50–690 nm for diamond cutter C.
Then, nanometric cutting is conducted using different tool edge radii (sharpness)
to investigate the upper bound of tool edge radius on ductile mode cutting of single
crystalline silicon wafer material. Six single crystal diamond cutters with different
tool edge radii are pre-prepared using the special designed lapping method as shown
in Fig. 6.4. The cutting conditions for tool cutting edge radius upper band tests
are listed in Table 6.4 [19]. Dry cutting is carried out for collecting chips easily.
Chips are collected using a double-sided tape, which is set onto the rake face of
Table 6.3 Nanometric cutting conditions for tool edge radius effect tests
Diamond tool Cutting condition
Spindle speed v Feed rate f DOC ao (nm) UCT d max (nm)
(rpm) (µm/rev)
Cutter A (r = 1000 5 10 10
45 nm) 20 20
30 30
40 40
50 50
70 66
Cutter B (r = 1000 5 900 118
335 nm) 2000 178
5500 298
5800 307
6000 320
6500 325
7000 337
Cutter C (r = 1000 5 70 50
647 nm) 210 100
740 200
2770 400
5000 543
6000 596
7200 655
8000 690
6.3 Experimental Details 111
Table 6.4 Nanometric cutting conditions for cutting tool edge radius upper band tests
No. Tool edge Cutting speed Feed rate Depth of cut UCT d max
radius r (nm) v (m/min) n (rpm) f (µm/rev) ao (µm) (nm)
Table 6.5 Nanometric cutting conditions for cutting performance evaluation tests
No. Spindle speed v (rpm) Feed rate f (µm/rev) DOC ao (nm) UCT d max (nm)
1 750 10 100 65
2 500 166
3 1000 242
4 1700 320
5 2000 348
6 2400 383
the diamond cutter. Lastly, nanometric cutting is conducted using a diamond cutter
with the tool edge radii (sharpness) of 335 nm to evaluate the ductile mode cutting
performance of single crystalline silicon wafer material. The cutting conditions for
cutting performance evaluation are listed in Table 6.5 [25]. Air and oil mist is used
as coolant for the tests.
The machined silicon wafer surfaces and chip formation are examined using a
SEM, TEM, OMIS, stylus surface profiler and AFM. The surface roughness Ra of
the machined silicon wafer is measured using the stylus surface profiler. Six points
on the machined wafer surface are selected for its surface roughness measurement
and the average value is taken from the six measurements.
112 6 Ductile Mode Cutting of Silicon
First SPDT test is to investigate the effect of tool edge radius on ductile mode cutting
of silicon wafers using three diamond cutters with different tool edge radii. SEM
micrographs of chips obtained in cutting of silicon wafers using diamond cutter A
with the tool edge radius of 45 nm under the cutting speed of 4 m/s (1000 rpm)
and feed rate of 5µm/rev (5 mm/min) are shown in Fig. 6.5, where the maximum
undeformed chip thickness are: (a) 10 nm, (b) 20 nm, (c) 30 nm, (d) 40 nm, (e)
50 nm and (f) 66 nm, respectively [7]. It can be seen that when the undeformed
chip thickness is smaller than 40 nm (Fig. 6.5a–d), the chips obtained are in the
form of ribbons and layers, i.e. the chips are formed continuously. The appearance
of such chips is similar to the chip generated when machining ductile metals. Hence,
such continuous chips obtained from the cutting of silicon at these undeformed chip
thickness indicate that the cutting is carried out in ductile mode cutting and plastic
deformation has taken place. On the other hand, as the undeformed chip thickness is
increased beyond 40 nm, discontinuous and fractured chips are obtained. It can be
seen that the chips generated are blocky and irregular in shape (Fig. 6.5e, f). They
also have sharp ends and such fragments are a result of crack propagation in fracture
process. Hence, the cutting under undeformed chip thickness being larger than 40 nm
indicates that the chip formation is in brittle mode cutting.
SEM micrographs of chips obtained in cutting of silicon wafers using the diamond
cutter B with the cutting edge radius of 335 nm under the cutting speed of 4 m/s
(1000 rpm) and feed rate of 5 m/rev (5 mm/min) are shown in Fig. 6.6, where the
undeformed chip thickness are: (a) 118 nm, (b) 178 nm, (c) 298 nm, (d) 307 nm,
(e) 320 nm, (f) 325 nm and (g) 337 nm, respectively [7]. It is found that when the
undeformed chip thickness is smaller than 320 nm, the chips obtained are in smooth
layers (Fig. 6.6a–e). As mentioned earlier, such continuous chips obtained are similar
to the chip formation during the cutting of ductile materials, where chip formation is
dominated by dislocation. Therefore, these chips show that the cutting is carried out in
ductile mode cutting under the cutting conditions. As the undeformed chip thickness
is increased further, irregular and blocky chips are once again obtained (Fig. 6.6f, g).
These chips are formed in a disorderly manner. Under the above conditions, brittle
fracture dominates the chip formation such that the cutting is performed under brittle
mode cutting.
Cutting of silicon wafers is made using diamond cutter C with the cutting edge
radius of 647 nm under the cutting speed of 4 m/s (1000 rpm) and feed rate of 5µm/rev
(5 mm/min). SEM micrographs of chips obtained are shown in Fig. 6.7, where the
undeformed chip thickness are: (a) 50 nm, (b) 100 nm, (c) 200 nm, (d) 400 nm, (e)
543 nm, (f) 596 nm, (g) 655 nm and (h) 690 nm, respectively [7]. The micrographs
show that the chips formed in cutting with the undeformed chip thickness being
between 50 nm to 596 nm are in the form of long ribbon, which is similar to the
chip generated in machining of ductile metals as shown in Fig. 6.7a–f. The chips
are continuous with fine ripples on the surfaces, indicating that the chip formation is
under ductile mode cutting dominated by dislocation. On the other hand, the chips in
6.4 Tool Edge Radius Effect 113
(a) Undeformed chip thickness: 10nm (b) Undeformed chip thickness: 20nm
(c) Undeformed chip thickness: 30nm (d) Undeformed chip thickness: 40nm
(e) Undeformed chip thickness: 50nm (f) Undeformed chip thickness: 66nm
Fig. 6.5 Chips formed at different undeformed chip thickness in cutting of Si-wafer using diamond
cutter A with a cutting edge radius of 45 nm
114 6 Ductile Mode Cutting of Silicon
(a) Undeformed chip thickness: 118nm (b) Undeformed chip thickness: 178nm
(c) Undeformed chip thickness: 298nm (d) Undeformed chip thickness: 307nm
(e) Undeformed chip thickness: 320nm (f) Undeformed chip thickness: 325nm
Fig. 6.6 Chips formed at different undeformed chip thickness in cutting of Si-wafer using diamond
cutter B with a cutting edge radius of 335 nm
6.4 Tool Edge Radius Effect 115
(a) Undeformed chip thickness: 50nm (b) Undeformed chip thickness: 100nm
(c) Undeformed chip thickness: 200nm (d) Undeformed chip thickness: 400nm
(e) Undeformed chip thickness: 543nm (f) Undeformed chip thickness: 596nm
(g) Undeformed chip thickness: 655nm (h) Undeformed chip thickness: 690nm
Fig. 6.7 Chips formed at different undeformed chip thickness in cutting of Si-wafer using the
diamond cutter C with the cutting edge radius of 647 nm
116 6 Ductile Mode Cutting of Silicon
cutting with the undeformed chip thickness being larger than 655 nm are fragmented
and discontinuous with non-uniform length as shown in Fig. 6.7g, h. These chips
are fractured particles and blocks with sharp ends. Such appearance indicates that in
the cutting process, the chip formation is dominated by crack propagation and brittle
fracture.
In summary, all the experimental results show that in cutting of silicon wafers
using diamond tools of different cutting edge radii, there is a transition from ductile
mode cutting to brittle mode cutting in the chip formation once the undeformed chip
thickness being increased to a certain level. However, comparing the results from
cutting with the three different tool edge radii, as shown in Figs. 6.5, 6.6 and 6.7, it
can be seen that the critical undeformed chip thickness corresponding to the ductile-
to-brittle transition of chip formation varies with the tool cutting edge radius. The
larger the tool edge radius, the larger critical value of undeformed chip thickness.
Also, it should be noted that the values of the critical undeformed chip thickness are
close to the values of the cutting edge radius. That is, the critical undeformed chip
thicknesses are 40, 320 and 596 nm when the tool cutting edge radii are 45, 335 and
647 nm, respectively.
The results from the cutting tests indicate that in cutting of silicon wafers under
certain cutting conditions, there is a critical value for the undeformed chip thickness,
at or below which the chip formation is under ductile mode cutting which generates
continuous chips. This observation is similar for all three cases of the tools with
different edge radii. Although there is a critical undeformed chip thickness for all
the three tools, this critical value differs. From the above Figs. 6.5, 6.6 and 6.7, it
can be seen that the critical value of the maximum undeformed chip thickness at the
ductile-to-brittle transition increases when the cutting edge radius of the diamond
cutter increasing. Meanwhile, the critical value of maximum undeformed chip thick-
ness at the ductile-to-brittle transition in cutting of silicon wafer coincides with the
tool cutting edge radius: the critical value of the maximum undeformed chip thick-
ness is 40 nm when the cutting edge radius of diamond cutter A is 45 nm in cutting
of silicon wafers, while the critical value is 320 nm when the cutting edge radius of
diamond cutter B is 335 nm, and the critical value is 596 nm when the cutting edge
radius of diamond cutter C is 647 nm. Figure 6.8 shows the relationship as the critical
undeformed chip thickness increases linearly with the tool cutting edge radius [7].
It is implied that ductile chip formation is a result of large compressive stresses
in the chip formation zone, which acts to stop the growth of pre-existing flaws in the
material by suppressing the stress intensity factor K I . This large compressive stress
can be generated by having an extremely small undeformed chip thickness and the
tool cutting edge radius being larger than the undeformed chip thickness. In addition,
ductile chip formation in cutting of brittle material can be a result of the enhancement
of material yield strength in the chip formation zone, which directly enhances the
material fracture toughness K C . This can be achieved by dislocation hardening and
strain gradient by having a nanometric scale of undeformed chip thickness during
the cutting process.
6.5 Tool Edge Radius Upper Bound 117
Thickness (nm)
400
300
200
Present Result
100
Yan et al Result [5]
0
0 100 200 300 400 500 600 700
Tool Cutting Edge Radius (nm)
Second SPDT test is to investigate the upper bound of tool edge radius on ductile
mode cutting of silicon wafers using six diamond cutters with different tool edge
radii (sharpness).
SEM and AFM photographs of the machined silicon wafer surfaces achieved in
cutting of silicon wafers using diamond cutters with different cutting edge radii at
the cutting speed of 150 m/min are shown in Figs. 6.9 and 6.10, respectively [19]. For
six diamond cutters with the tool edge radius of 23, 202, 490, 623 and 717 nm, one
test is conducted under the condition that undeformed chip thickness is less than the
tool edge radius and another test is conducted under the condition that undeformed
chip thickness is larger than the tool edge radius. For the diamond tool edge radius of
807 nm, both tests are conducted under the condition that undeformed chip thickness
is less than the tool edge radius.
Both SEM and AFM observations indicate that when the tool cutting edge radius
is not larger than 807 nm and undeformed chip thickness is smaller than the tool
cutting edge radius, the machined workpiece surfaces are very smooth and feed
marks are clearly displayed on the surfaces as shown in Figs. 6.9 and 6.10, which
show that the cutting is carried out under ductile mode cutting. On the other hand,
when the undeformed chip thickness is larger than the tool cutting edge radius, the
machined workpiece surfaces are very rough and fractured, which shows that the
cutting is conducted under brittle mode cutting. However, when the tool cutting edge
radius is 807 nm and even undeformed chip thickness is much smaller than the tool
cutting edge radius, the machined workpiece surfaces are very rough and fractured,
which show that the cutting is carried out under brittle mode cutting with the cutting
118 6 Ductile Mode Cutting of Silicon
r = 23nm
r = 202nm
r = 490nm
r = 623nm
r = 717nm
Fig. 6.9 SEM photographs of machined silicon wafer surfaces under different undeformed chip
thickness with different tool edge radius
6.5 Tool Edge Radius Upper Bound 119
r = 807nm
conditions. It is likely that the tool cutting edge radius of 807 nm should be an upper
bound for ductile mode cutting of silicon wafer material. As a result, in order to
achieve ductile mode cutting of silicon wafer material, two conditions should be
satisfied: (1) the diamond tool edge radius must be smaller than 807 nm and (2) the
undeformed chip thickness must be smaller than the tool cutting edge radius.
SEM photographs of chips formed in cutting of silicon wafers using diamond cutters
with different tool edge radii under the cutting speed of 150 m/min are shown in
Fig. 6.11 [19]. The cutting tests are conducted under the conditions mentioned in
Sect. 6.3.4. It is found that when the tool cutting edge radius is not larger than 807 nm
and undeformed chip thickness is smaller than the tool cutting edge radius, the chips
obtained seem to be formed in ductile mode cutting, as shown in Fig. 6.11. As men-
tioned earlier, such continuous chips obtained are similar to the chip formation during
the cutting of ductile materials, where chip formation is dominated by dislocation.
On the other hand, when the undeformed chip thickness is larger than the tool cutting
edge radius, the chips obtained seem to be formed in irregular shapes, showing that
the chips likely formed in brittle mode cutting. However, when the tool cutting edge
radius is 807 nm and even the undeformed chip thickness is much smaller than the
tool cutting edge radius, the chips obtained show that the cutting is carried out under
brittle mode cutting with the cutting conditions. It is believed that the tool cutting
edge radius of 807 nm should be an upper bound for achieving ductile mode cutting
of silicon wafer material, which is also verified by the machined workpiece surface
topography as shown in Figs. 6.9 and 6.10. Therefore, two conditions should be
satisfied for obtained ductile mode cutting of silicon wafer material: (1) the diamond
tool edge radius must be smaller than 807 nm and (2) the undeformed chip thickness
must be smaller than the tool cutting edge radius.
120 6 Ductile Mode Cutting of Silicon
r = 23nm
r = 202nm
r = 490nm
r = 623nm
Fig. 6.10 AFM photographs of machined silicon wafer surfaces under different undeformed chip
thickness with different tool edge radius
6.5 Tool Edge Radius Upper Bound 121
r = 717nm
r = 807nm
r = 23nm
r = 202nm
r = 490nm
r = 623nm
r = 717nm
Fig. 6.11 SEM photographs of chips formed in cutting of silicon wafers under different undeformed
chip thickness with different tool edge radius
6.6 Ductile Mode Cutting Performance 123
r = 807nm
Last SPDT test is conducted to evaluate the cutting performance in ductile mode
cutting of silicon wafers using the diamond cutter with tool edge radius of 335 nm.
Fig. 6.12 Surface roughness of machined silicon wafers achieved under different cutting modes
with different cutting edge radius
6.6 Ductile Mode Cutting Performance 125
SEM examination on the cross-section of the machined silicon wafer achieved under
ductile mode cutting is shown in Fig. 6.16, where the maximum undeformed chip
thickness is 320 nm [20]. No micro cracks and subsurface damages are observed,
which indicates that in nanometric ductile mode cutting of silicon wafers, the
machined workpiece surface is free of fracture and subsurface damages. It seems
that ductile mode cutting of silicon wafers is a fracture-free machining process for
the wafer preparation.
TEM pictures of chips generated in cutting of silicon wafer are shown in Fig. 6.17,
where the cutting conditions are: spindle speed v of 1000 rpm, feed rate f of
10 µm/rev, depth of cut ao of 2 µm and using the diamond cutter with tool edge
radius r of 647 nm [20]. It is found that the chips formed are thin ribbons, of which
the chip thickness is in nanometer scale. It well substantiates that the cutting is carried
out under ductile mode cutting.
126 6 Ductile Mode Cutting of Silicon
Fig. 6.14 SEM photographs of the machined silicon wafer surfaces with different undeformed chip
thickness (d max )
6.6 Ductile Mode Cutting Performance 127
Fig. 6.15 AFM photographs of the machined silicon wafer surfaces with different undeformed
chip thickness (d max )
128 6 Ductile Mode Cutting of Silicon
Fig. 6.16 Cross-section of the machined silicon wafer achieved in ductile mode cutting
Fig. 6.17 TEM pictures of chips generated under ductile mode cutting of silicon wafer
Tool wear test is conducted using a diamond cutter having the crystallographic plane
{110} as rake face. Figure 6.18 are SEM photographs of the cutting edges of the
diamond cutter: (a) fresh cutting edge, (b) after cutting of 2.15 km, (c) after cutting
of 3.36 km and (d) after cutting of 10 km [26, 27]. The fresh diamond cutter as shown
in Fig. 6.18a appears to be smooth and sharp revealing no visible defects under high
magnification. After cutting of 2.15 km as shown in Fig. 6.18b, a very small-scale
wear on the flank face and microscopic ruggedness on the tool edge are observed.
While the middle portion of the tool cutting edge shows greater wear. After cutting
of 3.36 km as shown in Fig. 6.18c, tool wear exhibits progressive wear on flank face.
Larger micro-grooves lie approximately at the mid-point of the tool cutting edge. As
6.6 Ductile Mode Cutting Performance 129
shown in Fig. 6.18d after cutting of 10 km, larger progressive wear is observed to
dominate the tool wear pattern with no obviously micro-groove wear and notch wear
on flank face.
Figure 6.19 illustrates the behavior of tool flank wear with respect to cutting
distance for the diamond tool having {110} crystallographic orientation as the rake
face [26, 27]. As shown in Fig. 6.19, tool wear initially tends to be slow, and then
increases significantly after cutting of around 3.0 km. Compared to wear height on
0.1
0.05
0
0 5 10 15
Cutting distance (km)
130 6 Ductile Mode Cutting of Silicon
0.02
0.01
0
0 5 10 15
Cutting distance (km)
flank face, wear width gets more dominant when increasing the cutting distance.
Even after cutting of 10 km, the maximum wear height on flank face is estimated
to be 14 µm, which is not expected to severely affect the cutting performance.
Here occurrence of progressive flank wear is basically caused by abrasion wear and
adhesion wear between cutting tool and machined surface, and possible thermo-
chemical effect. Furthermore, with the cutting distance increasing, tool wear grows
at the tool tip becoming larger both in height and width because the contact area
between the diamond cutter and work material increasing, thus forming a greater
wear land on flank face.
Figure 6.20 shows the tool wear behaviors of three diamond cutters with different
crystalline orientations as rake face in nanometric ductile mode cutting of silicon
wafers under the same cutting conditions. Although wear patterns as well as wear
mechanisms for three diamond cutters are same, the diamond cutter with {110} crys-
talline orientation as rake face shows greater wear resistance than those with {100}
and {111} orientations as rake face. Interestingly, diamond cutters with {100} and
{111} crystalline orientations as rake face exhibit very similar wear characteristics
with respect to cutting distance. Furthermore, the cutting directions, namely harder
and softer directions on a crystal plane, affect the tool wear rate. Thus the harder
direction ‹110› on flank face with {100} crystalline orientation may possibly be
responsible for less wear and longer tool life for diamond cutter with {110} crys-
talline orientation as rake face. Also the higher wear rate for diamond cutters with
{100} and {111} crystalline orientation as rake face may correspond to the softer
directions ‹100› and ‹112› on flank face with {100} and {112} crystalline orientation,
respectively.
Figure 6.21 shows the effect of depth of cut on surface roughness Ra of the machined
silicon wafers, which indicates that a good surface integrity is achieved in ductile
mode cutting when the depth of cut is less than 1700 nm, that is, the maximum
6.6 Ductile Mode Cutting Performance 131
300
200
100
0
0 500 1000 1500 2000 2500 3000
Depth of Cut (nm)
undeformed chip thickness is less than 320 nm [20]. The machined workpiece surface
roughness Ra achieved is less than 80 nm in nanometric ductile mode cutting of silicon
wafer.
According to Eq. (6.4), for the nanometric ductile cutting of silicon wafer, the
ideal surface roughness Ra of the machined silicon wafers can be calculated, which is
about 1.07 nm. Comparing the two surface roughness values, it is found that the actual
measured surface roughness Ra is much larger than the ideal value. Normally, it can
be attributed to chatter, vibration and inaccuracies of the machine tool movement.
But it is definitely different for the nanometric ductile mode cutting process. The
ultra-precision machine center has an excellent stiffness and movement accuracy.
Meanwhile, the depth of cut used is in micron scale so as to generate sub-Newton
scale cutting force, which is not power enough to activate any vibration during cutting
process. Therefore, the main reason causing the large machined surface roughness is
the diamond tool cutting edge contour accuracy, which is not perfect viewing under
a high magnification microscope even for a fresh one, and normally has a submicron
contour tolerance. Meanwhile, tool wear is occurred during the cutting process as
shown in Fig. 6.18, such that its cutting edge profile would become much rougher
than the fresh tool cutting edge. The tool cutting edge profile is duplicated on the
machined workpiece surface during cutting. The surface roughness of the machined
workpiece is mainly determined by the tool cutting edge contour rather than the tool
geometry and cutting conditions used. Therefore, the measured value of machined
workpiece surface roughness Ra is much larger than the ideal surface roughness
value.
The machined silicon wafer surface achieved in ductile mode cutting with the
maximum undeformed chip thickness of 320 nm is examined using an OMIS, as well
as the original polishing surface of a silicon wafer. These digital photos are shown in
Fig. 6.22 [20, 25]. Comparing the machined silicon wafer surface with the polished
wafer surface, it can be said that a good workpiece surface for silicon wafers can be
achieved by using the ductile mode cutting technology. Even through the machined
surface quality is far away from the final requirement of silicon wafers such as
chemical-mechanical polishing (CMP) silicon wafer surface, the machined silicon
132 6 Ductile Mode Cutting of Silicon
Fig. 6.22 Digital photos of ductile mode cutting and original polishing surfaces of silicon wafers
wafer surface with free of micro cracks and subsurface damage can be achieved
by ductile mode cutting, definitely which would significantly reduce the subsequent
CMP duration.
In general, the experimental results from the nanometric cutting tests indicate that
in cutting of silicon wafers, there is a critical undeformed chip thickness, at or below
which chip formation is under ductile mode cutting which generates continuous
chips. And the critical undeformed chip thickness differs when cutting of silicon
wafers using diamond cutters with different tool edge radii. The bigger diamond tool
edge radius, the larger critical undeformed chip thickness. It is implied that ductile
chip formation is a result of large compressive stresses in the chip formation zone,
which acts to stop the growth of pre-existing flaws in the material by suppressing the
stress intensity factor K I . This large compressive stress can be generated by having
an extremely small undeformed chip thickness and the undeformed chip thickness
being smaller than the tool cutting edge radius. In addition, ductile chip formation
in cutting of brittle material can be a result of the enhancement of material yield
strength in the chip formation zone, such that brittle material can undertake a much
large cutting stress in chip formation zone without fracture. This can be achieved by
dislocation hardening and strain gradient by having a nanometric scale of undeformed
chip thickness during cutting [21]. But there is an upper bound for diamond tool
edge radius, above which the chip formation is changed from ductile mode cutting
to brittle mode cutting even though the undeformed chip thickness remains to be
smaller than the tool edge radius. And the machined workpiece surface is fractured,
causing significant increase in the machined surface roughness. This could be due
to the decrease of the compressive stress in the chip formation zone, as discussed in
Sect. 6.2.
6.7 Concluding Remarks 133
Theoretical analysis on ductile mode cutting of silicon wafer indicates that machined
silicon surface with free of fracture and nanometer scale surface roughness is achieved
when dislocation dominates its chip formation rather than crack propagation. Nano-
metric cutting of silicon wafers using an ultra-precision CNC lathe with single crystal
diamond cutters are conducted to investigate tool edge radius effect on critical unde-
formed chip thickness and verify ductile mode cutting performance of silicon wafer.
Experimental results from the nanometric cutting tests indicate that in cutting of
silicon wafers, there is a critical undeformed chip thickness, at or below which chip
formation is under ductile mode cutting which generates continuous chips. Both duc-
tile mode and brittle mode of chip formation processes are observed in the cutting
tests. Critical undeformed chip thickness for ductile-to-brittle transition of chip for-
mation in cutting of silicon wafers varies with tool cutting edge radius. Larger tool
cutting edge radius, larger critical undeformed chip thickness for ductile-to-brittle
transition in chip formation. But there also has an upper bound for diamond tool edge
radius, above which chip formation is changed from ductile mode cutting to brittle
mode cutting even though the undeformed chip thickness remains to be smaller than
tool edge radius.
To obtain ductile mode cutting of single crystalline silicon wafers achieving
fracture-free surfaces, three basic conditions must be satisfied:
• Having an extremely small undeformed chip thickness, said in nanometric scale,
for cutting of silicon wafer.
• Undeformed chip thickness has to be smaller than tool edge radius of diamond
cutters.
• Tool edge radius of diamond cutters has to be smaller than the upper bound, being
between 700 and 800 nm for cutting of silicon wafers.
Surface roughness Ra obtained in ductile mode cutting of silicon wafers being
much larger than ideal surface roughness is mainly attributed by imperfect tool
cutting edge contour and tool wear. Experimental results are well substantiate the
analytical findings and nanometric ductile mode cutting of silicon wafer is achieved
successfully under the certain cutting conditions.
References
1. Pei ZJ, Billingsley SR, Miura S (1999) Grinding induced subsurface cracks in silicon wafers.
Int J Mach Tools Manufact 39:1103–1116
2. Zarudi I, Zhang L (1996) Subsurface damage in single-crystal silicon due to grinding and
polishing. J Mater Sci Lett 15:586–587
3. Brinksmeier E, von Schmieden W (1987) ID-cut-off grinding of brittle materials. CIRP Ann
36:219–222
4. Enomoto T, Shimazaki Y, Tani Y et al (1999) Development of a resinoid diamond wire con-
taining metal powder for slicing a slicing ingot. CIRP Ann 48:273–276
134 6 Ductile Mode Cutting of Silicon
5. Yan J, Yoshino M, Kuriagawa T et al (2001) On the ductile machining of silicon for micro
electro-mechanical system (MEMS), opto-electronic and optical applications. Mater Sci Eng
A 297:230–234
6. Puttick KE, Whitmore LC, Chao CL et al (1994) Transmission electron microscopy of nanoma-
chined silicon crystals. Phil Mag A 69:91–103
7. Liu K, Li XP, Rahman M et al (2007) A study of the effect of tool cutting edge radius on ductile
cutting of silicon wafers. Int J Adv Manufact Technol 32:631–637
8. Lucca DA, Chou P, Hocken RJ (1998) Effect of tool edge geometry on the nanometric cutting
of Ge. CIRP Ann 47:475–478
9. Blackley WS, Scattergood RO (1994) Chip topography for ductile-regime machining of ger-
manium. ASME T J Eng Ind 116:263–266
10. Fang FZ, Venkatesh VC (1998) Diamond cutting of silicon with nanometric finish. CIRP Ann
47:45–49
11. Moriwaki T, Shamoto E, Inoue K (1992) Ultraprecision ductile cutting of glass by applying
ultrasonic vibration. CIRP Ann 41:141–144
12. Shibata T, Fujii S, Makino E et al (1996) Ductile-regime turning mechanism of single-crystal
silicon. Precis Eng 18:129–137
13. Hung NP, Fu YQ (2000) Effect of crystalline orientation in the ductile-regime machining of
silicon. Int J Adv Manufact Technol 16:871–876
14. Komanduri R, Chandrasekaran N, Raff LM (2001) Molecular dynamics simulation of the
nanometric cutting of silicon. Philos Mag B 81:1989–2019
15. Yan JW, Syoji K, Tamaki J (2003) Some observations on the wear of diamond tools in ultra-
precision cutting of single-crystal silicon. Wear 255:1380–1387
16. Li XP, He T, Rahman M (2007) Tool wear characteristics and their effects on nanoscale ductile
mode cutting of silicon wafer. Wear 25:1207–1214
17. Zong WJ, Sun T, Li D et al (2008) XPS analysis of the groove wearing marks on flank face of
diamond tool in nanometric cutting of silicon wafer. Int J Mach Tools Manufact 48:1678–1687
18. Inamura T, Shimada S, Takezawa N et al (1999) Crack initiation in machining monocrystalline
silicon. CIRP Ann 48:81–84
19. Arefin S, Li XP, Liu K (2007) The upper bound of tool edge radius for nanoscale ductile cutting
of silicon wafer. Int J Adv Manufact Technol 31:655–662
20. Liu K, Zuo DW, Li XP et al (2009) Nanometric ductile cutting characteristics of silicon wafer
using single crystal diamond tools. J Vac Sci Technol B 27:1361–1366
21. Liu K, Li XP, Liang SY (2007) The mechanism of ductile chip formation in cutting of brittle
materials. Int J Adv Manufact Technol 33:875–884
22. Boothroyd G, Knight WA (1989) Fundamentals of machining and machine tools. Marcel
Dekker, New York
23. https://en.wikipedia.org/wiki/Silicon
24. Li XP, Rahman M, Liu K et al (2003) Nanoprecision measurement of diamond tool edge radius
for wafer fabrication. J Mater Proc Technol 140:358–362
25. Liu K, Li XP, Rahman M (2002) Study on surface topography in nanometric ductile cutting of
silicon wafers. In: 4th EPTC Proceedings, Singapore, pp 200–205
26. Uddin MS, Seah KHW, Li XP et al (2004) Effects of crystalline orientation on wear of diamond
tools for nano-scale ductile cutting of silicon. Wear 257:751–759
27. Uddin MS, Seah KHW, Rahman M et al (2007) Performance of single crystal diamond tools
in ductile mode cutting of silicon. J Mater Proc Technol 185:24–30
Chapter 7
Ductile Mode Cutting of Glass
7.1 Introduction
Glass is an inorganic material super-cooled from the molten state to a rigid condi-
tion without crystallizing (or amorphous), having excellent properties such as trans-
parency, attractive look, good corrosion resistance, temperature stability, nonporous,
various reflective indices homogeneity, high hardness and durability. It has been
widely used in optomechatronic systems, laboratory applications, wafer, semicon-
ductor components, optical lenses, windows of buildings, artwork, etc. Unfortunately,
glasses are very difficult to be machined and very often fractured during the machin-
ing process.
Many attempts have been made to assure the quality of the machined glass com-
ponents. Primary activities of previous research on machining of glass are focused
on ductile-regime grinding process. As early in 1975, improvement in precision
diamond grinding mechanism allowed the first reproducible evidence of grinding
ductility in brittle glass workpiece. Surface ground on glass workpiece using a sil-
icon carbide wheel exhibited extensive plastic flow over the surface, while surface
ground with diamond wheels appeared to have been generated by brittle fracture with
some evidence of localized plastic flow [1, 2]. A basic hypothesis was postulated for
ductile-regime grinding: all materials, regardless of their hardness or brittleness, will
undergo a transition from brittle machining regime to ductile machining regime if
the grinding infeed rate is made small enough. Below this threshold of infeed rate,
the energy required to propagate cracks is larger than the energy required for plas-
tic yielding, so plasticity becomes the predominant grinding mechanism [3]. The
float polishing technology was developed for optical materials, including borosili-
cate glasses, fuse silica and amorphous glasses [4, 5]. Very smooth surface on various
optical glass materials can be achieved to expect reducing the polishing time by using
cup-type resinoid-bonded diamond wheel [4]. The conventional grinding makes a
shape on glass parts by crashing or brittle fracture [6]. Optical glass was turned by
heating the work material with a point burner so as to increase the ductility of the
glass material in 1979 [7].
Machining of glass to the required finish for intended applications often poses
a serious challenge. Hardness and brittleness render them very difficult to finish
using conventional turning and grinding machine without causing substantial brittle
fracture. Thus, ductile cutting technology absorbs more and more research interesting
for machining of brittle material. Many effects have been made to investigate the
feasibility of cutting brittle material in ductile deformation mode [8–11]. Ultra-
precision cutting was carried out to investigate the cutting performance of ZKN7
glass using a single crystal diamond tool [12]. Ductile mode cutting was achieved
when depth of cut was less than 600 nm. Ultra-precision ductile cutting of glasses was
conducted by applying ultrasonic vibration on a single crystal diamond tool along
with its cutting velocity direction [13, 14]. It was found that in both the conventional
cutting and ultrasonic vibration assisted cutting, the machined grooves formed from
ductile mode cutting to brittle mode cutting when depth of cut exceeded a critical
value. Ultrasonic vibration would improve the ductile cutting performance of glass
in terms of critical depth of cut up to 1.4 µm, compared with the critical depth of
cut of 200 nm obtained in the conventional cutting. The profile of the tool cutting
edge is perfectly transferred to the cutting groove formed by the ultrasonic vibration
cutting compared with the conventional cutting [13]. This method proposed aims to
increasing the critical depth of cut and feed rate and improving the quality of the
machined surface. However, ductile cutting of glass has not been realized fully up to
now from the practical viewpoint.
Many studies have been done on the ductile-regime machining/grinding of glass,
but not much work has been done on ductile mode cutting of soda-lime glass. In
this chapter, grooving and ductile mode cutting of soda-lime glass are conducted to
investigate its cutting characteristics, including machined glass surface topography,
chip formation, surface roughness and tool wear. And ultra-precision cutting of soda-
lime glass by applying ultrasonic vibration has also been introduced to improve its
ductile mode cutting performance.
Soda-lime glass is the most common form of glass produced. Its nominal chemical
compositions are listed in Table 7.1 [15]. Soda-lime glass is inexpensive, chemically
stable, reasonably hard, and extremely workable, and capable being resoftened and
remelted numerous times if necessary. These qualities make it suitable for manufac-
turing a wide array of glass products, including light bulbs, windowpanes, bottles,
and art objects. Overall, soda-lime glass accounts for about 90% of the manufactured
glass. Table 7.2 lists the typical physical, thermal, mechanical and optical properties
of soda-lime glass [15].
Nanometer scale grooving, groove turning and cutting experiments of soda-lime
glass are carried out to investigate its cutting performance on an ultra-precision
lathe with 10 nm positioning resolution as showed in Fig. 6.3. Soda-lime glass with
75.5 mm in diameter and 2 mm in thickness is used as the workpiece. Glass is bonded
on an aluminium blank using heat-softened glue and then vacuum chucked on the
machine tool spindle. As the layer of glue may not be evenly spread out, pre-trimming
process is carried on the workpiece using a polycrystalline diamond tool (PCD) to
ensure the surface is smooth and ready for the tests. The trimming conditions used
are: spindle rotation speed of 1000 rpm, feed rate of 10 µm/rev, depth of cut of 5 µm
and dry cutting. Subsequently, manual polishing is carried out to provide a mirror-
like surface before the experiments to ensure that the glass surface is completely flat.
A single crystal diamond tool is used as the cutting tool. Its tool geometry parameters
are listed in Table 7.3 [16, 17].
In order to achieve the grooving with depth of cut increasing from zero to a large
value during each test, the workpiece is set to be inclined to the vertical plane at an
inclination of 0.02° as shown in Fig. 3.2. Its grooving speed is 1000 rpm. The cutting
conditions of the groove turning tests are: spindle rotation speed of 1000 rpm, depth
of cut of 10 µm, dry cutting, and no infeed in the radial direction as well. The cutting
conditions of nanometric cutting experiments are listed in Table 7.4 [16, 17]. The
maximum undeformed chip thickness d max is varied in the range of 34.38–319.7 nm
when the depth of cut ao is varied in the range of 40–1800 nm for the nanometric
cutting tests. All tests are conducted under dry cutting so as to collect cutting chips
easily. The machined workpiece surface topography is examined using a scanning
electron microscope (SEM) and an atomic force microscope (AFM). The groove
cross-section profile is measured using a stylus surface profilometer. Chip formation
is examined using SEM. Surface roughness is examined using AFM. Tool wear is
measured using an OMIS.
The grooved soda-lime glass work surfaces are firstly examined using the SEM.
The results are shown in Fig. 7.1 [16, 17]. It is shown that as the grooving started
A
7.3 Grooving Test 139
from the depth of cut as zero then gone on with the depth of cut kept increasing, the
machined workpiece surface is smooth at the beginning part, then changed in the
region near ductile-to-brittle transition section A-A to be rougher and rougher, with
cracks propagating into the workpiece. This indicates that there is a transition from
ductile mode cutting to brittle mode cutting in the grooving process. When the depth
of cut is below a critical level relating to the depth of cut at the ductile-to-brittle
transition section A-A, the chip formation is under ductile mode cutting. Otherwise,
when the depth of cut is exceeded the critical value, the chips are formed under brittle
mode cutting.
An AFM examination is taken on the grooved workpiece surface as shown in
Fig. 7.2 [16, 17]. The grooved workpiece surface is very smooth, showing that it is
obtained under a ductile mode cutting process. The critical depth of cut measured
at the ductile-to-brittle transition section A-A in the groove of soda-lime glass using
the stylus surface profilometer is about 560 nm. The experimental results indicate
that ductile mode cutting of soda-lime glass material can be achieved when the depth
of cut is extremely small in the grooving test.
Groove turning tests are also conducted to investigate the transition of ductile mode
cutting to brittle mode cutting in relation to a critical undeformed chip thickness
[16, 17]. Figure 7.3 shows the SEM photographs of the groove surface achieved in
glass turning processes: (a) full view and (b) close-up view of the turned groove.
140 7 Ductile Mode Cutting of Glass
(a) Full view of the turned groove (b) Close-up view of the turned groove
Fig. 7.3 SEM photographs of the groove surface obtained in the turning test
Figure 7.4 shows the surface profile of the groove obtained in the turning test, where
horizontal axis’s unit is millimeter and vertical axis’s unit is micrometer. It is shown
from Fig. 7.3a that the machined workpiece surface of the groove transited from very
smooth at the groove bottom center to very rough near the top and uncut shoulder
surface as the depth of cut being varied from an extremely small value to a maximum
value, where the smooth bottom area is achieved under ductile mode cutting (DMC)
and two fractured groove edge areas are obtained under brittle mode cutting (BMC).
Based on the experimental results of groove turning of soda-lime glass as shown
in Fig. 7.3, a schematic diagram of ductile-to-brittle transition (DBT) in cutting of
brittle material is illustrated in Fig. 4.3, which shows the rake face of the tool, being
perpendicular to the cutting direction, and d c is the critical undeformed chip thickness
Fig. 7.4 The profile of the groove obtained in the turning test
7.4 Groove Turning Test 141
to the fracture-pit transition on the uncut shoulder. The chip varies in thickness from
zero at the tool center to a maximum value at the top of the uncut shoulder. It shows
schematically that as the chip gets thicker toward the workpiece surface, deeper micro
fracture damage is propagated into the subsurface of the substrate. More details on
ductile-to-brittle transition of the groove can be observed from Fig. 7.3b.
The machined workpiece surfaces obtained under the pre-trimming process are exam-
ined using a SEM and AFM as shown in Figs. 7.5 and 7.6, which show the fractured
workpiece surfaces are achieved under brittle mode cutting. Microstructure of the
trimmed surface consists of large irregular fractures such as peaks and valley. There-
fore, a smaller depth of cut and feed rate is selected for subsequent experiments in
cutting of soda-lime glass using single crystal diamond tools to avoid the microstruc-
ture damage and fracture on the surface.
Fig. 7.7 SEM photographs of the machined soda-lime glass workpiece surface
used are: spindle rotation speed of 1000 rpm, feed rate of 5 µm/rev and depth of cut of
100 nm (its maximum undeformed chip thickness d max calculated being 63.43 nm).
Fractured workpiece surfaces achieved in brittle mode cutting of glass are shown in
Figs. 7.7b and 7.8b, where the cutting conditions used are: spindle rotation speed
of 1000 rpm, feed rate of 5 µm/rev and depth of cut of 1800 nm (its maximum
undeformed chip thickness d max calculated being 319.7 nm). Experimental results
indicate that ductile mode cutting of soda-lime glass and fracture-free surface can be
achieved when the undeformed chip thickness is not larger than the critical value.
7.5 Nanometric Cutting 143
Fig. 7.8 AFM photographs of the machined soda-lime glass workpiece surface
SEM observations on the nanometric ductile machined workpiece surfaces are shown
in Fig. 7.9. It has been clearly observed that machining of soda-lime glass with a single
crystal diamond tool having a tool cutting edge radius of 650 nm, all the machined
surfaces resulting from the undeformed chip thickness in range of 34.38–234.3 nm
are fracture-free. Transparent and smooth surfaces have been obtained through the
machining with no significant valleys across the entire surfaces. Hence, such fracture-
free surface obtained from the cutting of glass at these undeformed chip thickness
indicates the cutting is carried out under ductile mode cutting undertaking a large
compression stress in cutting region.
Surface roughness Ra of the machined workpiece varied with the cutting conditions
are listed in Table 7.5 [8]. The roughness values Ra of the machined workpiece
surfaces achieved in ductile mode cutting (Tests No. 1 to No. 5) varies within a
very narrow range. Compared with Ra values obtained in ductile mode cutting, the
roughness values Ra of machined workpiece surfaces achieved in brittle mode cutting
(Test No. 6) is very large, which indicates that cutting mode has a significant effect
on the machined work surface roughness. However, depth of cut has no significant
effect on the roughness value Ra of machined workpiece surfaces achieved in ductile
mode cutting.
144 7 Ductile Mode Cutting of Glass
Fig. 7.9 SEM photographs of the nanometric machined workpiece surfaces. a d max = 34.38 nm.
b d max = 49.99 nm. c d max = 161.1 nm. d d max = 234.3 nm
Figure 7.10 shows the SEM photographs of chips formed in the nanometric cutting
tests [16, 17]. As shown in Fig. 7.10, the chips are formed in terms of particles under
all cutting conditions with a large chip size comparing with the used depth of cut. The
chips are in irregular shapes and have sharp ends, with appearance to similar crushing
rock. Essentially, the experiment produces a smooth and fracture-free surface but the
7.5 Nanometric Cutting 145
Fig. 7.10 SEM photographs of chip formation. a d max = 49.99 nm. b d max = 161.1 nm. c d max =
234.3 nm. d d max = 319.7 nm
chips formed are in a discontinuous form. Therefore, the formed chips could not be
used as a sole criteria to assess the cutting modes achieved.
Figure 7.11 shows the single crystal diamond tool wear in the nanometric cutting of
glass [16]. The tool wear in cutting of glass mainly occurs on the tool flank face.
It shows that the diamond tool wear on flank face is a typical abrasion wear, which
is almost the same as the regular tool wear in metal cutting process. The VBmax
values of tool flank wear are listed in Table 7.6, showing that longer cutting distance,
larger tool flank wear VBmax . No significant crater wear is observed on the tool rake
face in nanometric cutting of glass. Due to very small depth of cut and feed rate
used in nanometric cutting of glass, the temperature rise in the work-tool interface is
expected to be low. Thus, oxidization, diffusion and graphitization are not expected
to be the dominant wear patterns in nanometric cutting of glass.
146 7 Ductile Mode Cutting of Glass
VBmax
Ultrasonic vibration assisted cutting technology has been developed and employed
to enhance the machinability of glass such as soda-lime glass, fused silica, etc., of
which ultrasonic vibration is applied to a single crystal diamond tool in the cutting
direction [13, 14, 18]. Thereafter, the critical depth of cut for ductile mode cutting of
soda-lime glass with ultrasonic vibration assistance is increased to be around serval
times as compared with that under conventional cutting. Continuous layer chip is
formed and smooth surface is obtained in ultrasonic vibration assisted cutting of
soda-lime glass as shown in Figs. 7.12 and 7.13, respectively [18]. The transparent
surface of soda-lime glass with a surface roughness of 16.19 nm in Ra measured
using a white light interferometer is obtained through face turning with ultrasonic
vibration assistance, which is clarified that ductile mode cutting of soda-lime glass is
achieved successfully [18]. But the extremely short tool life is the main constrain for
realizing the ultrasonic vibration assisted ductile mode cutting of glass in industry.
Fig. 7.12 Continuous layer chip formed in ultrasonic vibration assisted cutting of soda-lime glass
Fig. 7.13 Smooth surface obtained in ultrasonic vibration assisted cutting of soda-lime glass
References
1. Huerta M, Malkin S (1976) Grinding of glass: the mechanics of the process. ASME T J Eng
Ind 98:459–467
2. Huerta M, Malkin S (1976) Grinding of glass: surface strength and fracture strength. ASME T
J Eng Ind 98:468–473
3. Bifano TG, Dow TA, Scattergood RO (1991) Ductile-regime grinding: a new technology for
machining brittle materials. ASME T J Eng Ind 113:184–189
4. Namba Y, Abe M (1993) Ultra precision grinding of optical glasses to produce super-smooth
surfaces. CIRP Ann 42:417–420
5. Takeuchi Y (1996) Ultra precision 3D micro machining of glass. CIRP Ann 45:401–404
6. Komanduri R (1996) On material removal mechanisms in finishing of advance ceramics and
glasses. CIRP Ann 45:509–513
7. Brehm R, van Dun K, Teunissen JCG et al (1979) Transparent single point turning of optical
glass: a phenomenological presentation. Prec Eng 1:207–213
8. Liu K (2002) Ductile cutting for rapid prototyping of tungsten carbide tools. NUS PhD thesis,
Singapore
9. Liu K, Li XP (2001) Modelling of ductile cutting of tungsten carbide. T NAMRI/SME
29:251–258
10. Liu K, Li XP (2001) Ductile cutting of tungsten carbide. J Mater Proc Tech 113:348–354
11. Liu K, Li XP, Liang SY (2004) Nanometer scale ductile cutting of tungsten carbide. J Manu
Proc 6:187–195
12. Fang FZ, Chen LJ (2000) Ultra-precision cutting for ZKN7 glass. CIRP Ann 49:17–20
13. Moriwaki T, Shmoto E, Inoue K (1992) Ultraprecision ductile cutting of glass by applying
ultrasonic vibration. CIRP Ann 41:141–144
14. Zhou M, Wang XJ, Ngoi BK et al (2002) Brittle ductile transition in the diamond cutting of
glasses with the aid of ultrasonic vibration. J Mat Proc Tech 121:243–251
15. https://en.wikipedia.org/wiki/Soda%E2%80%93lime_glass
16. Liu K, Li XP, Liang SY et al (2005) Nanometer scale ductile mode cutting of soda-lime glass.
J Manu Proc 7:95–101
17. Liu K, Li XP, Liang SY et al (2003) Nanometer scale ductile cutting of tungsten carbide. T
NAMRI SME 31:153–160
18. Liu XD, Ding X (2002) Diamond turning of Stavax/glass assisted by ultrasonic vibration.
SIMTech Tech Re 51:1–9
Chapter 8
Ductile Mode Cutting of Tungsten
Carbide
8.1 Introduction
Tungsten carbide (or hard-metal, cemented carbide, sometimes also called sintered
carbide) has been widely used in the industry as cutting and forming tools, drilling
bites, making moulds and dies, hard-facings, and other application potentials due to
its excellent mechanical properties, such as superior strength, high hardness, good
fracture toughness, and high resistance to abrasion and wear. Tungsten carbide is
a bi-phase material, commonly made by powder metallurgy technique—containing
ceramic grains and a metallic binder—in which brittle refractory carbides of the
transition metals, such as WC, TiC, TaC or Cr3 C2 , are combined with a soft and
tough binder metal (most often cobalt, in some case nickel). The schematic diagram
of hard phase WC combined by the soft bonder phase cobalt is shown in Fig. 8.1 [1].
The bonding between carbides and the iron-group metals is very good. Chemical
bonding is especially favorable in the case of WC and cobalt due to their nearly
perfect wetting property, that is, the wetting angle of cobalt on WC is zero. The
excellent properties of tungsten carbide derive from the properties of its constituents.
Namely, the main component (carbide phase) of tungsten carbide provides hardness
and wear-resistance while the ductile binder contributes the toughness necessary for
most application.
The successes of tungsten carbide in engineering applications, however, are lim-
ited by its fabrication method, i.e. powder metallurgy technology. Tungsten carbide
is commonly made by a powder metallurgy process consisting of mould and die
making, powder ball milling, powder consolidating, powder pressing and forming,
pre-sintering, sintering and post-sinter forming [2]. However, for small quantity pro-
ductions and prototyping of tungsten carbide products, the powder metallurgy method
is obviously too costly and time consuming. Moreover, at the present, the tungsten
carbide tool design processes are mostly based on trial-and-error method for most
of the manufacturers. A successful final design may have to go through hundreds
WC
Tungsten monocarbide WC takes the most prominent place of all hard phases in
cemented carbide. More than 98% of all hard-metal grades contain WC, and more
than half of these are pure WC-Co alloys [3]. An important aspect of the W-C phase
diagram is the very small range of homogeneity of WC as shown in Fig. 8.2. The
carbon content corresponds to the theoretical value of the stoichiometric composition
(50 at.% or 6.13 wt% C). This means that there are neither carbon nor tungsten
4000
Liquid Liquid + C
3500
WC1-Z
3000 Liquid + W 2780°C
2500 W γ-W2C
Temperature, °C
β-W2C WC
2000 W + W 2C
α-W2C WC + W2C
1500 1300°C WC + C
1000
W + WC
500
0
0 10% 20% 30% 40% 50% 60%
W Atomic Percent Carbon
c W atoms
C atoms
a3
a2
a1
Fig. 8.3 Schematic diagram of simple hexagonal structure of tungsten carbide crystal
vacancies or interstitials other than those created by thermal activation [4]. Therefore,
characteristics of tungsten carbide, which are very important for our research work,
need to be reviewed since many engineering applications are largely based on its
excellent mechanical and thermal properties.
8.2.1.3 Hardness of WC
The range of fracture toughness values found for WC-Co alloys between 5 and 40
vol% cobalt is quite limited, at the most 8–26 MPa m1/2 . The critical stress intensity
factor and related parameters have been measured in three-point bending for 18
different combinations of different volume fractions of cobalt (5–37%) and grain
size of tungsten carbide (0.7, 1.1 and 2.2 µm) [5]. The experimental results and data
analyses indicated that the stress intensity factor is a unique function of the contiguity
of the carbide phase, which is supported by the role played in the fracture of the joints
between tungsten carbide crystals. The hardness to toughness relationship of fine-
grained WC-Co hardmetal was studied based on Palmqvist indentation toughness
measurements [6]. In general, the higher the hardness of the alloys, the longer is
the indentation cracks, indicating a decrease in fracture toughness with increasing
hardness. The sintering conditions play an important role in co-determining the
hardness to toughness relationship.
Unlike silicon and glass, tungsten carbide is a composite formed primarily by tung-
sten monocarbide (WC) and cobalt. Tungsten monocarbide is a kind of refractory
carbide. Cobalt is a metal having a much lower melting point compared to WC. In
achieving ductile mode cutting of tungsten carbide, the main concern is whether
or not dislocation can occur in tungsten carbide. The possibility of dislocation in
8.2 Tungsten Carbide Characteristics 153
tungsten carbide is analyzed through analyzing the crystal structure and bonding
characteristics.
The crystal structure of WC is a simple HCP structure with two atoms per unit cell
(see Fig. 8.3). The slip plane of WC crystal is {1100} at room temperature, which
is the plane of the densest packing. Slip occurs in directions 0001 (b = 2.83 Å)
and 1120 (b = 2.90 Å). The last direction is characterized by easier slip with
less critical shear stress, resulting from the possibility of the dissociation of perfect
dislocations with Burger’s vector in the direction 1123: 1/3[1120] → 1/6[1123]
+ 1/6[1123], where the Burger’s vector for 1/61123 is only 2.05 Å. Nevertheless,
both slip directions in tungsten require the expenditure of almost identical energy.
As a result, WC has significant microplasticity even at room temperature [7].
Cobalt has the stable hexagonal close-packed (HCP) phase in nature, the same
crystal structure as WC. Its lattice constants are a = 2.5071 Å and c = 4.0695 Å
with an almost ideal axial ratio between the a- and c-axes, c/a = 1.623 [8]. The slip
systems of cobalt crystal are slip plane {1120} and slip direction 0001, slip plane
{1121} and slip directions 1100 and 2113, and Burger’s vector 1/361126 [9].
Therefore, cobalt and WC have the same crystal structure and similar slip systems.
Grain Boundary
Slip Plane
Grain A Grain B
(WC) (Cobalt)
Tungsten carbide is mainly used as machining tools and molds in industry. The wear
behaviors and mechanisms of tungsten carbide as tools and molds have been widely
investigated and well known in the past decades. Little work has been done on the
material removal characteristics and mechanisms of tungsten carbide as workpiece
cut by other tools, such as single crystal diamond (SCD), polycrystalline diamond
(PCD) and cubic boron nitride (CBN). The work on the material removal mechanism
of tungsten carbide could be divided into the three main subtopics: slip characteristics,
sliding wear and grooving wear.
8.2 Tungsten Carbide Characteristics 155
The slip character and the brittle-to-ductile transition of single-phase solids were
investigated as early as in 1964. The results indicated that a transition from brittle
to ductile behaviour was shown experimentally to be a consequence of a change in
slip character rather than of a change in yield stress [10]. Determination of the slip
systems in single crystal of tungsten carbide was made and slip in single crystals of
tungsten monocarbide was observed around Vickers pyramid hardness indentations.
The slip plane is {1100} at room temperature and the slip directions are postulated
at the 0001 and 1120 [5]. Undissociated dislocations with Burgers vectors of
0001, 1/3110 and 1/31213 have been found and these appear to be glissile on
a variety of slip planes [11].
Dry sliding wear of hard materials, such as hot pressed Si3 N4 and SiC, 99% dense
Al2 O3 , WC + 6% Co (tungsten carbide) and Si, using pin-on-disc against a diamond
composite was conducted in air at room temperature [12]. The lowest wear rate was
obtained with tungsten carbide among those materials. The sliding wear of conven-
tional and nanostructured tungsten carbide and the effects of the microstructure on
the wear resistance were investigated by performing a series of unlubricated sliding
wear tests in air with pins of WC-Co composites against silicon nitride disks. The
sliding wear resistance of tungsten carbide depends on a somewhat complex manner
on the cobalt content, the grain size and the hardness of the composites. The sliding
wear of WC-Co composites occurs on a very small scale. No conventional fracture or
plastic deformation of the carbide grains is observed. The high sliding wear perfor-
mance of nanostructured WC-Co composites is commensurate with their hardness
[13]. Plastic flow and fracture in WC single crystals and WC-Co materials show that
deformation results in the development of high compressive stresses that encourages
slips in WC. Optical and transmission electron microscopy studies demonstrate that
plastic flow in the carbide phase always precede fracture [14].
The grooving wear of single-crystal tungsten carbide against diamond was evaluated
in single-tip scratch testing [15]. The single tip grooves were made with a Vickers
diamond indenter and the abrasion tests were performed with diamond and silicon
grits. The experimental results indicated that there was difference in both the amount
of wear and wear mechanisms between different crystallographic directions of WC.
Depending on the direction of the slip planes in relation to the groove direction, the
wear mechanisms change from ductile (grooves parallel to the slip planes) to brittle
(grooves perpendicular to the slip planes).
156 8 Ductile Mode Cutting of Tungsten Carbide
In the milling tests, a 5-axis CNC machining center is used to cut tungsten carbide
workpiece with CBN cutting tools.
8.3 Cutting Performance 157
Figure 8.5 shows the schematic illustration of the cutting experimental setup, where
v is cutting speed and ao is depth of cut in the axial direction [1, 21]. The diameter
of the cutting tool is 23.6 mm, axial rake angle γ p is 13°, radial rake angle γ f is −
4°, peripheral cutting edge angle ψ is 48°, clearance angle α is 20°, cutting edge
inclination angle λs is 0°, minor cutting edge angle κr is 3° and tool included angle
or point angle εr is 120°. The radius of the CBN tool cutting edge or sharpness r is
6.0 µm, tool nose radius R is 0.8 mm and rake angle of its chamfer γ is −41.4°. Only
one fresh CBN insert actually is used to remove the work material and another CBN
insert used as a balance does not activate as a cutting tool to remove work material.
The nanometer scale values of undeformed chip thickness are achieved by arranging
combinations of the tool nose radius R, depth of cut ao and feed rate f, as shown in
Fig. 8.5. The maximum undeformed chip thickness d max can be determined using
Eq. (4.23).
Before the experiments, every tungsten carbide workpiece, together with the fix-
ture used, is ground using a CNC grinder, so as to ensure that the workpiece top
surface is parallel to the bottom surface of the fixture. Here, the depth of cut is in a
micron scale. Nevertheless, the positioning resolution of the machine tool is ±1 µm.
and repeatability is ±1 µm. When comparing the depth of cut used with the machine
Main Spindle
Tool holder
Bolt
ao
Dynamometer
O1 O2
f
R
dmax
ao
tool’s resolution, it is necessary to study the effect of the machine tool’s resolution
on the cutting accuracy. One computer connected with a high sensitive dynamometer
is used to monitor the cutting processes. The cutting is considered as a successful
cutting only when the cutting forces are detected. Pre-experiments are carried out to
verify that the cutting accuracy could be guaranteed when the cutting processes are
monitored. A fresh tungsten carbide insert is used as the workpiece and a fresh CBN
insert is used as the tool for each experiment.
The experimental cutting conditions are shown in Table 8.1. All experiments
are conducted under dry cutting. The cutting forces are measured using a three-
component dynamometer. Every cutting experiment is repeated three times and the
average cutting forces of the three tests are taken. The machined workpiece surface
topography, chip formation, and tool wear are examined using a scanning electron
microscope (SEM). The tool wear on flank face VBmax is measured using an optical
measurement inspection system (OMIS). The surface roughness of the machined
tungsten carbide workpiece is measured using a stylus profiler. Ten points on the
machined workpiece surface are selected for the surface roughness measurement
and the average surface roughness of the ten points is taken.
Typical experimental cutting forces obtained under different cutting conditions are
shown in Fig. 8.6 [1, 21] and the corresponding maximum values of cutting forces
are shown in Table 8.2 [21]. Figure 8.6 displays the cutting force variation in the
duration of two cutting revolutions. It is found that in all cutting tests, the cutting
force F z is the largest one among those cutting forces. Meanwhile, the cutting force
Fz
80
40
0
0 0.005 0.01 0.015 0.02 0.025 0.03
-40
Time (S )
Fz
80
40
0
0 0.005 0.01 0.015 0.02 0.025 0.03
-40
Time (S )
Fz
180
120
60
0
0 0.005 0.01 0.015 0.02 0.025 0.03
-60
Time (S )
(c) v = 296.6m/min, f = 0.01mm/rev & ao = 5μm
160 8 Ductile Mode Cutting of Tungsten Carbide
F x is extremely smaller than other cutting forces. The cutting force F x is only about
one-seventh to one-eleventh of the cutting force F y . While it is one-eleventh to one-
sixteenth of the cutting force F z . Viewing from the measured cutting force signal, it
is found that forced vibrations occur with the nature of frequency of dynamometer
55 kHz during the work material removal process.
The key factors influencing the cutting force are found to be cutting speed, depth
of cut and feed rate. The effect of depth of cut, feed rate, and cutting speed on the
cutting forces in cutting of tungsten carbide are shown in Figs. 8.7, 8.8 and 8.9 [1, 21].
It is found that the cutting forces F x , F y, and F z all increased monotonically when
depth of cut being increased from 1 to 5 µm as shown in Fig. 8.7. As shown in Fig. 8.8,
when feed rate being increased from 0.005 to 0.03 mm/rev, the cutting forces F x ,
100
50
0
0 1 2 3 4 5 6
Depth of Cut ( μ m)
100
50
0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035
Feed Rate (mm/rev)
8.3 Cutting Performance 161
Cutting Force (N )
milling (f = 0.01 mm/rev, ao 160 Fy
= 2 µm) Fz
120
80
40
0
0 200 400 600 800 1000
Cutting Speed (m/min )
F y and F z first increase monotonically, and then reach the maximum values at the
feed rate of 0.02 mm/rev, finally F x , F y, and F z decrease very slowly. Comparing
with the effect of depth of cut and feed rate on cutting forces, the effect of cutting
speed on the cutting forces is more complex as shown in Fig. 8.9. All cutting forces
F x , F y, and F z reach the maximum values at the cutting speed of 296.6 m/min and
rapidly decrease towards the minimum values at the cutting speed of 444.8 m/min.
In all tests, both cutting forces F y and F z are much larger than the cutting force F x .
SEM photographs of the machined workpiece surfaces and the original ground sur-
face are shown in Fig. 8.10 [1, 21]. SEM observations of the machined workpiece
surface indicate that a ductile mode cutting surface with good surface integrity is
obtained as shown in Fig. 8.10b, which is much better than the ground surface as
shown in Fig. 8.10a. It should be noted that the machining tool leaves a periodical
profile—feed marks—on the machined workpiece surface as it moves across the
workpiece both in the cutting and grinding processes. When the undeformed chip
thickness is larger than 1054 nm, some fracture features are found on the machined
workpiece surfaces as shown in Figs. 8.10c and d, for fractured surface and semi
fractured surface, respectively.
The surface profiles of the machined workpiece and the original ground insert are
shown in Fig. 8.11 (vertical axis’s unit is 100 Å, and horizontal axis’s unit is µm) [1,
21]. The ductile cutting conditions in Fig. 8.11a are: cutting speed of 148.3 m/min,
feed rate of 0.01 mm/rev and depth of cut of 4 µm [1, 21]. Both surface profiles are
examined perpendicularly to the machining directions using the surface profiler. As
162 8 Ductile Mode Cutting of Tungsten Carbide
Fig. 8.10 SEM micrographs of the machined surfaces in face milling: a Original ground surface;
b v = 296.6 m/min, ao = 2 µm and f = 0.01 mm/rev; c v = 296.6 m/min, ao = 2 µm and f =
0.02 mm/rev; and d v = 296.6 m/min, ao = 5 µm and f = 0.01 mm/rev
shown in Fig. 8.11a, the surface roughness Ra of the machined workpiece achieved
in ductile mode cutting is 0.430 µm, which is better than that of the original ground
insert 0.6244 µm as shown in Fig. 8.11b.
Comparing the surface roughness and the surface profile achieved in the two pro-
cesses, the spacing between feed marks on the workpiece surface achieved in the
ductile mode cutting is a little larger than that achieved in the grinding as shown
in Fig. 8.11. Experimental results in ductile mode cutting of tungsten carbide indi-
cate that a good surface integrity could be achieved at certain cutting conditions.
Thus, the machined workpiece surface in ductile mode cutting of tungsten carbide
is comparable to the ground surface.
Effect of depth of cut on surface roughness of the machined workpiece is shown
in Fig. 8.12 (cutting speed of 296.6 m/min and feed rate of 0.01 mm/rev). Effect of
feed rate on surface roughness is shown in Fig. 8.13 (cutting speed of 296.6 m/min
and depth of cut of 2 µm). Effect of cutting speed on surface roughness is shown in
Fig. 8.14 (feed rate of 0.01 mm/rev and depth of cut of 2 µm). The surface roughness
increases monotonically when depth of cut being increased as shown in Fig. 8.12, as
8.3 Cutting Performance 163
well as feed rate being increased as shown in Fig. 8.13. However, cutting speed has
no significant effect on surface roughness as shown in Fig. 8.14 [1, 21].
SEM photographs of chips formed under different cutting conditions are shown in
Fig. 8.15, where cutting speed is 296.6 m/min used for all tests but 74.1 m/min for
Fig. 8.15f [1, 21]. SEM observations on the chip formations indicate that continuous
164 8 Ductile Mode Cutting of Tungsten Carbide
sliced chips are formed in ductile mode cutting when the undeformed chip thickness
is equal to or less than 1054 nm as shown in Figs. 8.15a, b, c, e and f. Discontinuous
chips are formed in brittle mode cutting when the undeformed chip thickness is larger
than 1054 nm as shown in Fig. 8.15d.
As shown in Table 8.2 and Figs. 8.6, 8.7, 8.8 and 8.9, in cutting of brittle materials,
such as tungsten carbide, when the maximum undeformed chip thickness d max is
smaller than the tool cutting edge radius r, thrust forces f z are much larger than
cutting forces f x . According to Eq. (2.26), the mean compressive stress σ s in the
cutting region is much larger than the shear stress on the curved shear plane τ s [1,
22]. As a result, crack propagation due to workpiece material pre-existing flaws
is blocked under the large compressive stress, and dislocations dominate the chip
formation process. Therefore, ductile chip formation can be achieved in cutting of
brittle materials under certain cutting conditions.
Tool wear experiments are carried out under different cutting speeds in face milling.
As the face milling is a discontinuous cutting process due to the small tungsten
carbide workpiece size. Thus, the cutting distance represents the actual cutting path
where the cutting tool passes through the workpiece. Figure 8.16 shows the effects
of the cutting distance on tool flank wear VBmax under different cutting speeds [1,
23]. The cutting conditions are: feed rate of 0.01 mm/rev, depth of cut of 3µm, and
cutting speed of 148.3 m/min, 370.7 m/min and 593.1 m/min, respectively, where the
undeformed chip thickness is 803 nm calculated from Eq. (4.23) based on the tool
nose radius, feed rate and depth of cut used. All milling tests are conducted under
dry cutting. It is found that the tool flank wear VBmax of the CBN inserts increases
Tool Flank Wear VBmax (mm)
0.6
0.5
0.4
0.3
0.2 593.1m/min
0.1 370.7m/min
148.3m/min
0
0 100 200 300 400 500 600 700
Cutting Distance (m)
Fig. 8.16 Effect of cutting distance on tool flank wear VBmax (feed rate 0.01 mm/rev and depth of
cut 3 µm)
8.3 Cutting Performance 167
Fig. 8.17 OMIS photographs of flank wear under same cutting distance 210 m (f = 0.01 mm/rev
and ao = 3 µm)
stably when the cutting distance increases under those three different cutting speeds
as shown in Fig. 8.16, and the CBN tool wear is faster when the cutting speed is
higher.
More details can be observed from Fig. 8.17, which shows the OMIS photographs
of the tool flank wear at a same cutting distance 210 m under different cutting speeds:
(a) 148.3 m/min, (b) 370.7 m/min and (c) 593.1 m/min, where the magnifications are
same (×50) [1, 23]. Other cutting conditions are: feed rate of 0.01 mm/rev, depth of
cut of 3µm and dry cutting. It is demonstrated that the higher the cutting speed, the
faster the cutting tool wear and the shorter the tool life.
With the increase of the tool flank wear VBmax , there is a natural and gradual
increase of cutting forces at different cutting speeds as shown in Fig. 8.18: (a) radial
forces F x (cutting component), (b) axial forces F y (feed component), and (c) tan-
gential forces F z (thrust component) [1, 23]. The cutting conditions are: feed rate of
0.01 mm/rev, depth of cut of 3µm, and cutting speeds 148.3 m/min, 370.7 m/min
and 593.1 m/min, respectively. All milling tests are conducted under dry cutting.
Viewed from Fig. 8.18, it is found that the radial forces F x at different cutting speeds
have a trend similar to the axial forces F y and the tangential forces F z . That is, all
the cutting forces F x , F y and F z increase rapidly at the beginning, and then remain
168 8 Ductile Mode Cutting of Tungsten Carbide
20
10
0
0 100 200 300 400 500 600 700
Cutting Distance (m)
148.3m/min
200
150
100
50
0
0 100 200 300 400 500 600 700
Cutting Distance (m)
500 148.3m/min
400
300
200
100
0
0 100 200 300 400 500 600 700
Cutting Distance (m)
0.4
0.2
0
0 100 200 300 400 500 600 700
Cutting Distance (m)
almost unchanged within a cutting duration, but finally increase rapidly again. The
radial forces F x at different cutting speeds are much smaller than the axial forces F y
and tangential forces F z . It is only about one sixth of the axial forces F y and one
thirteenth of the tangential forces F z . It is also indicated that the lower the cutting
speed, the longer the duration in which the cutting forces keep unchanged and the
longer the CBN tool life. The higher the cutting speed, the faster the cutting forces
increases, and the shorter the CBN tool life.
The influence of the cutting distance on surface roughness of the machined tung-
sten carbide workpiece at different cutting speeds (148.3, 370.7 and 593.1 m/min)
is shown in Fig. 8.19 [1, 23]. It is found that the surface roughness Ra values of
the machined tungsten carbide workpiece at different cutting speeds are distributed
randomly within a very narrow range between 0.1 and 0.2 µm as shown in Fig. 8.19.
Both in the rapid wear period and the smooth wear period, the surface roughness Ra
of machined tungsten carbide workpiece is almost unchanged. It is likely that the
cutting speed and cutting distance or cutting duration have no significant effect on
the surface roughness Ra of the machined workpiece in face milling. The unchanged
value Ra of the machined workpiece surface roughness achieved may possibly be
a result of the cutting conditions used rather than the effect of the cutting distance,
because the cutting speeds used are in a limited range and other cutting conditions
remain unchanged.
Examination of the wear surfaces of the CBN tools after cutting of tungsten carbide
material is carried out using a scanning electron microscope (SEM) and an energy
dispersive X-ray spectrometer (EDS). SEM observations on the wear surfaces of a
CBN tool after cutting tungsten carbide with the speed of 370.7 m/min are shown
in Fig. 8.20a and b. SEM photographs of Fig. 8.20 shows that the tool wear mainly
appears on the tool flank face, and no crater wear is formed on the tool rake face.
Chipping and micro-fracture occur on the chamfer of the tool rake face due to the
large compressive stress during the cutting process.
170 8 Ductile Mode Cutting of Tungsten Carbide
Flank Face
Flank Face
Chamfer
Rake Face
Rake Face
Fig. 8.20 SEM micrographs of the CBN tool wear (v = 370.7 m/min)
The high speed ductile mode cutting of tungsten carbide is conducted using the same
CNC machine tool and CBN tools, and cutting conditions are listed in Table 8.3,
where n is spindle rotation speed [1, 23]. All experiments are conducted under dry
cutting. Each cutting experiment is repeated three times.
Typical experimental cutting forces under different cutting conditions are shown in
Fig. 8.21 and the corresponding experimental values of cutting forces are shown
in Table 8.4 [1, 24]. The displayed in Fig. 8.21 is cutting force variations in the
-40
Time (second)
(a) f = 0.01mm/rev & ao = 2μm
120
Fx
90 Fy
Cutting Force (N)
Fz
60
30
0
0 0.002 0.004 0.006 0.008 0.01 0.012
-30
-60
Time (second)
(b) f = 0.02mm/rev & ao = 2μm
250
Fx
200 Fy
Cutting Force (N)
150 Fz
100
50
0
-50 0 0.003 0.006 0.009 0.012
-100
Time (second)
(c) f = 0.01mm/rev & ao = 4μm
duration of two cutting revolutions. It is found that in the all high speed cutting tests,
the cutting force F z is the largest one among those cutting forces; meanwhile, the
cutting force F x is extremely smaller than others. Normally, the cutting force F x is
only about one-seventh of the cutting force F y , and one-twelfth to one-fourteenth of
the cutting force F z .
Forced vibrations are also found to coexist with the work material removal process,
having the nature frequency of dynamometer 55 kHz within the measured cutting
force signal. But in the high speed cutting process, forced vibrations are much more
serious than that in face milling process, as shown in Fig. 8.21. Particularly, com-
paring the force signals shown in Fig. 8.21a with in Fig. 8.21b and c, it seems that
the larger feed rate and depth of cut produce larger forced vibration amplitude.
The influence of depth of cut on cutting forces in high speed cutting of tungsten
carbide is shown in Fig. 8.22 and the influence of feed rate on cutting forces is shown
in Fig. 8.23, where the cutting speed is 741.4 m/min (10,000 rpm), and feed rate is
0.01 mm/rev for Fig. 8.22 while depth of cut is 2µm for Fig. 8.23 [1, 24]. It can
be apparently seen that all cutting forces F x , F y and F z are increased monotonously
when depth of cut and feed rate being increased in the high speed cutting, as shown
in Figs. 8.22 and 8.23.
depths of cut Fy
200 Fz
150
100
50
0
0 1 2 3 4 5 6
Depth of Cut (μm)
rates 120 Fy
Fz
90
60
30
0
0 0.005 0.01 0.015 0.02 0.025 0.03
Feed Rate (mm/rev)
8.4 High Speed Ductile Mode Cutting 173
SEM photographs of the machined tungsten carbide surfaces achieved in high speed
cutting are shown in Fig. 8.24, where cutting speed used is 741.4 m/min (10,000 rpm)
[1, 24]. SEM observations on the machined workpiece surface indicate that a good
surface integrity can be achieved in high speed cutting as shown in Fig. 8.24a (depth
of cut of 2µm and feed rate of 0.01 mm/rev), which is much better than that of the
original ground workpiece surface as shown in Fig. 8.10a. Feed marks are left on the
machined surface as it moves across the workpiece. When feed rate is 0.02 mm/rev
(undeformed chip thickness of 1164 nm) or depth of cut exceeds 5 µm (undeformed
chip thickness of 1054 nm) in high speed cutting, some fracture features is achieved
on its surface and the produced surface is not satisfied as shown in Fig. 8.24b and c.
0.6
0.4
0.2
0
0 1 2 3 4 5 6
Depth of Cut (mm)
Surface Roughness Ra (μm)
0.6
0.4
0.2
0
0 0.005 0.01 0.015 0.02 0.025 0.03
Feed Rate (mm/rev)
SEM micrographs of chips formed under different cutting conditions in high speed
cutting of tungsten carbide are shown in Fig. 8.27 [1, 24]. The cutting conditions
8.4 High Speed Ductile Mode Cutting 175
Fig. 8.27 SEM photographs of chips formed under different cutting conditions
used are: (a) cutting speed of 741.4 m/min, depth of cut of 2 µm and feed rate of
0.01 mm/rev; (b) cutting speed of 741.4 m/min, depth of cut of 2 µm and feed rate
of 0.015 mm/rev; (c) cutting speed of 741.4 m/min, depth of cut of 2 µm and feed
rate of 0.02 mm/rev; and (d) cutting speed of 741.4 m/min, depth of cut of 5 µm and
feed rate of 0.01 mm/rev, respectively.
Observations on the chip formation using SEM indicate that in high speed cutting
of tungsten carbide, continuous slice chips are formed in ductile mode cutting when
feed rate is less than 0.02 mm/rev under cutting speed of 741.4 m/min and depth of
cut of 2 µm as shown in Fig. 8.27a and b. Continuous chips are also formed in ductile
mode cutting when depth of cut is 5 µm (undeformed chip thickness of 1054 nm) as
shown in Fig. 8.27d. Discontinuous chip is formed in brittle mode cutting when feed
rate is equal to or more than 0.02 mm/rev (undeformed chip thickness of 1164 nm)
as shown in Fig. 8.27c. That is, in high speed cutting of tungsten carbide, chips are
formed in ductile mode cutting when undeformed chip thickness is not larger than
1054 nm. As a result, undeformed chip thickness is the key factor influenced the
chip formation mode, which is determined by depth of cut, feed rate and cutting tool
geometry.
176 8 Ductile Mode Cutting of Tungsten Carbide
References
1. Liu K (2002) Ductile cutting for rapid prototyping of tungsten carbide tools. NUS Ph.D. thesis,
Singapore
2. Jenkins I, Wood J (1991) Cemented carbide powders and processing, powder metallurgy: an
overview. Institute of Metals, London, pp 312–330
3. Mohan K (1996) Microstructure of consolidated nanocomposite tungsten carbide – cobalt.
University of Connecticut, Ph.D. thesis, USA
4. Exner HE (1979) Physical and chemical nature of cemented carbide. Int Met Rev 24:149–174
5. Chermant JL, Osterstock F (1976) Fracture toughness and fracture of WC-Co composites. J
Mater Sci 11:1936–1951
6. Schubert WD, Neumeister H, Kinger G et al (1998) Hardness to toughness relationship of
fine-grained WC-Co hardmetals. Int J Refract Met Hard Mater 16:133–142
7. Takahashi T, Freise EJ (1965) Determination of the slip systems in single crystals of tungsten
monocarbide. Philos Mag 12:1–8
8. Korner A, Karnthaler HP (1983) Eeak-beam study of glide dislocation in h.c.p. cobalt. Philos
Mag A 48:469–477
9. Vaidya S, Mahajan S (1980) Accommodation and formation of {112̄1} twins in Co single
crystals. Acta Metall 28:1123–1131
References 177
10. Johnston TL, Davies RG, Stoloff NS (1965) Slip character and the ductile to brittle transition
of single-phase solids. Philos Mag 12:305–317
11. Greenwood RM, Loretto MH, Smallman RE (1982) The defect structure of tungsten carbide
in deformed tungsten carbide-cobalt composites. Acta Metall 30:1193–1196
12. Mehan RL, Hejna CI, McConnell MD (1985) Dry sliding wear of hard materials against a
diamond composite. J Mater Sci 20:1222–1236
13. Jia K, Fischer TE (1997) Sliding wear of conventional and nanostructured cemented carbides.
Wear 203–204:310–318
14. Rowcliffe DJ, Jayaram V, Hibbs MK et al (1988) Compressive deformation and fracture in WC
materials. Mater Sci Eng A 105(106):299–303
15. Engqvist H, Ederyd S, Axen N et al (1999) Grooving wear of single-crystal tungsten carbide.
Wear 230:165–174
16. Shetty DK, Wright IG, Mincer PN et al (1985) Indentation fracture of WC-Co cermets. J Mater
Sci 20:1873–1882
17. Schmid HG (1987) The mechanisms of fracture of WC-11 wt%Co between 20 °C and 1000 °C.
Mater Forum 10:184–197
18. Fischmeister HF, Schmauder S, Sigl LS (1988) Finite element modelling of crack propagation
in WC-Co hart metals. Mater Sci Eng A 105(106):305–311
19. James MN, Human AM, Luyckx S (1990) Fracture toughness testing of hardmetals using
compression-compression precracking. J Mater Sci 25:4810–4814
20. Han D, Mecholsky JJ Jr (1990) Fracture analysis of cobalt-bonded tungsten carbide composites.
J Mater Sci 25:4949–4956
21. Liu K, Li XP, Liang YS (2004) Nanometer-scale ductile cutting of tungsten carbide. J Manuf
Process 6:187–195
22. Liu K, Li XP, Liang SY (2007) The mechanism of ductile chip formation in cutting of brittle
materials. Int J Adv Manuf Technol 33:875–884
23. Liu K, Li XP, Rahman M et al (2003) CBN tool wear in ductile cutting of tungsten carbide.
Wear 255:1344–1351
24. Liu K, Li XP, Rahman M (2003) Characteristics of high speed micro cutting of tungsten carbide.
J Mater Process Technol 140:352–357
Chapter 9
Ductile Mode Cutting of Calcium
Fluoride
9.1 Introduction
Calcium fluoride (CaF2 ) single crystals are ionic mineral fluorites that commonly
serve as optical materials due to the wide range of transmittance [1, 2], excellent laser
thresholds [3], high refractive index homogeneity [4] and low axial and radial-stress
birefringence [5, 6]. CaF2 is also one of the hardest optical materials in tandem with
other fluoride crystals such as barium fluoride, magnesium fluoride, lithium fluoride
[7, 8]. Detailed tables of the material physical and optical properties of CaF2 can
be found in Table 9.1. With excellent optical properties and chemical stability, CaF2
is a prime candidate for various optical instruments such as lenses, beam splitters,
prisms and windows for UV and IR applications. It is also used in laser lithography
operated in shorter wavelengths of the UV region for precision optics that is prone
to inflict laser damage [3, 9, 10].
The Bridgman-Stockbarger technique and Czochralski techniques are the most com-
monly used methods to produce high purity CaF2 optical materials [20]. In the
Bridgman-Stockbarger technique, a seed crystal with the desired crystallographic
orientation and a mixture of raw materials are placed in a crucible and crystallized
by moving from a high-temperature region to a low-temperature region [21]. Large-
diameter CaF2 crystals along the (100), (110) and (111) crystallographic orientations
have been reported to be grown up to 250 mm in diameter using the vertical Bridgman
technique [1]. Post-processing is required for the manufacture of optical components
such as annealing processes which are essential to remove thermal stresses induced
during the crystal growth process [22]. Subsequently, mechanical machining meth-
ods are used to produce optical grade surface finishing and the various lens shapes.
Due to technological transition from 193 to 157 nm lithography, optical components
are subject to stringent requirements on their impurity levels that affect the absorp-
tion coefficients and refractive indices. CaF2 is the only material that is capable of
98% bulk transmission at the 157 nm wavelength [23], on the strict condition of low
absorption coefficient below 0.002 cm−1 [24]. Therefore, great efforts have been
made to improve the crystal growth process for quality optical lenses of CaF2 .
As part of the lens production process, post-growth mechanical processing is
subject to equally strict requirements to fulfil the criteria defined for low-wavelength
transmission. The surface quality of the optical window directly influences the optical
performance and its durability. Chemical-mechanical polishing is an advantageous
method that is used to machine superior CaF2 optical surfaces. A (111) oriented
surface of a roughness of 1.4 nm RMS produced by multi-step polishing has a trans-
mittance of 90.30%, but increases to 91.98% after adopting chemical-mechanical
polishing to achieve a surface roughness of 0.8 nm RMS [17]. Cerium polishing is
even capable of producing angstrom surface finishing (Ra < 2Å, RMS < 2Å) [25].
Magnetorheological finishing (MRF) is another successful technique used to pro-
duce high precision optical surfaces of RMS < 1 nm as a final machining process
9.1 Introduction 181
on CaF2 single crystal optical lenses [26–28]. Higher laser thresholds can also be
achieved through chemical-mechanical polishing [3]. Although the cleaved CaF2
(111) surfaces show the best surface finish achievable which directly relates to laser
thresholds that are readily three times higher than a polished surface [29], the reality
of producing freeform optics by cleaving has not been discovered. Float polish-
ing has been reported to produce the smoothest possible CaF2 single crystals by
mechanical means [30]. As good as the polishing technique may appear, these pro-
cesses present extremely low material removal rates that ultimately determine the
low overall machining productivity.
Semi-finished optical components can be easily manufactured in a relatively
shorter period of time using diamond turning as compared to conventional polishing
methods [31]. Diamond turning processes with high precision machine tool centers
are capable of producing nanometric surface structures on CaF2 for optical resonator
applications with high optical quality factors (Q) of 107 [32]. The use of single
crystal diamond cutting tools enables the fabrication of complex geometries with a
roughness of 3 nm Sa [33]. In this chapter, the mechanical machining process and
considerations for CaF2 single crystals will be discussed with emphasis on the funda-
mental material removal mechanisms of single point cutting in the sub-micrometric
range.
Having understood the stringent requirements of a finished surface for low wave-
length applications and laser damage resistance, the fundamental concept adopted
by most mechanical machining processes to achieve the necessary surface finishing
and subsurface damage will be discussed in this section. The low fracture tough-
ness and brittle nature of CaF2 single crystals classify the material as a difficult-
to-machine material. Careful management of the machining methods is required
to achieve optical-grade surface finishing, i.e. crack-free surfaces. The machining
methods stated in Sect. 9.1.2 embrace the concept of ductile-mode machining by
which material is removed by plastic deformation. The fundamental concept can
be understood in terms of a critical value below which lies the ductile regime, and
brittle failure occurs beyond such a material-dependent value [34, 35]. This critical
value can be defined in terms of indentation depth or undeformed chip thickness
during orthogonal cutting. There are different models adopted to predict the duc-
tile–brittle transition. Puttick et al. [34] modelled the critical value for indentation
after the material yielding characteristics that were described by Young’s modulus
and yield strength, while Blake and Scattergood [36] determined the value based on
Young’s modulus, hardness and fracture toughness. Blackley and Scattergood [36]
later used a similar concept to predict the ductile-regime cutting through adjustments
of machining parameters to define the fracture damaged zone relative to the effective
182 9 Ductile Mode Cutting of Calcium Fluoride
undeformed chip thickness, t, that is controlled by the feed. Based on the geometry
of a round-nose single-point cutting tool (Fig. 9.1), an uncut chip thickness above
the critical value t c will result in a microfracture damage zone extending a distance,
y, beneath the material surface. Final crack-free surfaces will be produced on the
machined plane on the condition that the length of the microfracture defined by yc
stops before the final surface as shown in Fig. 9.1. Blackley and Scattergood [36]
derived a governing equation (Eq. 9.1) that defines the machining parameters (tool
geometry and feed) for ductile-mode machining in face turning.
Z e2f f − R 2 dc2 dc + yc
= − 2 (9.1)
R2 f2 R
There are multiple ways to examine the surface features of brittle materials, some by
basic visual observation of a single plane surface and others using adaptive focusing
to produce a multi-layered image or 3-dimensional structure of the surface. The most
commonly used imaging devices on CaF2 are the basic optical microscope, white
light interferometry, atomic force microscopy, and transmission electron microscopy.
These optical devices are constantly advancing and merging technologies to achieve
9.1 Introduction 183
multiple analytical features in a single device. The fastest method for surface anal-
ysis is through optical imaging to visually evaluate the ductile–brittle transition and
crack morphologies on the top surface of a single flat plane as shown in Fig. 9.2.
In some cases, macrocracks are easily observed using an optical microscope but
the existence of microcracks may require higher magnifications for inspection and
becomes a limitation for the device, depending on the cleanliness and technology of
the optical microscope. Although this method is extremely quick for surface analysis,
it is generally unable to analyze curved surfaces due to the fixed focal length of the
microscope.
The atomic force microscopy (AFM) is capable of analyzing multiple layers of the
surface by running a contact probe across the sample surface. This method enables
the device to measure the surface form accuracy and arithmetic roughness of the
sample. In Fig. 9.3, AFM analysis is performed on a machined CaF2 surface exhibit-
ing brittle failure features along different cutting directions. The AFM produces the
Fig. 9.2 Optical image of a progressive microgroove performed on (111) CaF2 along the [01̄1]
direction
Fig. 9.3 Atomic force microscopic (AFM) images of the brittle failure surface features on CaF2
along different cutting directions: a lamellar fracture; b hybrid structure; c pyramidal cleavage [38]
184 9 Ductile Mode Cutting of Calcium Fluoride
Fig. 9.4 White light interferometry (WLI) images of CaF2 surfaces prepared by optical-polishing
and diamond turning, featuring scattered etch pits after a chemical etching process was done to
identify dislocation concentrations [39]
9.1 Introduction 185
Fig. 9.5 White light interferometry (WLI) images of microgrooves performed along different
crystallographic directions on (111) CaF2 single crystal [39]
Fig. 9.6 High resolution transmission electron microscopic (HR-TEM) images of the various loca-
tions (1–6) on the cross-sectional TEM image of CaF2 (blue arrow denotes the [01̄1](111) direction
[39]
186 9 Ductile Mode Cutting of Calcium Fluoride
Unlike amorphous materials, the machining of CaF2 single crystals involves an addi-
tional consideration to accommodate its anisotropic properties. CaF2 presents differ-
ent critical uncut chip thicknesses along different crystallographic orientations and
cutting directions. The resolved shear stress model is regularly used to describe the
directional dependencies based on the maximum shear stresses and tensile stresses
in relation to the slip systems and the cleavage plane. Plastic deformation will occur
when the resolved shear stress acting along the slip plane exceeds the material-defined
value, but brittle failure will be expected when the resolved tensile stresses acting
perpendicular to the cleavage plane exceed the material-defined value before the slip
system is activated [37]. The cubic ionic crystal structure of CaF2 comprises calcium
ions that are positioned at the corners and the centre of the faces of the unit cell,
while the fluoride ions sit in the tetrahedral lattice sites within. This fluorite structure
of CaF2 has the {100}110 primary slip system at room temperature [40]. Typically,
the (100), (110) and (111) plane orientations are found to exhibit 4-fold, 2-fold and
3-fold symmetries, respectively [40]. This means that an identical critical uncut chip
thickness and brittle failure surface morphology will be expected according to the
number of symmetries on each crystallographic plane. There are various methods to
study these anisotropic effects on CaF2 , such as orthogonal grooving, face turning,
directional microhardness indentation, etc.
Basic studies on the anisotropic properties can be done using micro-indentation
setups before translation to the complex deformation mechanisms like microcut-
ting testing. Knoop hardness testing uses an asymmetric pyramidal indenter with a
long diagonal that can be positioned according to the desired testing direction. As
a standard method for measurement, the hardness can be determined by calculat-
ing the indentation area relative to the applied indentation load. Figure 9.7 shows
a variation in hardness by orienting the asymmetric long diagonal indenter along
different crystallographic directions. O’Neil et al. [13] presented a model to describe
the anisotropy in terms of the mean effective resolved shear stress:
F cos ψ + sin γ
τe = cos φ cos λ (9.2)
A 2
where F is the applied force, A is the indentation area, λ is the angle between the
applied load and the slip plane, ψ is the angle between the crystal face and the axis
of rotation for the slip system, and γ is the angle between the crystal face and the
slip direction. Analysis of the anisotropic effects tends to become complex at higher
temperatures where additional slip systems are activated along more crystallographic
directions, i.e. higher degrees of deformation along more directions.
9.2 Effects of the Anisotropic Properties 187
Fig. 9.8 Schematic of Vickers hardness indentation measurement for fracture toughness
Fig. 9.9 Variations in critical depth of cut and calculated fracture toughness over different directions
on a CaF2 (111) surface. Reprinted by permission from Springer [42]
Face machining is also capable of revealing the anisotropic effect through symmetri-
cal features produced on the machined surfaces. A simple configuration of machining
parameters (e.g. nominal depth of cut and feed) would produce the symmetrical fea-
tures that often comprise of both ductile-regime and brittle-regime machining. The
model proposed by Blake and Scattergood [35] together with the direction-dependent
value of critical uncut chip thickness accurately describes the occurrence during face
turning. Accordingly, face machining of the 3-fold symmetric (111) crystallographic
plane of CaF2 produces tri-symmetrical features as outlined in Fig. 9.10. Within each
9.2 Effects of the Anisotropic Properties 189
group, three surface morphologies can be observed that are described by Yan et al.
[23] in Fig. 9.11 as pitted, cloudy and clear regions.
Fig. 9.12 Schematic of the cutting directions relative to the a (100) plane, b (110) plane, and
c (111) plane [43]
The study on micromachining often sees comparable material strain rates to the dis-
location movement, such that a volume of material being removed at a strain rate
that exceeds the speed for a dislocation to move will inhibit plastic deformation, i.e.
resulting in brittle failure. The measure of material deformation rates during micro-
machining is directly controlled by the effective cutting speed. Little work has been
done on this aspect for CaF2 to arrive at a concrete conclusion. Mizutomo et al.
[45] reported the decrease in critical uncut chip thickness with increasing cutting
speeds (100–2000 mm/min) but Chen et al. [42] claimed the opposite observations
with cutting speeds from 2000–4000 mm/min. In fact, Chen et al. [42] identified
a maximum cutting speed at 3500 mm/min before a decline in the critical uncut
chip thickness. The use of vibration-assisted machining provides an additional intu-
ition on the material response to cutting speeds. Due to the high elliptical vibrating
frequency of the cutting tool and its relative displacement to the workpiece during
machining, the effective cutting speeds can be considerably higher than that in a con-
ventional machining process. Like most cases using vibration-assisted machining,
the critical uncut chip thickness is reportedly improved for CaF2 [46]. Seemingly
being consistent with the results in Ref. [42], the evidence from vibration-assisted
machining tests cast doubt on the existence of a maximum cutting speed to increase
critical uncut chip thickness. However, vibration-assisted machining has its own set
of benefits such as small effective uncut chip thickness and the influence of effec-
tive cutting speeds on the machinability of CaF2 . A description on the machining
technique will be further detailed in Sect. 9.6.
192 9 Ductile Mode Cutting of Calcium Fluoride
Fig. 9.13 Experimental results of the cutting tool rake angle against the critical uncut chip thickness.
Reprinted by permission from Springer [42]
9.3 Effects of Cutting Parameters and Conditions 193
Fig. 9.14 Schematic of Merchant’s force circle with: a positive, and b negative rake angles
Cutting fluids are generally understood to enhance the machining process owing to
the reduction in localized heating and thermal expansion, and lubrication effects [50].
With these fundamental advantages of the use of cutting fluids during machining,
extended tool life has also been reported. However, Yan et al. [48] proposed that the
thermal cooling effect of cutting fluids may in fact be detrimental to the ductile-mode
machining process of CaF2 . Considering that high temperatures will be generated
during chip formation by plastic deformation, the heated machined surfaces will
be subjected to a high thermal gradient when it comes in contact with the cutting
fluids and may cause sudden increases in tensile stresses that leads to crack initiation.
The poor thermal properties of CaF2 (i.e. low thermal conductivity and high thermal
expansion rates) exacerbates the predominant issue. It was suggested that dry cutting
of CaF2 would enhance the machining process by avoiding the abrupt cooling effects
and allowing the heat generated from machining to gradually dissipate away from the
194 9 Ductile Mode Cutting of Calcium Fluoride
Fig. 9.15 Scanning electron microscopic image of a diamond turned CaF2 surface using kerosene
as coolant and exhibiting type A and B crack morphologies [48]
cutting zone (Fig. 9.14). Yan et al. [48] observed the formation of two types of crack
morphologies during diamond turning—type A and type B as shown in Fig. 9.15.
Type A microfracture features are manifested in tens of micrometres owing to both
size effects and crystal anisotropic properties, while Type B microfractures are more
apparent in the hundreds of micrometres and were suggested to be caused by the
thermal effects.
Although it was shown by Yan et al. [14] that the cracks formed due to thermal
effects were mitigated as a result of dry cutting, other benefits of using cutting fluids
on top of increasing the critical uncut chip thickness remain relevant in the machining
process of CaF2 , such as extended tool life and chip evacuation. Figure 9.16 shows
Fig. 9.16 Orthogonal plunge-cuts performed on CaF2 along the (111)[01̄1] direction with and
without air coolant
9.3 Effects of Cutting Parameters and Conditions 195
Numerical simulations on the continuum scale are commonly used to model a wide
variety of applications with finite element method (FEM). Orthogonal microcut-
ting simulations may be performed using various FEM software (e.g. ABAQUS,
DEFORM, etc.). Due to the single crystalline nature of CaF2 , the crystal plasticity
finite element method (CPFEM) is used to best model the microcutting process in
accordance with the crystallographic slip system and Schmid’s law [37]. The con-
stitutive equations governing the plastic deformation can be programmed into the
ABAQUS software using a UMAT subroutine [53, 54]. The plastic shear rate Ḟp is
defined in Eq. 9.4 by the slip rate for a particular slip system γ̇ (α) , and s and n are
the respective slip direction and normal to the slip plane. The slip rate on slip system
(α) is defined based on the critical resolved shear stress τ̇c(α) as a result of strain hard-
ening on slip system (β), strain hardening factor h αβ , and the rate sensitivity factor
m in Eqs. 9.5–9.7. The various parameters for the simulation of CaF2 are defined in
Table 9.2.
Ḟp = γ̇ (α) s(α) ⊗ n(α) Fp (9.4)
α
γ̇0(α) τ (α) /τc(α) sgn τ (α)
1/m
γ̇ (α) = (9.5)
196 9 Ductile Mode Cutting of Calcium Fluoride
τ̇c(α) = h αβ γ̇ (β) (9.6)
β
(β)
a
τc
h αβ = qαβ · h 0 1− (9.7)
τs
Wang et al. [38, 51] employed the CPFEM simulation to demonstrate the asymme-
try of the workpiece stress distributions during an orthogonal microcutting process
of a (111) CaF2 workpiece. By modelling a thin block of material, the side views of
the stress distribution on the work material can be observed in Fig. 9.17. The cutting
tool was removed from the figures for clearer observation of the von Mises stress
distribution ahead of the cutting tool. Differences in the stress distribution along dif-
ferent cutting directions can provide insights into the ductile–brittle transition point
in microcutting of a single crystal. In Fig. 9.17, simulation results, A, B, D, and E,
show high concentrations of stresses that extend into the workpiece material, whilst
only a confined area of stress is observed in C. The simulated results correspond
directly to experimental results, which see the highest critical uncut chip thickness
in microcutting along the [01̄1] direction and significant drops in the ductile–brittle
transition point along the other respectively simulated directions that lead to macro-
crack formation [52]. Normal stress distributions (S 22 ) are also capable of evaluating
the anisotropic critical uncut chip thickness. In the example of Fig. 9.18, the simula-
tion results of microcutting along the [1̄1̄2] direction display high tensile stresses in
the workpiece material that extends ahead of the cutting tool and propagate deeper,
relative to the uncut chip thickness. This agrees well with the experimentally known
result to exhibit brittle failure at low uncut chip thickness along this cutting direction.
On the other hand, the tensile stresses are only observed to extend far ahead of the
cutting tool and within the uncut chip thickness, when cutting along the [01̄1] direc-
tion that exhibits the highest critical uncut chip thickness. Furthermore, the absence
of tensile stress extending downwards at the primary deformation zone indicates the
suppression of subsurface pre-existing microcracks [55].
The asymmetric characteristic observed in the CPFEM model further expounds
the anisotropic properties of the single crystal. The normal stress (S11) distributions
of a microgroove on CaF2 along the [011̄](111) direction in Fig. 9.19a, b display
9.4 Theoretical Simulations 197
Fig. 9.17 CPFEM simulated Mises stress distributions on each side during microcutting on the
(111) plane with cutting directions along: a [1̄1̄2], b [1̄2̄3], c [01̄1], d [13̄2], and e [12̄1] [53]
obvious differences in the stresses from both side views of the simulated cell. Node
a is positioned ahead of the cutting tool where compressive stresses are mainly
observed, and nodes b and c are positioned at the secondary deformation zone beneath
the tool clearance face to study the machined subsurface quality. Node c shows high
concentrations of tensile forces that are conducive for surface crack propagation,
while node b exhibits compressive stresses that inhibits crack propagation. In other
words, the simulated results predict asymmetric crack formation on a microgroove,
which was experimentally shown in an orthogonal progressive microcutting process
in Fig. 9.19c, where lamellar cracks are observed on one side of the groove and a
clear surface is produced on the other. Further studies of the machined subsurface
lattice rotation (Fig. 9.19d) are also experimentally validated using high resolution
transmission electron microscopy (HR-TEM) in Fig. 9.20, both showing a lattice
rotation by ~17° in the immediate subsurface region of CaF2 along the [011̄](111)
direction.
198 9 Ductile Mode Cutting of Calcium Fluoride
Fig. 9.18 CPFEM simulated normal stress distributions (S 22 ) on each side during microcutting on
the (111) plane with cutting directions along: a [1̄1̄2], and b [01̄1] [53]
Despite the advantages of CPFEM simulation such as the inexpensive and accu-
rate modelling of the anisotropic characteristic and the subsurface lattice rotation,
particularly in the case for CaF2 , it has its shortcomings of being able to study the full
micromachining process of brittle materials (i.e. chip formation, form accuracies).
This is due to the lack in material plasticity characterization and severe element dis-
tortion associated with the finite deformation problem of cutting process simulation.
The limitations on the continuum scale method pave way for the use of atomic-scale
methods to resolve the dynamics of machining brittle materials. At present, there is
no reported discrete dislocation dynamic models that have been developed to study
the micromachining of CaF2 . The molecular dynamics (MD) simulation method
models a material as a collection of atoms, in which the trajectories of the atoms
are numerically calculated based on Newton’s equations of motion. The interatomic
reaction forces are defined by an appropriate empirical potential energy function
that utilizes the material properties such as the equation of state and stability, lattice
constants, elastic constants, and energy of sublimation [56]. CaF2 is best modelled
as a face centered cubic structure with space group Fm-3m (lattice constant a =
5.4646 Å) using a Buckingham potential, comprising of Coulomb and pairwise short-
range interaction calculations. The potential energy of each pair of atoms U is defined
as a function of the distance r between the two atoms A and B. The potential energy
assumes the form:
9.4 Theoretical Simulations 199
Fig. 9.19 a–b CPFEM simulated normal stress distributions (S 11 ) on each side of a microgroove on
CaF2 along the [01̄1](111) cutting direction; c white light interferometry image of the microgroove
produced along the [01̄1] direction; d CPFEM simulated results of the subsurface lattice rotation at
nodes a, b, and c [38]
Fig. 9.20 a Cross-sectional transmission electron microscopy image of the machined CaF2 sub-
surface along the [01̄1](111) direction; b–c high resolution TEM (HR-TEM) images of the lattice
order in the immediate subsurface layer and the bulk layer, respectively [38]
200 9 Ductile Mode Cutting of Calcium Fluoride
qAqB r AB C AB
U (r AB ) = + A AB · exp − − (9.8)
4π ε0 r AB ρ AB (r AB )6
where q is the electric charge of each Ca2+ and F− ion, ε0 is the vacuum dielectric
constant, and A, ρ and C are potential parameters. Three terms exist in this poten-
tial. The first term describes the long-range Coulombic interactions based on the
electric charge of each ion, such that the like-charges will result in a repulsion and
opposite charges in attraction. The second term represents the Born-Mayer term that
considers the repulsive interaction due to the overlap in electron clouds over shorter
interatomic distances. It describes the decay in repulsive forces with the increase in
interatomic distances. The last term addresses the short-ranged van der Waals forces.
The potential parameters for CaF2 that have been used by various authors are stated
in Table 9.3. Since double charges on the cations indicate that the Ca2+ ions will
be separated over large interatomic distances, the short-range potential parameters
can be assumed to be negligible (i.e. A++ and C ++ = 0) [57]. In the same way, the
attraction is highly anticipated between Ca2+ and F− , reducing any van de Waals
interaction (i.e. C ± = 0) [57].
At present, theoretical simulation of the nanometric cutting process using molec-
ular dynamics simulation has not been reported on CaF2 but a similar dynamic
deformation process has been performed on the nanoindentation process by Lodes
et al. [59]. For the sake of future developments in dynamic deformation process mod-
elling using molecular dynamics, the supercell details used in [60] will be briefly
discussed. A 48 × 50 × 48 unit cell was created with periodic boundary conditions
and oriented along 111, 211 ¯ and 011̄, respectively. The layer of atoms at the
bottom of the cell was defined to be frozen to restrict the degrees of freedom as
the indenter approaches the (111) surface from the top. The NVE constants were
selected for the indentation simulation, which defines the constant number of atoms
(N), a constant volume (V) and constant internal energy (E). It is crucial to include an
additional step to allow the cells to relax (i.e. to reach equilibrium potential energy),
prior to the actual deformation process.
A network of dislocations can be modelled during the indentation, which cor-
responded well with the threefold symmetry of the deformed crystal orientation as
shown in Fig. 9.21. Dislocation movements coincided with the slip systems on the
{100} planes and the 110 direction. In this indentation problem resolved by Lodes
et al. [59], the indenter was modelled as a force with a radius which corresponds
well with common indentation experimental work. However, most orthogonal cut-
ting problems involve tools with unique geometries such as the rake and clearance
angles, and tool edge radius.
9.5 Techniques to Improve Machinability 201
Multiple variants in the use of a vibrating cutting tool have been considered to
machine difficult-to-machine materials, including linear vibration (1-dimensional)
and elliptical vibration (2-dimensional) as shown in Fig. 9.22. In this machining
process, the nominal cutting speed defined on the machining centre is preferably
lower than the tool vibrating speed to enable separation between the cutting tool
and the workpiece material. Consequently, a nominal cutting speed above a critical
Vcrit = 2π f A (9.9)
formation [65, 66]. In essence, the ion implantation technique is bound to create
lattice structural damage in the substrate material [67]. Understandably, the distortion
and reduction in lattice integrity are expected to have a negative effect on the substrate
material properties. Microcutting of hydrogen ion implanted silicon (H-Si) showed
a delay in the ductile–brittle transition as shown in Fig. 9.23. Cutting and thrust
forces were also reportedly lowered by half during the raster milling process of the
hydrogen implanted silicon (H-Si) sample [68].
At this point, there is no reported result on the enhanced machinability for CaF2
by directly applying ion implantation on the material. The costly process is a major
factor that limits the integration of this technique in fabrication of optical compo-
nents. However, another aspect rises with the integration of ion implantation on the
cutting tools. Considerations for the material subsurface integrity must be made in
the fabrication process of optical components. While most machining processes aim
to increase the production efficiency by increasing the material removal rates, an
Fig. 9.23 Optical images (left) and scanning electron microscopic (SEM) images of microgrooves
produced by raster milling on: a hydrogen ion implanted silicon; and b unmodified silicon. Reprinted
from [68] with permission from Elsevier
204 9 Ductile Mode Cutting of Calcium Fluoride
increase in subsurface damage (SSD) may effectively reduce the overall manufac-
turing productivity with the need for additional post-processing to remove the dam-
aged layer. Mizumoto et al. [69] emphasized on optimizing the cutting conditions to
improve the subsurface integrity, with the use of a boron-doped nano-binderless poly-
crystalline diamond (B-NPD) cutting tool that is reported to reduce friction forces on
the tool rake face. SSD analysis through transmission electron microscopy (TEM)
showed a decrease in the SSD layer thickness (interpreted as a blur layer below the
machined surface) from 250 nm (Fig. 9.24), when machining with a single crystal
diamond tool, to 80 nm when using the B-NPD tool in Fig. 9.25. It was concluded that
the formation of an oxide film on the irradiated diamond surface could improve the
tool friction problem in the case of B-NPD [69, 70]. Hartley [71] observed a similar
trend in reduction of friction of diamond with nitrogen implantation in gramophone
styli wear tests.
However, the ion implantation technique does present a highly complex problem,
particularly regarding the interatomic reactions between different combinations of
projectile atoms and host atoms. Friction reduction could be observed with optical
materials and titanium carbide coatings, but a contrasting phenomenon occurs with
nickel phosphorous (NiP) workpiece materials. It was reported that gallium ion irra-
diated single crystal diamond tools promoted adhesion of workpiece material on the
tool due to an increased affinity and enhanced the tool wear process, resulting in
poor surface finish [72]. Kawasegi et al. [72] also observed that the wear resistance
of the irradiated tool could be improved through a post-irradiation heat treatment
process. Non-diamond phases (amorphous carbon) were explained to be eliminated
during the heat treatment process and thereafter restored the tool mechanical prop-
erties [73]. Subsequent annealing processes of irradiated single crystal silicon were
also reported to have similar effects of lattice restoration [74]. However, it seems that
the cutting tool preparation process to ensure that the mechanical strength of the tool
is sufficient for wear resistance consists of too many procedural steps, which could
hinder the overall productivity.
As mentioned earlier, the ion implantation technique uses two main parameters to
characterize the process (i.e. the implantation energy and fluence), which indicates
an optimization of these parameters alone could achieve the positive effects of an ion
implanted diamond cutting tool. McKenzie et al. [75] showed the influence of the
irradiation dosage on the diamond lattice integrity and observed a maximum dose
of 3 × 1013 ions/cm2 of gallium ions at 30 keV before structural damage occurred.
Therefore, the optimization of the ion irradiation process could potentially preserve
the cutting tool mechanical properties. However, it is uncertain if the reduced coeffi-
cient of friction is exclusive to certain ion species and specific workpiece materials.
The full understanding of an ion implanted cutting tool for micromachining oper-
ations remains a premature technology that has shown promising prospects with
proper ionization parameter controls.
9.5 Techniques to Improve Machinability 205
Fig. 9.24 Transmission electron microscopy (TEM) analysis of a (010)-oriented CaF2 subsurface
machined by a boron-ion implanted nano-binderless polycrystalline diamond cutting tool: a cross-
sectional TEM image; b high resolution TEM images of regions I, II, and III; c Fast Fourier transfer
(FFT) analyses of areas A, B, and C [69]
206 9 Ductile Mode Cutting of Calcium Fluoride
Fig. 9.25 Transmission electron microscopy (TEM) analysis of a (010)-oriented CaF2 subsurface
machined by a single crystal diamond cutting tool: a cross-sectional TEM image; b high resolution
TEM images of regions I, II, and III; c Fast Fourier transfer (FFT) analyses of areas A, B, and C
[69]
9.6 Concluding Remarks 207
Calcium fluoride single crystal is a highly sought-after optical material for its wide
transmission range, which extends its applicability across multiple industries. How-
ever, its brittleness and anisotropic properties create a challenge for the fabrication
of high-quality optical lenses with intricate features. This chapter touches base on
the machining conditions that can be optimized, and the necessary considerations
required to account for during micromachining of the brittle material. In addition,
innovative machining enhancement techniques that have been applied on calcium flu-
oride are reviewed. Numerical modelling techniques are also discussed to elucidate
the understanding of the influence of the material anisotropy on the machinability.
References
16. Phillips WL (1961) Deformation and fracture processes in calcium fluorade single crystals. J
Am Ceram Soc 44:499–506
17. Retherford RS, Sabia R, Sokira VP (2001) Effect of surface quality on transmission perfor-
mance for (111) CaF2 . Appl Surf Sci 183:264–269
18. Hahn D (2014) Calcium fluoride and barium fluoride crystals in optics: multispectral optical
materials for a wide spectrum of applications. Opt Photonik 9:45–48
19. Gupta R, Burnett JH, Griesmann U, Walhout M (1998) Absolute refractive indices and thermal
coefficients of fused silica and calcium fluoride near 193 nm. Appl Opt 37:5964
20. Horowitz A, Biderman S, Amar GB, Laor U, Weiss M, Stern A (1987) The growth of single
crystals of optical materials via the gradient solidification method. J Cryst Growth 85:215–222
21. Manutchehr-Danai M (2009) Bridgman–Stockbarger technique. In: Dictionary of gems and
gemology. Springer, Berlin, p 111
22. Nachimuthu S, Aoshima M, Shimizu S, Sumiya K, Ishibashi H (2007) Improvement in optical
properties of CaF2 single crystals used for nanolithography. In: Hitachi Technical Report, vol
48, pp 7–12
23. Yan J, Syoji K, Tamaki J (2004) Crystallographic effects in micro/nanomachining of single-
crystal calcium fluoride. J Vac Sci Technol B Microelectron Nanom Struct 22:46
24. Rothschild M, Bloomstein TM, Fedynyshyn TH, Kunz RR, Liberman V, Switkes M, Efremow
NN, Palmacci ST, Sedlacek JHC, Hardy DE, Grenville A (2003) Optical lithography. Lincoln
Lab J 14:221–236
25. Darcangelo CM, Sabia R, Stevens HJ, Williamson PJ (2002) Angstrom polishing of calcium
fluoride optical VUV microlithography lens elements and preforms. United States Patent
26. Golini D, Jacobs SD, Kordonski WI, Dumas P (1997) Precision optics fabrication using mag-
netorheological finishing. Adv Mater Opt Precis Struct 67:251–274
27. Shorey AB, Gregg LL, Romanofsky HJ, Arrasmith SR, Kozhinova I, Hubregsen J, Jacobs SD
(1999) Study of material removal during magnetorheological finishing (MRF). Opt Manuf Test
Iii 3782:101–111
28. Tricard M, Dumas PR, Golini D (2004) New industrial applications of magnetorheological
finishing (MRF). In: Frontiers in Optics 2004, Optical Society of America—Technical Digest
29. Gogoll S, Stenzel E, Reichling M, Johansen H, Matthias E (1996) Laser damage of CaF2 (111)
surfaces at 248nm. Appl Surf Sci 96–98:332–340
30. Namba Y, Ohnishi N, Yoshida S, Harada K, Yoshida K, Matsuo T (2004) Ultra-precision float
polishing of calcium fluoride single crystals for deep ultra violet applications. CIRP Ann Manuf
Technol 53:459–462
31. Marsh ER, John BP, Couey JA, Wang J, Grejda RD, Vallance RR, Marsh ER, John BP, Couey
JA, Wang J, Grejda RD, Vallance RR (2005) Predicting surface figure in diamond turned
calcium fluoride using in-process force measurement Predicting surface figure in diamond
turned calcium fluoride using in-process force measurement. J Vac Sci Technol B Microelectron
Nanom Struct Process Meas Phenom 23:84–89
32. Grudinin I, Savchenkov A, Matsko A, Strekalov D, Ilchenko V, Maleki L (2006) Ultra high Q
crystalline microcavities. Opt Commun 265:33–38
33. Gläbe R, Riemer O (2010) Diamond machining of micro-optical components and structures.
In: Proceedings of SPIE, vol 7716. Micro-optics
34. Puttick KE, Rudman MR, Smith KJ, Franks A, Lindsey K (1989) Single-point diamond machin-
ing of glasses. Proc R Soc A Math Phys Eng Sci 426:19–30
35. Blake PN, Scattergood RO (1990) Ductile-regime machining of germanium and silicon. J Am
Ceram Soc 73:949–957
36. Blackley WS, Scattergood RO (1991) Ductile-regime machining model for diamond turning
of brittle materials. Precis Eng 13:95–103
37. Nakasuji T, Kodera S, Hara S, Ikawa N (1990) Diamond turning of brittle materials for optical
components. CIRP Ann 39:89–92
38. Wang H, Riemer O, Rickens K, Brinksmeier E (2016) On the mechanism of asymmetric
ductile-brittle transition in microcutting of (111) CaF2 single crystals. Scr Mater 114:21–26
References 209
39. Wang H, Senthil Kumar A, Riemer O (2018) On the theoretical foundation for the microcutting
of calcium fluoride single crystals at elevated temperatures. Proc Inst Mech Eng Part B J Eng
Manuf 232:1123–1129
40. Mizumoto Y, Kakinuma Y (2018) Revisit of the anisotropic deformation behavior of single-
crystal CaF2 in orthogonal cutting. Precis Eng 53:9–16
41. Fang T, Lambropoulos JC (2002) Microhardness and indentation fracture of potassium dihy-
drogen phosphate (KDP). J Am Ceram Soc 85:174–178
42. Chen X, Xu J, Fang H, Tian R (2017) Influence of cutting parameters on the ductile-brittle
transition of single-crystal calcium fluoride during ultra-precision cutting. Int J Adv Manuf
Technol 89:219–225
43. Azami S, Kudo H, Mizumoto Y, Tanabe T, Yan J, Kakinuma Y (2015) Experimental study
of crystal anisotropy based on ultra-precision cylindrical turning of single-crystal calcium
fluoride. Precis Eng 40:172–181
44. Azami S, Hiroshi K, Tanabe T, Yan J, Kakinuma Y (2014) Experimental analysis of the sur-
face integrity of single-crystal calcium fluoride caused by ultra-precision turning. Proc CIRP
13:225–229
45. Mizumoto Y, Aoyama T, Kakinuma Y (2011) Basic study on ultraprecision machining of
single-crystal calcium fluoride. Proc Eng 19:264–269
46. Suzuki N, Nakamura A, Shamoto E, Harada K, Matsuo M, Osada M (2004) Ultrapreci-
sion micromachining of brittle materials by applying ultrasonic elliptical vibration cutting.
In: Micro-nanomechatronics and human science, 2004 and the fourth symposium micro-
nanomechatronics for information-based society, pp 113–138
47. Dahlman P, Gunnberg F, Jacobson M (2004) The influence of rake angle, cutting feed and
cutting depth on residual stresses in hard turning. J Mater Process Technol 147:181–184
48. Yan J, Tamaki J, Syoji K, Kuriyagawa T (2004) Single-point diamond turning of CaF2 for
nanometric surface. Int J Adv Manuf Technol 24:640–646
49. Patten J, Gao W, Yasuto K (2005) Ductile regime nanomachining of single-crystal silicon
carbide. Trans ASME 127:522–532
50. Chan CY, Lee WB, Wang H (2013) Enhancement of surface finish using water-miscible nano-
cutting fluid in ultra-precision turning. Int J Mach Tools Manuf 73:62–70
51. Luo XC, Sun JN, Chang WL, Ritchie JM (2012) Single point diamond turning of calcium
fluoride optics. Key Eng Mater 516:408–413
52. Lee YJ, Chong JY, Chaudhari A, Wang H (2019) Enhancing ductile-mode cutting of calcium
fluoride single crystals with solidified coating. Int J Precis Eng Manuf Technol (in press)
53. Wang H, Riemer O, Brinksmeier E (2015) Study on the critical chip thickness in microcut-
ting CaF2 single crystals with crystal plasticity finite element method. In: 15th International
conference of the European society for precision engineering and nanotechnology (euspen)
54. Lee WB, Wang H, Chan CY, To S (2013) Finite element modelling of shear angle and cutting
force variation induced by material anisotropy in ultra-precision diamond turning. Int J Mach
Tools Manuf 75:82–86
55. Zhang Q, Lambropoulos JC (2007) A model of CaF2 indentation. J Appl Phys 101:043105
56. Komanduri R, Raff LM (2001) A review on the molecular dynamics simulation of machining
at the atomic scale. Proc Inst Mech Eng Part B J Eng Manuf 215:1639–1672
57. Brass A (1989) Molecular dynamics study of the defect behaviour in fluorite structure crystals
close to the superionic transition. Philos Mag A Phys Condens Matter Struct Defects Mech
Prop 59:843–859
58. Yoshino M, Ogawa Y, Aravindan S (2005) Machining of hard-brittle materials by a single point
tool under external hydrostatic pressure. J Mater Sci Eng 127:837–845
59. Lodes MA, Hartmaier A, Göken M, Durst K (2011) Influence of dislocation density on the
pop-in behavior and indentation size effect in CaF2 single crystals: Experiments and molecular
dynamics simulations. Acta Mater 59:4264–4273
60. Catlow CRA, Hayns MR (1972) A computational study of the F-F interionic potential. J Phys
C Solid State Phys 5:L237–L240
210 9 Ductile Mode Cutting of Calcium Fluoride
61. Zhang J, Cui T, Ge C, Sui Y, Yang H (2016) Review of micro/nano machining by utilizing
elliptical vibration cutting. Int J Mach Tools Manuf 106:109–126
62. Shamoto E, Moriwaki T (1994) Study on elliptical vibration cutting. CIRP Ann Manuf Technol
43:35–38
63. Li ZJ, Fang FZ, Gong H, Zhang XD (2013) Review of diamond-cutting ferrous metals. Int J
Adv Manuf Technol 68:1717–1731
64. Dearnaley G (1975) Ion implantation. Nature 256:701–705
65. Bourgoin JC, Massarani B (1976) Threshold energy for atomic displacement in diamond. Phys
Rev B 14:3690–3694
66. Kinchin GH, Pease RS (1955) The displacement of atoms in solids by radiation. Rep Prog Phys
18:1–51
67. Hunsperger RG, Wolf ED, Shifrin GA, Marsh OJ, Jamba DM (1971) Measurement of lattice
damage caused by ion-implantation doping of semiconductors. Radiat Eff 9:133–138
68. To S, Wang H, Jelenković EV (2013) Enhancement of the machinability of silicon by hydrogen
ion implantation for ultra-precision micro-cutting. Int J Mach Tools Manuf 74:50–55
69. Mizumoto Y, Amano H, Kangawa H, Harano K, Sumiya H, Kakinuma Y (2018) On the
improvement of subsurface quality of CaF2 single crystal machined by boron-doped nano-
polycrystalline diamond tools. Precis Eng 52:73–83
70. Sumiya H, Ikeda K, Arimoto K, Harano K (2016) High wear-resistance characteristic of boron-
doped nano-polycrystalline diamond on optical glass. Diam Relat Mater 70:7–11
71. Hartley NEW (1982) Mechanical property improvements on ion implanted diamond. In:
Metastable mater form by ion implant, pp 295–301
72. Kawasegi N, Niwata T, Morita N, Nishimura K, Sasaoka H (2014) Improving machining
performance of single-crystal diamond tools irradiated by a focused ion beam. Precis Eng
38:174–182
73. Kawasegi N, Ozaki K, Morita N, Nishimura K, Sasaoka H (2014) Single-crystal diamond tools
formed using a focused ion beam: Tool life enhancement via heat treatment. Diam Relat Mater
49:14–18
74. Xiao YJ, Fang FZ, Xu ZW, Wu W, Shen XC (2013) The study of Ga+ FIB implanting crystal
silicon and subsequent annealing. Nucl Instr Meth Phys Res Sect B Beam Interact Mater Atoms
307:253–256
75. McKenzie WR, Quadir MZ, Gass MH, Munroe PR (2011) Focused ion beam implantation of
diamond. Diam Relat Mater 20:1125–1128
76. Dominguez-Rodriguez A, Cheong DS, Heuer AH, Dominguez-Rodriguez A (1991) High-
temperature plastic deformation in Y2 O3 -stabilized ZrO2 single crystals: IV. The secondary
slip systems. Philos Mag A Phys Condens Matter Struct Defects Mech Prop 64:923–929
77. Wang Y, Shi J, Ji C (2014) A numerical study of residual stress induced in machined silicon
surfaces by molecular dynamics simulation. Appl Phys A Mater Sci Process 115:1263–1279
78. Xiao G, To S, Zhang G (2015) Molecular dynamics modelling of brittle-ductile cutting mode
transition: Case study on silicon carbide. Int J Mach Tools Manuf 88:214–222
Part III
Hybrid Ductile Mode Cutting
Chapter 10
Ultrasonic Vibration Assisted Ductile
Mode Cutting
10.1 Introduction
As critical undeformed chip thickness for most brittle material is usually quite small
for conventional cutting process, in most industry fields, it is not recommended to
apply direct cutting to obtain high quality surface on brittle material. Fortunately, the
appearance of ultrasonic vibration assisted cutting (UVC) method has provided a new
direction for achieving better cutting performances in cutting of brittle material. Since
more than two decades ago, UVC has been successfully applied in ultra-precision
diamond turning process, in order to machine non-diamond-turnable metals, like
stainless steel [1] and tungsten alloy [2], as well as brittle material, like silicon
[3] and WC [4]. Such superior cutting performance includes better surface quality
[5, 6], longer tool life [7] and suppression of chatter vibration [8], etc. A lot of
experiments using UVC method have shown that better cutting performance can be
achieved in machining of various materials compared to the conventional cutting
(CC) method. More importantly, experimental investiagtions demonstrated that the
critical undeformed chip thickness for ductile-to-brittle transition can be significantly
increased in UVC of brittle material [9]. The critical undeformed chip thickness could
be several times larger than the one achieved in CC process under identical machining
conditions. Such significantly larger critical undeformed chip thickness makes it
possible to machine brittle material in a convenient and sustainable way to achieve
high quality mirror surface. Due to the unique intermittent cutting mechanics in
UVC, the critical undeformed chip thickness has strong correlation with the nominal
cutting speed, which is observed in CC process [10].
A theoretical prediction model of critical undeformed chip thickness for ultra-
precision UVC of brittle material is presented and described in detail in this chapter.
To achieve this, the material deformation process in each vibration cutting cycle is
studied thoroughly. The energy consumption in material removal by both elastic-
plastic deformation and crack propagation is calculated for each cutting cycle, and
then the specific cutting energies for both ductile mode cutting and brittle mode
cutting can be predicted given specific machining conditions and material properties.
The transitional undeformed chip thickness is believed to be the threshold value at
which the specific cutting energy switches from the ductile mode cutting to the brittle
mode cutting. Finally, the proposed model is verified through experimental tests on
UVC of single crystal silicon workpiece with varying nominal cutting speeds.
In the UVC process, the tool vibrates harmonically along the x-axis at an ultrasonic
frequency f, and the workpiece is fed against the tool with a nominal cutting speed
of vc . Figure 10.1 shows a schematic view of the UVC process, where the points D
and F represent the theoretical cutting-start and cutting-end points, respectively [3].
The tool position relative to the workpiece can be given in the following equation:
ω = 2π f (10.2)
Elastic Elastic
recovery deformation
α
Tool
r
to
F F' D D'
vc/f θf s
vc
In fact, (vt )max acts as a critical nominal cutting speed, above which tool rake
face never separates from uncut work material. Speed ratio is considered as one of
the most essential parameters in UVC [5], which is defined as the ratio of nominal
cutting speed to maximum vibration speed:
vc < ωa (10.5)
(a) (b)
7 6
D F Principal force
6 5 Thrust force
Principal force
5 Thrust force 4
Cutting force, N
Cutting force, N
4
3
3
2
2
1
1
0 0
D' F'
-1 -1
0 0.5 1 1.5 2 2.5 3 0 π/2 π 3π/2 2π
Time, s Phase of vibration, rad
Fig. 10.2 Comparison between experimental transient cutting force in a conventional cutting, and
b vibration cutting (f = 0.25 Hz, a = 20 µm) [3]
216 10 Ultrasonic Vibration Assisted Ductile …
material stress will be released gradually until the elastic energy is fully dissipated at
the actual tool/workpiece disengagement point F as shown in Fig. 10.2b. Although
there still exists interactive force between tool and workpiece during the elastic
rebound process, there is no cutting energy consumption because the cutting tool does
not do work. Such elastic deformation and recovery process is also schematically
illustrated in Fig. 10.1.
A number of experiments and analytical studies have been conducted on the size
dependence of specific cutting energy in conventional cutting process [11, 12]. As the
undeformed chip thickness of a material is reduced to nano-metric scale, its specific
cutting energy will increase sharply to a very high value. It begets a scenario where a
round edge tool attempts to remove the solid single grain of the material resulting in a
high value of specific cutting energy. Such specific cutting energy required to machine
at a very low chip thickness values could not be explained by the energy required
for plastic shearing and for overcoming friction on tool rake face. This phenomenon
should be mainly caused by two reasons. Firstly, the ploughing or elastic recovery of
workpiece along tool flank face plays a significant role when machining with chip
thickness values approaching the edge radius of cutting insert used. Although the
small flank force constitutes an insignificant proportion during macro-scale material
removal process, such force cannot be ignored when the undeformed chip thickness
is reduced to micro scale because the elastic recovery force is usually assumed
to be constant at all scales of material removal. Secondly, in macro-scale material
removal process, undeformed chip thickness is significantly larger than the tool edge
radius, and hence shearing occurs along the easy slip plane out of several available
configurations. However, in micro/nano-scale cutting process, as undeformed chip
thickness is comparable to or even smaller than tool edge radius, the effective shear
angle may become significantly smaller, leading to a larger shear force and hence a
larger specific cutting energy.
In the CC process, the energy in ultra-precision machining process is consumed
at primary/shear zone, secondary zone on tool rake face, and tertiary zone on the tool
flank face. The primary, secondary and tertiary zones are respectively associated with
the energy consumption in plastic deformation of the material along the shear plane,
in overcoming friction at tool-chip interface and tool flank face interface. However,
in the UVC process, there exists an additional step of elastic deformation (D to D as
shown in Fig. 10.1) before plastic deformation during each cutting cycle compared to
the CC process, which will lead to a different formulation of specific cutting energy
for unit volume of material removal. The specific cutting energy in UVC will not
only change with the variation of machining parameters, but also with the vibration
parameters, due to the unique intermittent cutting process.
In conventional scale machining process, rake angle and shear angle are treated
to be constant because undeformed chip thickness is usually much higher than tool
10.3 Model Development 217
where α is nominal rake angle (see Fig. 10.1), α e is instantaneous effective rake
angle, ϕ e is instantaneous shear angle, t 0 is undeformed chip thickness, r is tool
edge radius (see Fig. 10.1), and r c is cutting ratio. It is reported that cutting ratio r c
varies between 0.29 and 0.36 for most of the micro machining process [14]. This
assumption is true for micro machining of silicon and cutting ratio can be taken as
0.3 for single-crystal silicon machining [15].
Although there is a significantly reduction of the measured cutting force in UVC
compared to CC, such extremely smaller force may be actually caused by the low
natural frequency (≤5 kHz) of the force measurement system, which acts as a low
pass filter attenuating the high frequency peak force measured by the sensor. There-
fore, the artificially small cutting forces are probably recorded, and it is believed that
some reported results underestimate actual cutting forces in UVC by a factor of 2–5
due to such force attenuation [5]. In fact, according to the low frequency vibration
cutting experiments which eliminates the effect of force attenuation (see Fig. 10.2),
the transient cutting force in plastic deformation step point D to F is just reduced
by 7–20% compared to that in CC [16]. The exact decreasing value of cutting force
depends on the tool/workpiece properties, machining parameters, vibration condi-
tions and so on. Therefore, in this model, the cutting force in plastic deformation step
of UVC is assumed to be equal to that in CC, and is predicted based on an existing
force model.
A tool force model has developed by considering elastic recovery of workpiece
on tool flank face, which is proved to be applicable in precision machining [17]. The
model accounts for chip formation as well as elastic recovery of work material on
tool flank face. Hence, the principal cutting force during plastic deformation step can
be given:
H Ac H
Fc− p = √ + Ac + μ f δ f A f (10.8)
3 3 sin ϕe 3
In groove cutting process, the contact area at flank face due to elastic recovery
has a parabolic shape. Figure 10.3 shows the schematic tool-workpiece contact area
on tool flank face due to material spring back [3]. The contact area at flank face can
be determined by the following equation:
2s
Af = W (10.9)
3 tan θ f
where W is width of cut, θ f is tool clearance angle, and s is elastic recovery of the
freshly machined workpiece surface (see Fig. 10.1). There two parameters can be
calculated from the following equations [17]:
W = 2 t0 (2R − t0 ) (10.10)
H
s = k1 r (10.11)
E
where R is tool nose radius, t 0 is undeformed chip thickness (see Fig. 10.1), k 1 is a
scaling constant for the best fit of Eq. (10.11), and E is elastic modulus.
The material stress on tool flank face can be derived from the following equation:
H
δ f = k2 H (10.12)
E
where k 2 is a scaling constant. Therefore, Eq. (10.8) can now be rewritten as:
H Ac cos ϕe H
Fc− p = √ + 1 + μ f A f K2 H (10.13)
3 3 E
10.3 Model Development 219
Hence, cutting energy spent for plastic deformation step of each cutting cycle in
ductile mode UVC is given by the following expression:
vc
E d− p = Fc− p (10.14)
f
As found in the cutting tests as shown in Fig. 10.2, elastic force always exists in
each cutting cycle due to elastic deformation before plastic deformation step. Hence,
additional cutting energy is consumed during elastic deformation process. According
to Hooke’s law, elastic force is in direct proportion with material deformation. Hence,
transient cutting force for elastic deformation step of each cutting cycle can be given
by the following equation:
L x H Ac cos ϕe H
Fc−e = √ + 1 + μ f A f k2 H (10.15)
Le 3 3 E
L e
E d−e = Fc−e d L x (10.16)
0
Based upon the above two sections of cutting energy consumption in vibration
cutting cycle, total energy for one cutting cycle spent in ductile mode cutting can be
given in the following equation:
E d = E d− p + E d−e (10.17)
where θ is included sector angle for tool nose on a plane perpendicular to cutting
direction at a given t 0 , and its value can be derived through the following equation:
√
−1 t0 (2R − t0 )
θ = tan (10.19)
R − t0
Ed
E sp−d =
Ac vc f
H Ac √ ϕe
cos
+ 1 + μ f A f k2 H HE (1 + s) s f H Ac √ ϕe
cos
+1
3 3 6vc 3
= +
Ac Ac
(10.20)
Cm
Lateral
crack
Median crack
10.3 Model Development 221
It is important to note that median cracks produce subsurface damage and lateral
cracks determine the removal of material in the form of hemi-spherical packets from
bulk material. The removal rate with reference to Fig. 10.5 is given by:
1
Vb = πCl2 vc (10.21)
2
where C l is lateral crack length. It was identified in abrasive machining that the
surface damage could approximately be correlated to the subsurface damage [18]:
Cm
Cl ∼
= (22)
7
where median crack length C m is determined by the critical load of fracture and
fracture toughness of the material [19]:
2/3
4χ Pc
Cm = (10.23)
Kc
where χ is a geometric constant, and its value is 0.064 for most of brittle material;
Pc is the load at the critical point aligned in the direction of median crack; and K c is
work material fracture toughness.
Fracture energy associated with radial and median cracks is due to the creation
of new surfaces and is determined by their surface energy of work material. Hence,
fracture energy associated with generation of new surfaces considering radial and
lateral cracks per unit time is given by:
E f = A s γs
= (2πCl + 2Cm )vc γs (10.24)
E b = E f + f E d− p (10.25)
Hence, the specific cutting energy for brittle mode machining in UVC is given
by:
Eb
E sp−b =
Vb
222 10 Ultrasonic Vibration Assisted Ductile …
2 (2πCl + 2Cm )γs + H Ac √ ϕe
cot
+ 1 + μ f A f k2 H HE
3 3
= (10.26)
πCl2
In order to validate the proposed model for predicting the critical undeformed chip
thickness of ductile-to-brittle transition, a series of UVC tests are performed on a 4-
axis ultra-precision CNC machining center (Toshiba ULG-100H3 ), which has a 10 nm
positioning resolution. A fresh single crystal diamond tool is attached on one tool
holder, which is connected with an ultrasonic vibration device. The ultrasonic vibrator
vibrates along the nominal cutting direction at a fixed frequency of 38.87 kHz, and the
vibration amplitude is set as 2 µm. Polished silicon wafers are used as the workpiece,
which is held on the vacuum chuck. Figure 10.6 shows the experimental setup of
the grooving tests [3]. Table 10.1 shows the cutting conditions and tool/workpiece
properties, including the machining and vibration parameters, tool specifications and
workpiece material properties [3].
The workpiece was glued to a flat plate which is attached on the vacuum chuck,
and face turning was performed to remove the unevenness on the workpiece surface
caused by gluing. Then, the workpiece is tilted at a tiny angle, and groove tests
were then performed by feeding the tool at different nominal cutting speeds (see
Table 10.1) into the workpiece. The onset of brittle fracture was noted by careful
observation of the grove under an optical microscope.
After performing the groove tests, the groove’s widths are measured at the onset
of brittle fracture. The corresponding undeformed chip thickness is calculated from
the measured width of cut using the following equation:
2
Wc
tc = R − R2 − (10.27)
2
Vacuum
chuck
Single crystal
diamond insert
Single crystal
Ultrasonic
silicon wafer
vibration
where W c is critical width of cut for the critical undeformed chip thickness.
The specific cutting energies can be analytically modelled by considering the
workpiece material properties, tool geometry, machining and vibration parameters.
The ductile-to-brittle transition point is considered as the critical undeformed chip
thickness at which ductile mode energy transits into brittle-mode energy. Such point
is determined by simulating the modelled cutting energies for a fairly sufficient range
of undeformed chip thickness based on Eqs. (10.20) and (10.26). Furthermore, for a
given undeformed chip thickness, the specific cutting energies with varying nominal
cutting speeds can also be calculated, and a ductile-to-brittle transitional speed for
UVC of single crystal silicon can be determined.
Given the specific machining parameters and tool/workpiece properties, Fig. 10.7a
shows the specific cutting energy in ductile mode cutting (blue line) and brittle mode
cutting (red line) with the variation of undeformed chip thickness at a nominal cutting
speed of 400 mm/min. It can be seen that the specific cutting energy in ductile mode
cutting decreases nonlinearly with the increase of undeformed chip thickness, while
the specific cutting energy in brittle mode cutting increases in the meantime. The
predictive value of transitional undeformed chip thickness comes out to be 0.39 µm.
Figure 10.7b shows the specific cutting energy with different nominal cutting
speeds at undeformed chip thickness of 0.4 µm [3]. It is interesting to note that the
specific cutting energy in ductile mode cutting in UVC decreases nonlinearly with
the increase of nominal cutting speed, due to the unique elastic-plastic deformation
process in each cutting cycle of the UVC process. The predictive value of critical
speed for ductile-to-brittle transition comes out to be 388 mm/min. It means that, at
224 10 Ultrasonic Vibration Assisted Ductile …
100 Brittle-mode
80
Transition point
60
40
20
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-6
Undeformed chip thickness (m) x 10
(b) 80
70
Specific cutting energy (GPa)
60 Ductile-mode
50
40
Transition speed Brittle-mode
30
20
10
0 500 1000 1500 2000 2500 3000 3500 4000
Nominal cutting speed (mm/min)
0.4 µm undeformed chip thickness, ductile mode material removal can be achieved
in UVC of single crystal silicon below such critical speed.
Besides UVC of single crystal silicon, a comparison grooving test is also con-
ducted using the same silicon material and same diamond tool without applying
ultrasonic vibration [3]. Table 10.2 shows the captured microscope photographs of
the machined grooves on the single crystal silicon under conventional cutting and
UVC methods. The experimental critical undeformed chip thickness is measured to
be 0.20 and 0.45 µm in CC and UVC respectively at 400 mm/min nominal cutting
speed. From Table 10.2, it can also be observed that the ductile-to-brittle transition
point appears at a much higher undeformed chip thickness in UVC than in CC of
single crystal silicon. In the brittle mode cutting region, much less crack has been
observed in UVC compared to those in CC, and the cracks in UVC can only be found
at the center of the groove, where the largest undeformed chip thickness is located.
Such results could further prove the significant advantages of UVC compared to CC
10.4 Experimental Verification 225
Table 10.2 Machined grooves using specified cutting methods (vc = 400 mm/min)
Machining method Ductile mode Transition mode Brittle mode
Conventional
cutting
Ultrasonic
vibration assisted
cutting
on improving surface quality and increasing the critical undeformed chip thickness
in cutting of brittle material.
Table 10.3 shows the predicted and experimental results of critical undeformed
chip thickness with different nominal cutting speeds (or speed ratios) in UVC of sin-
gle crystal silicon. The blue and red lines represent the ductile-mode and brittle-mode
specific-cutting-energy variation curves respectively. The ductile-to-brittle transi-
tion position is located at the intersection point of ductile-mode and brittle-mode
curves. From the predicted specific cutting energy for ductile and brittle modes, it
can be observed the brittle-mode curve stays unvaried with the change of nomi-
nal cutting speed. As the specific cutting energy in ductile-mode UVC decreases
with the increase of nominal cutting speed for a specific undeformed chip thickness,
the ductile-mode specific-cutting-energy curve will move downwardly, leading to
the decrease of transitional undeformed chip thickness for the ductile-brittle mode
transition. Furthermore, the optical images captured by microscope showing ductile-
to-brittle transition area and experimental critical undeformed chip thickness are also
listed in Table 10.3. The critical width of cut W c is measured by the microscope, and
corresponding critical undeformed chip thickness is calculated using Eq. (10.27).
Figure 10.8 shows the predicted and experimental critical undeformed chip thick-
ness with varying nominal cutting speeds. Noted that the predicted t c matches well
226 10 Ultrasonic Vibration Assisted Ductile …
Table 10.3 Predicted and experimental results of critical undeformed chip thickness in UVC of
single crystal silicon
Cutting Speed Predicted Experimental results
speed ratio, Rs results
(mm/min) Predicted Optical photographs Exp. W c Exp. t c
t c (µm) (µm) (µm)
100 0.003 0.75 65 0.75
0.8
Experimental value
0.7 Predicted value
0.5
0.4
0.3
0.2
0.1
0
0 500 1000 1500 2000 2500 3000 3500 4000
Nominal cutting speed (mm/min)
Fig. 10.8 Predicted and experimental critical undeformed chip thickness with varying nominal
cutting speeds in UVC of single crystal silicon
with the experimental values, and the proposed analytical model is able to simu-
late the variation trend of experimental results very well. The predicted and experi-
mental undeformed chip thicknesses increase with the decrease of nominal cutting
speed, which is caused by the decrease of ductile-mode specific cutting energy (see
Table 10.3). According to the analytical study, when the nominal cutting speed is
decreased, the number of vibration cycles needed for per unit volume of material
removal will be larger. Hence, the work material has to experience more cycles for
elastic-plastic deformation to accomplish the material removal process, which will
lead to higher elastic cutting energy and accordingly higher specific cutting energy,
as shown in Eq. (10.20).
References
18. Bifano TG, Fawcett SC (1991) Specific grinding energy as an in-process control variable for
ductile-regime grinding. Prec Eng 13:256–262
19. Marshall DB, Lawn BR (1986) Indentation of brittle materials, microindentation techniques
in materials science and engineering. ASTM STP 889:26–46
20. Leung TP, Lee WB, Lu XM (1998) Diamond turning of silicon substrates in ductile-regime. J
Mater Proc Technol 73:42–48
21. Yoshino M, Ogawa Y, Aravindan S (2005) Machining of hard-brittle materials by a single point
tool under external hydrostatic pressure. J Manufact Sci Eng 127:837–854
Chapter 11
Ultrasonic Vibration Assisted Cutting
of Tungsten Carbide
11.1 Introduction
improved significantly by applying the elliptical vibration to the cutting tool. A finite
element analysis was performed to understand the fracture process of ultrasonically
cutting through brittle material [12]. Compared with conventional cutting in achiev-
ing ductile mode chip formation for grooving of tungsten carbide, the critical depth
of cut for the transition from ductile mode cutting to brittle mode cutting in ultrasonic
vibration assisted grooving was found to be several times larger than that in grooving
without ultrasonic vibration assistance [13]. The critical maximum undeformed chip
thickness for ductile-to-brittle transition in cutting of silicon wafers was varied with
the tool cutting edge radius. The larger the tool cutting edge radius, the larger the
critical maximum undeformed chip thickness [14].
In fact, ductile mode cutting has been demonstrated for cutting of a number of
brittle material, and ultrasonic vibration assisted cutting technology has been verified
showing improvement on ductile cutting performance for some brittle material. In this
chapter, ultrasonic vibration assisted cutting of tungsten carbide has been conducted
to investigate its cutting performance, such as the cutting forces, machined workpiece
surface topography and chip formation.
dx
νu = = ωa sin ωt = 2π f a sin 2π f t (11.3)
dt
Thus, the actual cutting speed in ultrasonic vibration assisted cutting vt can be
given by
vt = v + vu = v + 2π f a sin 2π f t (11.4)
1D ultrasonic vibration cutter is employed in the grooving test to examine the per-
formance of ductile mode cutting of tungsten carbide work material.
234 11 Ultrasonic Vibration Assisted Cutting …
The grooved tungsten carbide workpiece surfaces using 1D ultrasonic vibration cutter
are examined by an optical measurement inspection system (OMIS) and a scanning
electron microscope (SEM). The OMIS photograph of one grooved tungsten carbide
sample surface using 1D ultrasonic vibration cutter is shown in Fig. 11.3 [13], where
11.3 1D Ultrasonic Vibration Assisted Grooving 235
Fig. 11.3 OMIS photograph of the machined groove using 1D ultrasonic vibration cutter
(a) Full view of the groove (b) Close-up view of the groove
Fig. 11.4 SEM micrographs of the machined groove using 1D ultrasonic vibration cutter
the cutting direction is from right to left. It is shown that as the grooving started
from the depth of cut of zero then continued to increase, the machined workpiece
surface is very smooth at the beginning, then changed in the region near the section
A-A to be rougher, with cracks propagating into workpiece. More details of the same
grooved surface are observed using SEM as shown in Fig. 11.4. Here, Fig. 11.4a is a
full view of the machined groove and Fig. 11.4b is a close-up view of the machined
groove nearby the brittle-to-ductile transition region A-A. It is demonstrated that
there is a brittle-to-ductile transition during the grooving of tungsten carbide using
1D ultrasonic vibration cutter, as shown in Figs. 11.4 and 11.5 [13].
Figure 11.5 shows more details on cutting modes of tungsten carbide work material
using 1D ultrasonic vibration cutter nearby brittle-to-ductile transition region: (a)
ductile mode cutting and (b) brittle mode cutting. During the grooving process,
when the depth of cut is below a critical value, its chip formation is under ductile
236 11 Ultrasonic Vibration Assisted Cutting …
Fig. 11.5 SEM micrographs of cutting modes of tungsten carbide groove surface nearby the brittle-
to-ductile transition region using 1D ultrasonic vibration cutter
mode cutting. More details can be seen from the OMIS and SEM observations of the
grooved tungsten carbide workpiece surface nearby the brittle-to-ductile transition
region A-A, as shown in Figs. 11.3, 11.4 and 11.5a.
Otherwise, when the depth of cut is larger than or equal to the critical value, its
chip formation is under brittle mode cutting. More details can be seen from the OMIS
and SEM observations of the machined tungsten carbide workpiece surface nearby
the brittle-to-ductile transition region A-A, as shown in Figs. 11.3, 11.4 and 11.5b.
This is in good agreement with the theoretical prediction in Chap. 4, of which in the
grooving of tungsten carbide work material without ultrasonic vibration assistance,
there is a ductile-to-brittle transition when the undeformed chip thickness or depth of
cut is increased. It should be noted that the vibration marks produced on the groove
surface with the fixed frequency of 19 kHz during 1D ultrasonic vibration assisted
grooving process, are clearly displayed on the machined workpiece surface, both in
ductile mode cutting or brittle mode cutting.
The critical values of depth of cut in 1D ultrasonic vibration assisted grooving tests
are determined by measuring the cross-section profile of the machined grooves at the
brittle-to-ductile transition region A-A using surface profilers Alpha-Step 500 and
Form Talysurf. The transition region A-A is identified by examining the machined
groove surface, as shown in the OMIS photograph and SEM micrographs (see
Figs. 11.3 and 11.4). The criterion for identifying the location of A-A is to set it
as in between the end of the smooth surface and the start of rough surface on the
machined grooves.
The cross-section profiles of two machined grooves at the brittle-to-ductile tran-
sition region A-A using 1D ultrasonic vibration cutter are shown in Fig. 11.6 [13]:
11.3 1D Ultrasonic Vibration Assisted Grooving 237
Fig. 11.6 Profiles of the grooves at the brittle-ductile transition regions formed in the ultrasonic
vibration grooving (both horizontal and vertical axis’s unit is µm)
238 11 Ultrasonic Vibration Assisted Cutting …
Table 11.1 Critical values of depth of cut and material removal ratios in the ultrasonic vibration
assisted grooving tests
No. Critical depth of cut (µm) Material removal ratio f ab
1 24.17 0.998
2 13.81 0.997
3 10.82 0.983
4 11.21 0.995
5 16.61 0.936
6 19.35 0.974
Average 15.995 0.981
(a) the 1st groove, and (b) the 5th groove, where both the horizontal and vertical
axis’s unit is µm. It can be seen that the critical values of depth of cut are 24.017 and
16.61 µm, respectively. Six such machined grooves using the ultrasonic vibration
cutter are obtained. Those critical values of depth of cut are listed in Table 11.1. The
average critical value of depth of cut as 15.995 µm. That is, during 1D ultrasonic
vibration assisted grooving process, the cutting transited from ductile mode cutting
to brittle mode cutting when the depth of cut exceeds the critical value of 15.995 µm.
Based on the surface profiles of machined grooves using 1D ultrasonic vibration
cutter, the amount of work material removal AW and the ratio of material removed
f ab could be calculated from Eqs. (4.24) and (4.25), respectively. Here, the average
ratio of material removal f ab for these given grooves listed in Table 11.1 is 0.981,
which indicates that the material removal mode of tungsten carbide work material
in 1D ultrasonic vibration assisted grooving test at the brittle-to-ductile transition
region A-A is almost under an ideal-cutting mode. That is, tungsten carbide work
material could be grooved under ductile mode cutting using 1D ultrasonic vibration
cutter when the depth of cut is less than 15.995 µm for the given CBN tools used.
Comparing the critical values of depth of cut obtained in the two grooving tests,
it is found that the critical value of depth of cut of 15.995 µm obtained in 1D ultra-
sonic vibration assisted grooving process is much larger than the value of 4.761 µm
obtained in the conventional grooving process. This is in good agreement with the
theoretical result analyzed in Chap. 2, which shows that ultrasonic vibration could be
used to improve the ductile mode cutting performance of tungsten carbide material.
Ultrasonic vibration assisted cutting tests are conducted using a CNC lather, where
ultrasonic vibration is activated on the CBN tools used. Same ultrasonic vibration
11.4 1D Ultrasonic Vibration Assisted Cutting Performance 239
devise and CBN tools are used for 1D ultrasonic vibration assisted cutting tests as
the previous grooving tests. Tungsten carbide workpiece is pre-trimmed using CBN
cutting tools. The cutting conditions of 1D ultrasonic vibration assisted cutting exper-
iments are shown in Table 11.2, where n is spindle rotation speed. All experiments
are conducted under dry cutting. The cutting forces are measured using a three-
dimensional dynamometer. The machined workpiece surface topography and the
chip formation are examined using a scanning electron microscope (SEM). Accord-
ing to the ultrasonic vibration amplitude and frequency used in the experiments, the
maximum velocity of the ultrasonic vibration vu is 107.4 m/min, the actual cutting
speed v is ranged from 127.8 m/min to 766.8 m/min. In order to ensure the accuracy
of experimental results, one fresh tungsten carbide insert and one fresh CBN tool
insert are used for each experiment. Each cutting experiment is repeated three times
and the average cutting forces of the three tests are taken.
Typical experimental cutting forces under different cutting conditions are shown
in Fig. 11.7 and corresponding experimental values of cutting forces are shown in
Table 11.3 [17]. The displayed in Fig. 11.7 is cutting force variation in the duration
of two cutting revolutions. It is found that in all cutting tests, radial cutting force F x is
much larger than tangential cutting force F z and axial cutting force F y . Actually, axial
0.5
Cutting Force (N)
0
0 0.08 0.16 0.24 0.32
-0.5
-1
Fx
-1.5 Fy
Fz
-2
Time (second)
(a) v = 255.6m/min, f = 0.01mm/rev & ao = 5µm
0.5
0
Cutting Force (N)
-1
Fx
-1.5 Fy
Fz
-2
Time (second)
(b) v = 511.2m/min, f = 0.01mm/rev & ao = 5µm
Fig. 11.7 Cutting forces in typical ultrasonic vibration assisted cutting tests
cutting force F y is only about one-seventh of radial cutting force F x , and tangential
cutting force F z is one-third of radial cutting force F x .
Comparing the typical cutting forces obtained under different cutting conditions
as shown in Fig. 11.7, it is found that key factors influencing cutting force are cutting
speed, depth of cut and feed rate. The effects of experimental cutting conditions on
cutting forces are shown in Fig. 11.8 [17]: (a) cutting speed, (b) feed rate, and (c)
depth of cut. It is indicated that the radial cutting force F x is much larger than the axial
cutting force F y and tangential cutting force F z in all tests. As shown in Fig. 11.8a,
all cutting forces F x , F y and F z reach their minimum values when cutting speed is
383.4 m/min. When feed rate is increased from 0.005 to 0.03 mm/rev, radial cutting
force F x is increased slowly and monotonously, meanwhile, axial cutting force F y
and tangential force F z remain almost unchanged, as shown in Fig. 11.8b. All cutting
forces F x , F y and F z are increased rapidly when depth of cut is increased from 2.5 to
25 µm, as shown in Fig. 11.8c. That is, tangential cutting force F z , axial cutting force
F y and radial cutting force F x are increased by 775%, 790% and 850%, respectively.
Comparing the three figures, it can be seen that depth of cut has a big effect on cutting
force increment than that of feed rate and cutting speed, and the effect of cutting speed
on cutting force is in a more complex behavior (see Fig. 11.8a) than that of feed rate
and depth of cut on cutting force (see Fig. 11.8b, c). Likely, it would result from the
compound effects of cutting temperature rise, thermal-depended friction coefficient
between CBN tools and tungsten carbide work material, and actually effective rake
angle of the cutting tools.
Figure 11.9 shows the SEM photographs of chips formed in ultrasonic vibration
assisted cutting tests under different cutting speeds: (a) 127.8 m/min, (b) 255.7 m/min,
(c) 511.2 m/min, (d) 639.0 m/min and (e) 766.8 m/min [17]. Other cutting conditions
used are depth of cut of 5 µm and feed rate of 0.01 mm/rev. As shown in Fig. 11.9,
continuous sliced chips are formed in ductile mode cutting under all cutting speeds.
It seems that the cutting speed has no significant effect on the chip formation in
cutting of tungsten carbide material. But the lower the cutting speed, the thicker the
continuous chips formed together with few discontinuous chips.
Figure 11.10 shows the SEM photographs of chips formed in ultrasonic vibration
assisted cutting tests under different feed rates: (a) 0.005 mm/rev, (b) 0.01 mm/rev,
(c) 0.015 mm/rev, (d) 0.02 mm/rev, (e) 0.025 mm/rev and (f) 0.03 mm/rev [17]. Other
cutting conditions used are cutting speed of 511.2 m/min and depth of cut of 5 µm.
242 11 Ultrasonic Vibration Assisted Cutting …
(e) v = 766.8m/min
Fig. 11.9 SEM photographs of chips formed under different cutting speeds
244 11 Ultrasonic Vibration Assisted Cutting …
Fig. 11.10 SEM photographs of chips formed under different feed rates
11.4 1D Ultrasonic Vibration Assisted Cutting Performance 245
Their maximum undeformed chip thickness can be calculated from the tool nose
radius, feed rate and depth of cut [1]. Continuous sliced chips are formed in ductile
mode cutting when the feed rate is less than 0.03 mm/rev as shown in Fig. 11.10a–e.
That is, when the maximum undeformed chip thickness is less than 3.172 µm, the
cutting is under ductile mode in cutting of tungsten carbide material with ultrasonic
vibration assistance. The smaller the feed rate, the thinner the continuous chips
formed in ultrasonic vibration assisted cutting of tungsten carbide. On the other
hand, discontinuous chips are formed under brittle mode cutting when the maximum
undeformed chip thickness is larger than 3.172 µm, i.e. the feed rate is larger than
0.03 mm/rev, in cutting of tungsten carbide, as shown in Fig. 11.10f. The experimental
results indicate that in cutting of tungsten carbide, once the other cutting conditions
and cutting tool geometry remain unchanged, the cutting modes are dominated by
feed rate.
Figure 11.11 shows the SEM photographs of chips formed in ultrasonic vibration
assisted cutting tests under different depths of cut: (a) 2.5 µm, (b) 5 µm, (c) 10 µm,
(d) 15 µm, (e) 20 µm and (f) 25 µm [17]. Other cutting conditions used are: cutting
speed of 511.2 m/min and feed rate of 0.01 mm/rev. As shown in Figs. 11.11a–e and
11.10b, continuous sliced chips are formed under ductile mode cutting when the depth
of cut is less than 25 µm in ultrasonic vibration assisted cutting of tungsten carbide.
That is, when the maximum undeformed chip thickness is less than 3.369 µm, the
cutting is under ductile mode in cutting of tungsten carbide with ultrasonic vibration
assistance. On the other hand, discontinuous chips are formed under brittle mode
cutting when the maximum undeformed chip thickness is larger than 3.369 µm,
i.e. depth of cut is larger than 25 µm in cutting of tungsten carbide, as shown in
Fig. 11.11f. The experimental results indicate that in cutting of tungsten carbide
with ultrasonic vibration assistance, once the other cutting conditions and cutting
tool geometry remain unchanged, the cutting modes are dominated by depth of cut.
As a result, in cutting of tungsten carbide with ultrasonic vibration assistance,
ductile mode chip formation is mainly determined by undeformed chip thickness.
That is, maximum undeformed chip thickness is mostly dominated by feed rate
and depth of cut used when the tool edge radius remains unchanged. Experimental
results indicate that in cutting of tungsten carbide with ultrasonic vibration assistance,
ductile mode cutting will be achieved when maximum undeformed chip thickness
is less than 3.172 µm, i.e. depth of cut is less than 25 µm or feed rate is less than
0.03 mm/rev. On the other hand, in ultrasonic vibration assisted cutting of tungsten
carbide, chip formation can be achieved under brittle mode cutting when maximum
undeformed chip thickness is larger than 3.172 µm.
246 11 Ultrasonic Vibration Assisted Cutting …
Fig. 11.11 SEM photographs of chips formed under different depths of cut
11.4 1D Ultrasonic Vibration Assisted Cutting Performance 247
SEM photographs of the machined tungsten carbide workpiece surfaces are shown
in Fig. 11.12, where cutting speed used is 511.2 m/min [17]. SEM observations
on the machined tungsten carbide workpiece surfaces indicate that a good surface
integrity is achieved under the depth of cut of 5 µm and feed rate of 0.01 mm/rev as
shown in Fig. 11.12a, compared to the surfaces achieved under the depth of cut of
5 µm and feed rate of 0.03 mm/rev as shown in Fig. 11.11. Figure 11.12b and under
the depth of cut of 20 µm and feed rate of 0.01 mm/rev as shown in Fig. 11.12c. Feed
marks are left on the machined workpiece surface as it moves across the workpiece
surface. SEM observations indicate that when the feed rate is less than 0.03 mm/rev
or depth of cut is less than 25 µm, that is, maximum undeformed chip thickness is
less than 3.172 µm, smooth surface like Fig. 11.12a is achieved under ductile mode
cutting. On the other hand, when feed rate is larger than 0.03 mm/rev or depth of cut
exceeds 25 µm, i.e. maximum undeformed chip thickness is larger than 3.172 µm,
some fractures are found on the machined workpiece surfaces in cutting of tungsten
carbide with ultrasonic vibration assistance, as shown in Fig. 11.12b, c, respectively.
Experimental tests are conducted to study the effects of different cutting and vibra-
tion conditions on surface integrity and tool wear for realizing ductile mode cutting
with ultrasonic vibration assistance [15]. The machining performance of different
combination of cutting and vibration parameters including thrust directional vibra-
tion amplitude, spindle rotation speed and tool nose radius are evaluated through a
series of grooving and face turning tests using both single crystal diamond (SCD)
and polycrystalline diamond (PCD) tools, in terms of diamond tool flank wear and
machined surface quality. The machining tests are performed using a 5-axis ultra-
precision machining system (Moore Nanotech 350FG) as shown in Fig. 11.13, which
Fig. 11.13 Experimental setup for grooving and face turning tests: a Over view, b Close-up view,
and c schematic view of ultrasonic vibration assisted turning test
11.5 2D Ultrasonic Vibration Assisted Cutting Parameters 249
has a 1 nm position resolution for all the 3 linear axes. The ultrasonic vibration cut-
ting system enables the installed tool to vibrate along the cutting direction (a) and the
thrust direction (b) at a fixed frequency of 38.87 kHz, and the peak to zero vibration
amplitudes are adjustable ranging from 0 to 2 µm.
Fresh diamond tools are installed and tightly locked at the tool holder located at
the end of ultrasonic horn. Tool centering with respect to the air spindle is conducted
for each diamond tool before the cutting tests using a small brass workpiece. Dry
air without mist is blown to the cutting zone to take away the heat generated and
blow away the chips for all the tests. The workpiece used is made of ultra-fine
(0.5 µm grain size) tungsten carbide grain bound with 12 wt% cobalt. Its HV and
HRA hardness values given are 17.3 GPa and 92.5 GPa, respectively. The tungsten
carbide workpiece is in a cylindrical rod shape and locked in a fixture, which is then
held on the vacuum chuck of the machining system. Before the machining tests are
conducted, the end face of the tungsten carbide rod is pre-trimmed to obtain a flat
surface.
To compare the cutting performance of 1D ultrasonic vibration and 2D ultrasonic
vibration in terms of surface quality and generation of brittle fracture, 4 grooving
tests are conducted on tungsten carbide workpiece using an SCD tool with 0.2 mm
tool nose radius, −30° rake angle and 10° clearance angle, as shown in Fig. 11.14a
[15]. A nominal cutting speed of 1200 mm/min is applied in the tests, and the cal-
culated speed ratio (Rs ) is 0.041, which is significantly smaller than the maximum
speed ratio derived. After performing the groove tests, microscopic pictures of the
machined grooves are taken and shown in Fig. 11.14b, with respect to different groov-
ing depths. From Fig. 11.14b, for the grooves machined by 2D ultrasonic vibration,
it can be observed that less brittle fracture is observed with the reduction of thrust
directional vibration amplitude (b) at all grooving depths. In particular, when the
thrust directional vibration amplitude is reduced to zero, minimum brittle fracture
is observed on the bottom surface of grooves, for all conditions with either small or
large grooving depth. The phenomenon of increased brittle fracture at 2D vibration
cutting of tungsten carbide could be caused by the punching impact induced by micro
elliptical digging motion of the diamond tool. It has been proven theoretically that 2D
ultrasonic vibration will lead to a much smaller shear plane angle due to the reversed
friction between tool and chip, a smaller partial pneumatic pressure applied on the
tool-workpiece contact zone is expected, which will in turn weaken the ductile mode
cutting capability with ultrasonic vibration assistance.
To confirm the effect of thrust direction vibration amplitude on the suppression of
brittle fraction in cutting of brittle material, the same experimental set up is utilized to
conduct grooving tests on single crystal silicon with higher brittleness using another
single crystal diamond tool with 0.6 mm nose radius. From the results shown in
Fig. 11.15, it can be proven that 1D ultrasonic vibration does perform better than
2D ultrasonic vibration in suppressing fracture generation on the machined brittle
workpiece surface, and lower thrust directional vibration amplitude leads to better
ductile-mode cutting [15].
To evaluate the diamond tool wear conditions in ultrasonic vibration assisted
turning of tungsten carbide workpiece with different combination of vibration and
250 11 Ultrasonic Vibration Assisted Cutting …
Fig. 11.14 a Schematic section view of ultrasonic vibration grooving test on tungsten carbide
workpiece, and b microscopic images of machined grooves at different depths of cut
Fig. 11.15 Microscopic images of ultrasonic vibration cut grooves on silicon at different thrust
directional amplitudes
11.5 2D Ultrasonic Vibration Assisted Cutting Parameters 251
cutting conditions, a series of face turning experiments are conducted using PCD tools
with depth of cut of 4 µm and feed rate of 2.5 µm/rev, and total cutting distance is
calculated to be approximately 31 m. The PCD tools used have a 0° rake angle, 11°
clearance angle and grade number of DA150. Table 11.4 lists the selected machining
parameters used. Constant surface speed is applied to all face turning tests, and their
spindle rotation speed at the outer diameter (OD, φ = 10 mm) is shown in Table 11.4
as well. For example, a nominal cutting speed of 20 rpm at OD is 314 mm/min, and
spindle rotation speed will increase accordingly when the diamond tool approaching
to the center and its relative radial position decreasing. Figure 11.16 shows the rake
and flank tool wear conditions captured using a microscope as well as the values of
flank wear (VB) for these 5 cutting tests.
Table 11.4 Conditions for face turning tests of tungsten carbide with ultrasonic vibration
Exp. no. Vibration Spindle speed at OD (rpm) Tool nose radius (mm)
amplitude
(µm)
a b
I 2 0.5 40 0.2
II 2 0.5 20 0.2
III 2 0 20 0.2
IV 2 0.5 20 0.4
V 2 0 20 0.4
100μm
Tool cutting
Chipped Chipped No chipping No chipping No chipping
edge condition
Flank tool wear, VB, μm
40
37
30
32
28
20
21
10 16
0
I II III IV V
Experiment No.
Fig. 11.16 PCD tool fracture and wear conditions for the 5 face turning tests (31 m cutting distance)
with different cutting conditions
252 11 Ultrasonic Vibration Assisted Cutting …
From Fig. 11.16, by comparing experiments II and III as well as IV and V, it can be
confirmed that 1D ultrasonic vibration performs better than 2D ultrasonic vibration
in terms of flank tool wear rate for both tool nose radius, corresponding with the
result obtained in the grooving tests as illustrated above. Moreover, by comparing
experiments II and IV as well as III and V, it can also be concluded that, under
the same vibration and cutting conditions, a large tool nose radius would lead to a
reduced tool flank wear. This result is caused by the reduced maximum undeformed
chip thickness d max with a larger tool nose radius, and machining of brittle material is
usually very sensitive to this parameter. By comparing experiments I and II, reducing
nominal cutting speed or spindle rotation speed would lead to a smaller tool flank
wear.
Few years ago, nano-polycrystalline diamond (NPD) has been introduced to eliminate
the anisotropy effect of single crystal diamond with better tool wear resistance in
ordinary diamond turning of brittle material. To identify the type of diamond tool
to be used in ultrasonic vibration assisted turning of tungsten carbide, one SCD and
one NPD tool are utilized in 1D ultrasonic vibration (i.e. b = 0) assisted cutting of
tungsten carbide to compare their tool wear condition. The same tungsten carbide
workpiece, nominal cutting speed of 20 rpm at OD, feed rate of 2.5 µm/rev and depth
Fig. 11.17 a Diamond tool wear conditions against the increment of cutting distance, b machined
tungsten carbide workpiece surface taken by microscope
11.6 Effect of Diamond Type on Cutting Performance 253
of cut of 4 µm with the above test using PCD tools are applied in the face turning
test. Here, the diamond tool used has a tool nose radius of 0.2 mm and −30° rake
angle.
It can be observed from Fig. 11.17a that flank wear of NPD tool is always not
smaller than SCD tool with increasing cutting distances, which means that NPD
does not perform better than SCD in ultrasonic vibration assisted turning of tungsten
carbide, not like the ordinary diamond turning process [15]. No catastrophic chipping
on cutting edge has been observed from both types of diamond tools, meaning that
abrasive wear dominates the tool wear mechanism in cutting of tungsten carbide.
The same trend on surface quality of machined tungsten carbide workpiece can
be observed from the microscopic images in Fig. 11.17b, and more obvious feed
marks are found on the machined workpiece using the NPD tool for all the 3 cutting
distances.
when maximum undeformed chip thickness is smaller than a critical value. Corre-
sponding to the variation of chip formation mode from ductile to brittle, three types
of machined workpiece surfaces are obtained: fracture free surface, semi-fractured
surface and fractured surface. Cutting speed had no significant effect on cutting mode
of chip formation.
References
1. Liu K (2002) Ductile cutting for rapid prototyping of tungsten carbide tools. NUS Ph.D. thesis,
Singapore
2. Liu K, Li XP (2001) Ductile cutting of tungsten carbide. J Mater Process Technol 113:348–354
3. Liu K, Li XP (2001) Modelling of ductile cutting of tungsten carbide. T NAMRI/SME
29:251–258
4. Liu K, Li XP, Rahman M et al (2003) CBN tool wear in ductile cutting of tungsten carbide.
Wear 255:1344–1351
5. Moriwaki T, Shamoto E, Inoue K (1992) Ultraprecision ductile cutting of glass by applying
ultrasonic vibration. CIRP Ann 41:141–144
6. Kim JD, Choi IH (1997) Micro surface phenomenon of ductile cutting in the ultrasonic vibration
cutting of optical plastics. J Mater Process Technol 68:89–98
7. Kim JD, Choi IH (1998) Characteristics of chip generation by ultrasonic vibration cutting with
extremely low cutting velocity. Int J Adv Manuf Technol 14:2–6
8. Nerubai MS (1987) Characteristics of contact interaction in ultrasonic cutting of difficult-to-
machine materials. Sov J Fri Wear 8:452–458
9. Lucas M, Graham G, Smith AC (1996) Enhanced vibration control of an ultrasonic cutting
process. Ultrasonics 34:205–211
10. Astashev VK, Babitsky VI (1998) Ultrasonic cutting as a nonlinear (vibro-impact) Process.
Ultrasonics 36:89–96
11. Moriwaki T, Shamoto E (1995) Ultrasonic elliptical vibration cutting. CIRP Ann 44:31–34
12. Smith A, Nurse A, Graham G et al (1996) Ultrasonic cutting—a fracture mechanics model.
Ultrasonics 34:197–203
13. Liu K, Li XP, Rahman M et al (2004) Study of ductile mode cutting in grooving of tungsten
carbide with and without ultrasonic vibration assistance. Int J Adv Manuf Technol 24:389–394
14. Liu K, Li XP, Rahman M et al (2007) A study of the effect of tool cutting edge radius on ductile
cutting of silicon wafers. Int J Adv Manuf Technol 32:631–637
15. Zhang XQ, Huang R, Liu K et al (2018) Suppression of diamond tool wear in machining of
tungsten carbide by combining ultrasonic vibration and electrochemical processing. Cera Int
44:4142–4153
16. Rozenberg LD, Kazantsev VF, Makarov LO et al (1964) Ultrasonic cutting (English ed. by
Balamuth L). Consultants Bureau Enterprises, New York, pp 4–55
17. Liu K, Li XP, Rahman M (2008) Characteristics of ultrasonic vibration assisted ductile cutting
of tungsten carbide. Int J Adv Manuf Technol 35:833–841
Chapter 12
Thermally Assisted Ductile Mode Cutting
12.1 Introduction
the ductile regime, prior to the formation of localised fractures during micro-cutting
of brittle materials. The ideology of ductile-mode cutting is essential for the pro-
duction of crack-free optical grade surface finish on brittle materials by removing
work-material by plastic deformation [10] and can be achieved through controlled
machining conditions [11]. While the realisation of the ductile-regime cutting has
certainly broadened our perspective of brittle-material removal processes, productiv-
ity is still limited by the ductile–brittle transition zone at the sub-micrometre scale.
Enhanced ductile-mode machining techniques include vibration-assisted machining
[12–14] and thermally enhanced machining [15, 16]. It is interesting to observe
that thermally assisted machining plays a significant role in generally improving
the machinability of difficult-to-machine materials. However, this chapter will place
emphasis on the ductile-mode machining of brittle materials under the influence of
elevated temperatures. The integration of heat during machining would therefore
work on the principle of decreased material yield strength and hardness below the
fracture strength with increase in temperature, and thereafter delaying the onset of
brittle failure [17, 18].
Selection of heating methods must be carefully considered to avoid undesir-
able metallurgical changes in the work material [19–21]. Basic requirements for
thermally-assisted machining include the incorporation of a temperature-regulating
device and an instrument to rapidly impart intensive heat to the work material ahead
of the cutting point [22]. Several heating methods have undergone investigation in the
recent years on difficult-to-machine materials (e.g. laser-assisted machining, plasma-
enhanced machining [23, 24], induction heating, etc.). Table 12.1 summarizes the
various heating methods.
Plasma-assisted machining utilizes a plasma arc to soften the work-material ahead
of the cutting tool. A typical setup includes a tungsten electrode centralized in a
copper nozzle with a stream of ionized gas flowing down the annulus to the nozzle
orifice. Figure 12.1 shows a pictorial schematic of the described setup. An initial low-
current arc is used to kick start the process of a high-current arc used for heating.
Temperatures of the arc range between 16,000 and 30,000 °C [25], while the cutting
zone is heated between 400 and 700 °C [20]. In general, cutting forces induced
during machining of Inconel-718 nickel-based super alloy is reduced with the use of
plasma-assisted machining [26], although an increasing in cutting speed could reduce
its effectiveness due to the smaller time frame for heating of the work-material ahead
of the deformation zone. The plasma-assisted machining technique has also been
reported on brittle materials, Pryex, mullite and silicon nitride but failed to enhance
the machinability of others such as alumina and zirconia [27]. Excessive heating
of the materials also proved detrimental toward achieving good machined surface
quality. Another disadvantage of plasma-assisted machining is the large heating spot
size ranging in millimetres [20] in comparison to the unique characteristic of the
micrometric and even the nanometre scale machining of brittle materials to achieve
ductile-mode cutting.
In the meantime, the electrical, induction coil and gas flame heating methods
as described in Table 12.1 are relatively simple and inexpensive to setup but are
reportedly inaccurate and difficult to control. These methods can be classified under
globalised heating methods, which have been recently revisited in enhancing the
ductile-mode cutting of calcium fluoride (CaF2 ) single crystals [28]. While a general
consensus indicates preference over a localised heat source at a small zone [17], the
globalised heating has been proposed to accommodate the material thermal expan-
sion upon heating, which effectively changes the cutting conditions (i.e. undeformed
chip thickness) [29]. However, the localised heating technology has not been fully
developed to accommodate dynamic rotary machining processes such as turning,
milling and drilling. The concept follows by gradually supplying heat to the work-
piece material as a whole to reduce the tendencies for thermal shock to occur in
the workpiece as shown in Fig. 12.2. While the method of heating is undeniable
difficult to control, the concept of a globalised form of heating shows promising
potential, particularly during thermally assisted machining of work materials with
high thermal expansion rates, in this case of calcium fluoride single crystals with a
thermal expansion rate of 18.85 × 10−6 /°C. In theory, the precise control of heat
258 12 Thermally Assisted Ductile Mode Cutting
Fig. 12.2 Schematic of the globalised heating setup: 1—tool holder; 2—cutting tool; 3—work-
piece; 4—heating element; 5—heat shield; 6—spindle
Fig. 12.3 White light interferometry measurements of grooves performed at room temperature and
100 °C: a groove image; b depth profile of each groove [28]
Fig. 12.4 Orthogonal micro-cutting forces with globalised thermal-assistance: a experimental cut-
ting forces; b theoretical thrust forces [30]
260 12 Thermally Assisted Ductile Mode Cutting
use of a CO2 laser rating up to 20 kW [35] but a mere 1.4 kW was required to assist
machining of stainless steels and Udimet 700 [32]. While high power lasers are easily
acquired on an industrial scale, much lower ratings (<400 W) are needed for laser-
assisted machining of non-metallic materials due to the higher adsorption and low
thermal dissipation that could result in the melting or vaporisation of the material
[36]. Any higher laser power would result in laser cutting and induce additional
defects such as a heat affected zone.
The laser-assisted machining methodology encompasses additional parameters
that can be varied according to the work materials to be processed. Basic laser param-
eters include the power intensity, spot size, and beam distribution [37]. Optimisation
of the laser power intensity is essential to avoid the large thermal stress gradients
arising from heating of the material below its melting point, which could eventually
lead to thermal cracking or local buckling [37]. The laser beam normally takes the
shape of a Gaussian profile (Fig. 12.6) that influences the resultant heat transfer (i.e.
the highest laser intensity is located at the centre but exponentially decreases along
the radial direction).
2
x + y2
q = qmax exp − (12.1)
r B2
where q is power density of the heat source, and r B is average radius of the Gaussian
beam or the radius where power density has reduced to 1/e the peak value qmax . The
change in laser intensity effectively influences the resultant temperature distribution
of the heated material.
A novel technique was also adopted to further reap the benefits of laser-assisted
machining with the integration of chemical reagents, otherwise coined as “excimer
laser-assisted chemical machining (ELACM)” [37]. Silicon carbide undergoes tribo-
chemical reactions when in contact with water to form a hydrous oxide that can be
easily machined [38], but still tends to be a time-consuming process. Hibi et al. [39]
12.1 Introduction 261
Fig. 12.8 Schematic of the high-pressure-phase-transition (HPPT) zone located ahead of the cutting
tool. Reprinted from [45] with permission from Elsevier
silicon (Si-I). Substrate temperatures during scratch tests on (100) silicon single crys-
tal along the [110] crystallographic direction with an applied load of 150 mN and
a 45 W laser were reportedly higher than 1000 °C [45]. The notion is not entirely
inaccurate, but as mentioned earlier, the material undergoes plastic deformation in
the Si-I structure at temperatures above 350 °C and negligible metallic phases can be
observed in the quasi-static examination by indentation. Hence, there exists a pos-
sibility that the annealing effects are not the predominant causes for the enhanced
machinability but the fact that the material is simply deformed by an ease of dis-
location movement at higher temperatures. Nonetheless, the optimised laser power
identified to machine silicon single crystals indefinitely improves the machinability
of the brittle material and produces surfaces of pristine crystalline quality. This pre-
dominantly requires the effective heating temperature of the work material to enable
thermal softening through both kinetic excitations of the atoms and the progressive
phase transitions of silicon.
The ductility by high pressure phase transformation has also been reported for other
brittle materials such as germanium [46, 47], silicon carbide [48], and silicon nitride
[49]. However, there are other brittle materials that exhibit plastic behaviour without
the formation of high-pressured metallic phases, such as calcium fluoride (CaF2 )
single crystals. Instead, these materials deform following the fundamental under-
standing of slip deformation and cleavage fracture based on Schmidt’s law. Slip
264 12 Thermally Assisted Ductile Mode Cutting
deformation along the plane with the highest density of atoms and the direction of
the shortest lattice spacing will occur when the resolved shear stress, τ r , exceeds the
critical value, τrcrit . By the same token, cleavage fracture occurs when the resolved
tensile stress on the cleavage plane, σ , favours the formation of a crack based on the
surface energy required to create a surface. These essential values may be calculated
as:
where σ a is applied stress, θ is the angle between the applied stress and the normal
to the slip plane, ζ is the angle between the applied stress and the slip direction on
the slip plane, and ϕ is the angle between the applied stress and the normal to the
cleavage plane.
Wang et al. [50] performed orthogonal plunge-cuts on (111)-plane oriented CaF2
single crystals and analysed the subsurface damage of groove along the ductile-
regime only to find rotated lattice orientations and the absence of an amorphous region
in the immediate subsurface layer (i.e. negligible phase transition). This observation
was also confirmed by Mizumoto et al. [51] with added perspectives of the influence
of single crystal anisotropy on the subsurface damage and concluding that the slip
system was the dominating factor that influences the material plasticity. Wang et al.
[30] adopted the notion and introduced secondary slip system activation in numeri-
cal simulated micro-cutting of CaF2 by the crystal plasticity finite element method
(CPFEM) and observed the reduction in machining forces as well as the ease on
anisotropic characteristic. In their previous work [50], they observed the asymmet-
ric crack formation on the groove along the (111)[01̄1] cutting direction as shown
in Fig. 12.9 (i.e. cracks forming on half portion of the groove) and correlated it
to the asymmetric nature of the cleavage planes relative to the cutting direction.
Numerical modelling through commercially available finite element modelling soft-
ware ABAQUS/Standard reflected the asymmetric subsurface stress concentration
during micro-cutting with the primary slip system {100}110 [50], but displayed
increasingly symmetrical stress features with the activation of secondary slip sys-
tems {111}100 and {110}100 [30]. In the case of CaF2 , the secondary slip systems
activate at 90 °C and 400 °C, respectively [52].
With these fundamentals of Schmidt’s law in mind, the indentation test becomes a
relatively easy method to visualise the relationship between temperature-dependent
activation of slip systems and the enhanced ductility. The (111) cleavage plane and
slip planes of CaF2 can be easily drawn with respect to different crystallographic
planes as shown in Fig. 12.10. At room temperature, indentations on the (110) plane
show the lowest resistance to crack formation at 0.5 N indentation loads with the
cleavage plane located directly perpendicular to the cleavage plane. By resolving
Eq. 12.3, one can easily recognise that cleavage is likely to occur on the condition
of brittle failure being energetically favourable for plastic deformation by slip. With
12.2 Effects on the Ductile–Brittle Transition 265
Fig. 12.9 White light interferometric image of the asymmetric crack formation feature along the
(111)[01̄1] groove. Reprinted from [50] with permission from Elsevier
the slip planes parallel to the indentation direction of the (100) crystal plane and the
resultant forces directed relatively parallel to the cleavage plane, the material is likely
to deform plastically before crack propagates at higher loads. Chaudhari et al. [53]
performed indentations on CaF2 single crystals at elevated temperatures and found
not only the decrease in material hardness with micro-crack suppression across all
crystallographic planes, but also the eventual convergence of material hardness that
was indicative of the ease in single crystal anisotropy as shown in Fig. 12.11.
266 12 Thermally Assisted Ductile Mode Cutting
Fig. 12.10 Schematic of the slip and cleavage planes during indentation of the (111), (110), and
(100) crystal planes of calcium fluoride single crystals
Fig. 12.11 Vickers micro-hardness of CaF2 single crystals at elevated temperatures and the respec-
tive scanning electron microscopic (SEM) images of the indentations
12.2 Effects on the Ductile–Brittle Transition 267
a series of research work on scratch testing and micro-cutting has been performed on
a variety of brittle materials with micro-laser assisted machining (μ-LAM) technol-
ogy integrated with existing ultra-precision machining platforms (Fig. 12.13). The
concept of application differs from conventional laser-assisted machining techniques
where the irradiation source is separated from the cutting region. Instead, the μ-LAM
technology directs the laser through an optical assembly to the rear polished surface
of the diamond tool and travel further into the tool to be finally positioned at the tool
cutting edge as shown in Fig. 12.14. The absorption of the laser light by up to 60% of
the laser irradiance enables the material to be heated in a localised area of <300 μm2
[55].
The humble origins of this promising technology were first reported in 2008
where an infrared diode laser with a wavelength of 1.48 μm and a maximum power
of 400 mW was integrated with a conical diamond stylus to perform laser-assisted
scratch tests [58]. The scratch test is one of the easier methods of characterising the
dynamic ductile–brittle transition during machining, where a load is applied onto the
work-material through a sharp diamond tip and produce a groove on the workpiece.
Depending on the desired results, loading conditions may be constant or with increas-
ing order. Smoother grooves with the absence of brittle cracks would indicate the
material deformation in the ductile-regime and the formation of surface cracks typi-
cally occurring on the groove edges would conversely denote the brittle-regime. The
integration of the micro-laser in a universal material tester directs the laser through
Fig. 12.13 Optimus T+1 and laser control station on an ultra-precision lathe machining a convex
silicon lens [57]
12.3 Micro-Laser Assisted Machining 269
Fig. 12.14 Schematic of the laser delivery system used in the Optimus T+1 [55]
the diamond tip to heat the material during scratching as shown in Fig. 12.15. Force
interactions between the diamond probe and the work-material follow the conven-
tions used in machining (i.e. the cutting force is along the machining direction and
the thrust force is directed perpendicular to the cutting plane or machined surface).
Silicon single crystals were one of the first few materials tested in view of its brittle
nature alike many other ceramics and its application in the semiconductor industry
for integrated circuits and infrared optics [60]. However, its brittleness presents an
issue in the microfabrication of intricate features that are conventionally produced
by grinding, lapping and polishing [61]. In addition, the anisotropic nature of single
Table 12.2 Results of micro-laser assisted scratch testing with increasing loads from 2 to 70 mN
Material Si [62] 4H-SiC [63]
Laser (λ = 1.48 μm) N Y N Y N Y
Thrust force (mN) 20.0 20.0 35.0 42.0 35.0 40.0
Cutting force (mN) 6.0 5.5 9.0 8.0 12.0 14.0
Groove depth (nm) 280 400 480 710 105 240
Scratch morphology Ductile Ductile DBT DBT DBT DBT
the laser dwelling time required for enhanced ductility must not compromise on the
total machining time to produce the final component.
Micromachining involves several types of processes such as milling, drilling and
turning. Laser-assisted machining techniques in milling and drilling follow the con-
cepts used in orthogonal micro-cutting (i.e. straight scratching) where the material is
heated ahead of the cutting tool. In these scenarios, the effective laser dwelling time
remains relatively constant as the tool moves along its designated path. However,
it is still important to determine the relationship between the tool movement speed
and the output power of the heat source to achieve the necessary temperatures on the
material being removed. Patten et al. [59] performed orthogonal scratch tests using
a conical diamond stylus tip under a constant load of 25 mN and cutting speeds of 1
and 305 μm/s along the 1010 crystallographic direction of 6H-SiC single crystals.
At grooving speeds of 305 μm/s, the average groove depth only increased by a small
degree from 41 to 46 nm. In contrast, the average groove depth increased from 54 to
90 nm at grooving speeds of 1 μm/s. It becomes clear that the grooving speed has
a significant influence on the effectiveness of this technique, in that there exists a
minimal time required to thermally soften the material for enhanced ductility.
In turning processes, the ductile–brittle transition differs from orthogonal scratch-
ing in that the critical uncut chip thickness can be highly dependent on the feed (f )
rather than the nominal depth of cut (ap ). Figure 12.16 illustrates a typical cylin-
drical turning and face turning process. Assuming a round-nosed cutting tool and
a relatively small feed in nanometres, one can easily recognise that the thickness
of material being removed is relatively small. While in theory it may also be pos-
sible to achieve ductile-mode machining with a small nominal depth of cut in the
nanometre range and large feed values, the resultant surface finish will be extremely
poor with large peak-valley values across the machined surface. The characterisation
of the ductile–brittle transition using this method was validated by Yan et al. [64]
with a straight-nosed cutting tool for easier representation. Nonetheless, both tool
geometries evoke the same concept to define the critical machining parameters to
ensure ductile mode machining. The feed in Eq. 12.4 determines the feed rate (vf )
Fig. 12.16 Schematics of the cylindrical turning and face turning process
272 12 Thermally Assisted Ductile Mode Cutting
(i.e. lateral velocity of the tool path) and the spindle rotational speed (N), which ulti-
mately defines the effective cutting speed (vc ). The simplified total machining time
(t tot ) to produce simple structures such as a flat plane by face turning and a circular
rod by cylindrical turning is defined by the number of passes to remove the desired
thickness of material as represented in Eq. 12.5. Although the total machining time
does not directly equate to the effective time taken to heat an instantaneous spot on
the material, it provides an indication of the average laser dwell time along each
position on the workpiece. In face turning, the thickness of the material is defined
by h and in cylindrical turning the thickness is defined by the outer radius r 0 and the
final desired radius r 1 of the cylindrical workpiece. L is the length of material to be
removed along the concentric axis of the rod.
vf
f = ; f < f crit (12.4)
N
⎧
⎨ h r0 face turning
= r p−r f L
a v
ttot (12.5)
⎩ 0 1 cylindrical turning
ap vf
The allowable time to heat the material in the example lies with the spindle
rotational speed, which remains relatively constant in cylindrical turning (vc = 2π Nr)
but varies drastically in face turning due to the decreasing radius as the tool moves
towards the centre of the workpiece. It is now easy to understand that while thermally-
assisted machining can increase the critical uncut chip thickness or feed in this
scenario, the time taken to heat the material is also essential with the cutting speed
governed by the rate of heat supplied to an instantaneous area.
There have been comments on the effects of feed rate. The conventional notion
of employing a small feed rate to achieve lower surface roughness could lead to the
additional heating of the material [56]. This was apparent in the study of diamond
turning of silicon crystals during the finishing process at small feeds of 1 μm/rev while
the laser beam diameter was reportedly larger and caused pre-heating of the material
before the actual material removal process as shown in Fig. 12.17. Truncation of the
beam size would help with the process but a separate feature to modify the optical
transmission of the diamond tool with the built-in laser will have to be developed to
conform well with the machining conditions. Mohammadi et al. [56] avoided the issue
with a lower laser output power and observed a drastic change in machined surface
roughness from 25.43 nm Ra at 30 W laser power to 9.78 nm Ra at 20 W laser power.
The concept considers the overall heat transferred into the work-material as the beam
irradiates the pre-machined surfaces, where heat losses by conduction to cutting fluids
and the beam dwelling time in terms of cutting speed must be considered for effective
heating of the work-material and the enjoyment of the benefits in thermally-assisted
machining.
Under conventional notions, the lower cutting forces correspond to lower cutting
tool wear, although the causes are attributed to a wide variety of explanations such
as the generation of heat energy in the deformation zones that are dissipated through
the cutting tool and result in thermal softening of the tool. Naturally, one would
expect a similar result in thermally-assisted machining with lower resistance to plastic
deformation in these brittle materials. But there is a controversial notion that the
enhanced heating in the cutting zone would reduce cutting tool wear as reported
on multiple occasions with the use of the micro-laser assisted machining setup [56,
65, 66]. Shahinian et al. [55] reported the extended tool life in micro-laser assisted
machining of silicon single crystals by a magnitude of 2.5 by measuring the cutting
distance before fracture occurred on the cutting edge. A Nd:YAG infrared laser (λ =
1.064 μm) was used in these experiments. They also claimed that the temperature of
the cutting zone was between 500 and 600 °C. Interestingly, the temperature range lay
dangerously close to the oxidation temperature of diamond at 627 °C under normal
ambient air conditions [67]. Overshadowed by the catastrophic diamond tool wear
problem when machining ferrous metals, oxidative wear of diamond cutting tools is
often neglected. In fact, oxidation is one of the chemo-mechanical methods employed
in diamond polishing [68]. Graphitisation, the diamond-graphite thermodynamic
transformation process, is another critical chemical wear process of diamond that
can be observed at 727 °C and accelerates even further at higher temperatures [69].
While it can be assumed that the diamond absorbs only ~30% of the laser irradiation
[55], it is without a doubt that thermal transfer will occur at the tool-workpiece
interface that may initiate the aforementioned chemical wear processes of diamond
cutting tools. Extra considerations are necessary to identify the limits to which the
μ-LAM technique can be effective.
Brittle ceramics that undergo phase transitions require high temperatures for effec-
tive transformation but can be reduced under applied pressures [55]. This could poten-
tially reduce the need for higher heating temperatures and still achieve the desired
material softening effects. The increase in machining pressures can be reached by
increasing in uncut chip area or the further increase in negativity of the rake face
to induce further compressive stresses. Yan et al. [41] introduced a simple model
to determine the pressures induced during micromachining of silicon based on the
274 12 Thermally Assisted Ductile Mode Cutting
cutting tool geometry and the equivalent tool-workpiece contact area. The tool-chip
contact and contact between the workpiece and the flank face were neglected in the
model with the main focus solely on the material ahead of the cutting tool. Consider-
ing a cutting process with depth of cut d, tool rake angle γ , tool edge radius Re , and
tool nose radius Rn , the pressure acting on the subsurface can then be derived from
the effective tool-workpiece contact area denoted by the blue line in Fig. 12.18. Two
cross-sections of the effective contact area can be modelled with one intersecting
the uncut surface and the other intersecting the onset of the machined surface and
parallel to the rake face as shown in Fig. 12.18. The individual length parameters
can be defined as follows [41]:
Re
h1 = (1 − sin γ ) (12.6)
cos γ
2
a1 = 2Rn cos γ d − d 2 (12.7)
cos γ
2
b1 = 2Rn cos γ (d − h 1 cos γ ) − (d − h 1 cos γ )2 (12.8)
cos γ
d
h2 = (12.9)
cos γ
Fig. 12.18 Schematic of the tool-workpiece contact geometry and the cross-sections of the effective
contact areas. Reprinted from [41] with permission from Elsevier
12.3 Micro-Laser Assisted Machining 275
1
S1 = h 1 b1 + π h 1 (a1 − b1 ) (12.10)
4
−1 a1 a1
S2 = Rn sin
2
− (Rn − h 2 ) (12.11)
2Rn 2
π
Seq = S1 cos φ + S2 cos −γ −φ (12.12)
2
F
p= (12.13)
Seq
Accordingly, the machining pressure increases with negative rake angles to induce
larger machining pressures and evoke the corresponding phase transformations of
silicon. Also, the increase in cutting depth naturally increases the effective tool-
workpiece contact area which results in an overall increase in machining pressure.
Therefore, it is intuitive to understand that there are measures that can be taken to
reduce the effective work-material heating temperatures and lower the tendencies
for chemical wear of the diamond cutting tools. Till date, there is yet to be a clear
understanding on the potential effects of changing the cutting conditions on tool
wear during micro-laser assisted machining. Mohammadi et al. [56] pointed out
that the use of a highly negative rake angle had negative effects on the machined
surface roughness (a comparison between −25° and −45°) attributing the result
to the inhibition of chip flow. Further parametric experimentation revealed a more
favourable combination with rake angle and the laser output power considering that
even lower heat will be required with the additional machining pressures induced by
the negative rake angle. As it stands in conventional diamond machining of silicon,
highly negative rake angles are normally used ranging between −30° and −50° to
achieve the highest critical uncut chip thickness [70].
Formation of secondary compounds on the diamond surface during the occurrence
of chemical wear will most assuredly influence the optical transmission of the tool.
But the optical properties of diamond alone lower the laser transmission efficiency,
resulting in heavy power losses. With the use of the diode laser (λ = 1.48 μm) rated
with an output power of 350 mW, the resultant power output at the diamond tip was
reported to be 140 mW (57% loss) due to absorbance and reflection in the diamond
tool [58, 71]. These losses are commonly associated with the optical properties such
as transmissivity, absorption, reflectance and the refractive index of the transmission
medium. Natural diamond is a favourable material choice for laser windows due to
its low adsorption coefficient (~0.05 cm−1 ) and high thermal conductivity [72], but
has a transmittance rate between 60% and 70% over a range of wavelengths from
225 to 2500 nm. In general, the high refractive index (n) of diamond decreases with
increase in wavelength from 2.729 (λ = 225 nm) to 2.383 (λ = 2000 nm) [73]. A good
approximation on the refractive index of diamond can be calculated by Eq. 12.14
[74]:
276 12 Thermally Assisted Ductile Mode Cutting
0.3306λ2 4.3356λ2
n2 = 1 + + (12.14)
λ2 λ2 − (106)2
Although the refractive index does not deviate much across natural diamonds with
different nitrogen impurity concentrations (e.g. type I and type II), different grades
of synthetic diamonds possess different optical properties such that the stabilisation
of structural lattice defects by annealing would reduce n and the ion implantation of
natural diamonds would exponentially increase n [73]. Recently, the development
of doped-cutting tools has been reported to reduce the friction during machining
and consequently the subsurface damage of the optical material [75]. The positive
outcomes of the two innovative techniques may attract initiatives to integrate both
systems to reap the reported benefits. However, caution must be made with recal-
ibrating the resultant power output at the tool cutting edge for effective control of
thermally-assisted system. Temperature also plays a significant role in the refractive
index that may not be affected much at relatively low temperatures below 340 K but
follows a quadratic increase in n at higher temperatures (>400 K) [76].
To ensure proper control of the amount of heat supplied to the cutting region, a
detector and feedback system must be instituted for precise real-time measurements
and optimisation of the cutting process. Negligence of the temperature control could
result in undesirable mechanical effects, phase-changes and even physio-chemical
interactions [37]. Interestingly, the discussion on physio-chemical reactions during
micromachining, aside from thermally enhanced machining, has recently resurfaced
in micromachining of copper and aluminium under the influence of a surface-active
medium [77, 78]. Hence, the temperature effects that accelerate chemical reactions or
remove surface moisture from the work-material must be well-thought-out, begin-
ning with the temperature ranges attained during thermally enhanced machining.
Surprisingly, most research work using laser-assisted machining techniques rarely
measures the temperature at the cutting zones but simply presents the influence of
machining conditions and laser beam parameters on the resultant machined sur-
faces of various materials and material removal processes (i.e. turning, drilling, and
milling).
A laser pyrometer [79, 80] was constructed to obtain surface temperature mea-
surements during laser-assisted machining of silicon nitride ceramics. This required
the precise alignment of the CO2 laser heating spot and the laser pyrometer along the
same circumferential location along the work-material during a cylindrical turning
process. Figure 12.19 illustrates the concept of circumferential temperature measure-
ment at the vicinity of the cutting tool [32]. Infrared thermal imaging is also often
used to determine the temperature of the cutting regions [81] and the heating zone
[28, 53]. Figure 12.20 shows the thermal imaging of the globalised heating method
for micro-cutting of brittle materials realised by electrical heating. Proper calibra-
tion of the emissivity and reflectance values to ensure measurement accuracies. The
12.4 Temperature Measurement and Control 277
Fig. 12.19 Schematic of an in-process temperature measurement with a pyrometer. Reprinted from
[32] with permission from Elsevier
Fig. 12.20 Infrared thermal imaging of the micro-cutting experimental setup for globalized heating
[28]
reflectance may be measured using an FT-IR spectrometer and the spectral emissivity
may be determined from the reflectance measurements [82]. It is important to note
that the explained temperature measurements were conducted on the macroscopic
scale where cutting depths ranged in millimetres. To date, the in-situ measurement
of temperature in the microscopic cutting regions remains a challenging task and
needs further experimental justifications.
278 12 Thermally Assisted Ductile Mode Cutting
• Heat supply
• Heat distribution
• Thermal gradients between heating zones and surrounding components.
Heat supply is considerably consistent with respect to the localised heating tech-
niques in terms of power intensity and power supply with reference to laser-assisted
machining. However, the heat distribution requires refinement according to the
machining conditions such as the feed and the Gaussian profile of the laser beam as
mentioned in Sects. 12.3.2 and 12.1, respectively. Temperature distribution over a
globalised heating setup can be fabricated with proper control of the energy supply
and positioning of heat sources. Thermal gradients between heating zones and the
surrounding systems are critical in terms of heat transfer rates that would eventually
influence the effective material geometric form (i.e. a higher temperature difference
would promote heat transfer rates that could influence the material expansion and
contraction). For example, the temperature difference between the heated workpiece
and the tool holder would result in higher heat transfer by convection and radiation
that would influence the resultant temperature distribution of the workpiece, even
under an ideal situation of an evenly distributed heat source. The differences in tem-
perature across the material surface would change its effective distance relative to
the cutting tool (based on thermal expansion as shown in Fig. 12.21) and ultimately
the produced form accuracy after machining. An ideal incorporation of events dur-
ing thermally-assisted micromachining would account for all aspects of heat transfer
and the in-situ adjustment of the tool-workpiece relative contact according to the
instantaneous temperature measured that would define its expanded form.
Till date, temperature control measures are only in its infancy stages on providing
insulation between the heating regions and the integrated precision machining cen-
tre. The simplest method is to simply include a layer of insulating material such as
polytetrafluoroethylene (PTFE), further known as Teflon, at the workpiece-machine
interface. The polymer is a relatively soft material with excellent heat insulation prop-
erties that can be easily machined to form the workpiece holder. However, material
Fig. 12.21 Schematic of a the ideal thermal-assisted machining heat transfer and b the actual event
accounting for thermal expansion and tool heating
280 12 Thermally Assisted Ductile Mode Cutting
flexing tends to occur after repeated cycles of heating and cooling that results in the
material expansion and contraction. Ceramics can also provide good thermal insula-
tion but are difficult to machine as a brittle material. Thermal barrier coatings that are
commonly used in the aerospace, nuclear reactor and marine automobile industries,
are also a method of insulation to be considered [83].
Machine tool components have also been reported to undergo thermal expansion,
which leads to the conceptualisation of a cooling stages and cooling tool holders.
Gavrunov and Ravindra [84] fabricated a cooling workpiece holder platform made
of aluminium alloy 6061 to act as an insulation layer between the heated workpiece
and the lathing spindle. 42 internal cooling cylindrical channels of varying diame-
ters were designed along the circumference of the stage to “scoop” ambient air as
the cooling fluid during the revolving motion of the workpiece, such that the stage
becomes more effective at higher spindle speeds. Compressed air supply could also
be directed toward the rotating cooling channels for more effective supply of the
coolant and the air-cooled staged was claimed to maintain temperatures below 30 °C
under the influence of a 50-W laser. Chaudhari et al. [85] identified the issue of heat
transfer to the tool holder by radiation and conduction that could lead to errors in
the tool-workpiece contact. A single component tool holder made of stainless steel
was designed and fabricated with the selective laser sintering (SLS) additive man-
ufacturing process to include intricate features such as the tool-tip cooling section
that is 1 mm thick and 4 mm tall as shown in Fig. 12.22. Internal channels enabled
Fig. 12.22 Schematic of the tool holder with internal cooling features [85]
12.4 Temperature Measurement and Control 281
Fig. 12.23 Numerical model simulation of the temperature distribution on the tool holder and
the experimental infrared thermal imaging measurement: a without applying internal cooling and
b with internal cooling [85]
the constant flow of cooling fluids to mitigate the adverse heat transfer from the
thermally-assisted machining process by conduction. Simple fluids such as water
are sufficient at discouraging heat transfer to the bulk of the tool holder as shown in
Fig. 12.23. Finite element method (FEM) modelling was performed using the com-
mercial software, ANSYS Workbench 16.2, where the cutting tip and the internal
cooling channel was defined at 100 °C and 27 °C, respectively, to simulate the supply
of cooling fluids.
process has not yet been realised. These additional researches include the refinement
of the process parameters to account for diamond tool wear rates at elevated temper-
atures and the variation of laser beam size during machining. Temperature control
is a particularly important factor that is often neglected and is undeniably important
regarding material expansion and contraction under thermal influence for the work-
piece, cutting tool and workpiece holders. There is a need for auxiliary technologies
to ensure precise temperature control of the machining environment to enjoy the full
benefits of thermally-assisted machining at the micrometric and nanometric scales.
References
20. López de Lacalle LN, Sánchez JA, Lamikiz A, Celaya A (2004) Plasma assisted milling of
heat-resistant superalloys. J Manuf Sci Eng 126:274
21. Özler L, İnan A, Özel C (2001) Theoretical and experimental determination of tool life in hot
machining of austenitic manganese steel. Int J Mach Tools Manuf 41:163–172
22. Jeon Y, Lee CM (2012) Current research trend on laser assisted machining. Int J Precis Eng
Manuf 13:311–317
23. Novak JW, Shin YC, Incropera FP (1997) Assessment of plasma enhanced machining for
improved machinability of Inconel 718. J Manuf Sci Eng 119:125–129
24. Leshock CE, Kim JN, Shin YC (2001) Plasma enhanced machining of Inconel 718: modeling
of workpiece temperature with plasma heating and experimental results. Int J Mach Tools
Manuf 41:877–897
25. Madhavulu G, Ahmed B (1994) Hot machining process for improved metal removal rates in
turning operations. J Mater Process Technol 44:199–206
26. Chen SH, Tsai KT (2017) The study of plasma-assisted machining to Inconel-718. Adv Mech
Eng 9:1–7. https://doi.org/10.1177/1687814017735789
27. Kttagawa T, Maekawa K (1990) Plasma got machining for new engineering materials. Wear
139:251–267
28. Lee YJ (2019) Thermal expansion control in heat assisted machining of calcium fluoride single
crystals. In: 19th international conference of the European Society for Precision Engineering
and Nanotechnology (euspen)
29. Lee YJ, Chaudhari A, Zhang J, Wang H (2019) Thermally assisted microcutting of calcium
fluoride single crystals. In: Simulation and experiments of material-oriented ultra-precision
machining. Springer, Singapore, pp 77–127
30. Wang H, Senthil Kumar A, Riemer O (2018) On the theoretical foundation for the microcutting
of calcium fluoride single crystals at elevated temperatures. Proc Inst Mech Eng Part B J Eng
Manuf 232:1123–1129
31. Lei S, Shin YC, Incropera FP (2001) Experimental investigation of thermo-mechanical char-
acteristics in laser-assisted machining of silicon nitride ceramics. J Manuf Sci Eng 123:639
32. Anderson M, Patwa R, Shin YC (2006) Laser-assisted machining of Inconel 718 with an
economic analysis. Int J Mach Tools Manuf 46:1879–1891
33. Shayan AR, Poyraz HB, Ravindra D, Patten JA (2009) Pressure and temperature effects in
micro-laser assisted machining (μ-LAM) of silicon carbide. Trans North Am Manuf Res Inst
SME 37:75–80
34. Shayan AR, Poyraz HB, Ravindra D, Ghantasala M, Patten JA (2009) Force analysis, mechan-
ical energy and laser heating evaluation of scratch tests on silicon carbide (4H-SiC) in micro-
laser assisted machining (μ-LAM) process. In: ASME 2009 international manufacturing
science and engineering conference. American Society of Mechanical Engineers, pp 827–832
35. Cline HE, Anthony TR (1977) Heat treating and melting material with a scanning laser or
electron beam. J Appl Phys 48:3895–3900
36. Spalding IJ (1974) Lasers—their applications and operational requirements. Opt Laser Technol
6:263–272
37. Chryssolouris G, Anifantis N, Karagiannis S (1997) Laser assisted machining: an overview.
Trans ASME 119:766–769
38. Tomizawa H, Fischer TE (1987) Friction and wear of silicon nitride and silicon carbide in
water: hydrodynamic lubrication at low sliding speed obtained by tribochemical wear. A S L
E Trans 30:41–46
39. Hibi Y, Enomoto Y, Kikuchi K, Shikata N, Ogiso H (1995) Excimer laser assisted chemical
machining of SiC ceramic. Appl Phys Lett 66:817–818
40. Chavoshi SZ, Xu S (2018) Temperature-dependent nanoindentation response of materials.
MRS Commun 8:15–28
41. Yan J, Asami T, Harada H, Kuriyagawa T (2009) Fundamental investigation of subsurface
damage in single crystalline silicon caused by diamond machining. Precis Eng 33:378–386
42. Domnich V, Aratyn Y, Kriven WM, Gogotsi Y (2008) Temperature dependence of silicon
hardness: experimental evidence of phase transformations. Rev Adv Mater Sci 17:33–41
284 12 Thermally Assisted Ductile Mode Cutting
43. Khayyat MMO, Hasko DG, Chaudhri MM (2007) Effect of sample temperature on the
indentation-induced phase transitions in crystalline silicon. J Appl Phys 101:83515
44. Khayyat MM, Banini GK, Hasko DG, Chaudhri MM (2003) Raman microscopy investigations
of structural phase transformations in crystalline and amorphous silicon due to indentation with
a Vickers diamond at room temperature and at 77 K. J Phys D Appl Phys 36:1300–1307
45. Ravindra D, Ghantasala MK, Patten J (2012) Ductile mode material removal and high-pressure
phase transformation in silicon during micro-laser assisted machining. Precis Eng 36:364–367
46. Yan J, Maekawa K, Tamaki J, Kuriyagawa T (2005) Micro grooving on single-crystal germa-
nium for infrared Fresnel lenses. J Micromech Microeng 15:1925–1931
47. Clarke DR, Kroll MC, Kirchner PD, Cook RF, Hockey BJ (1988) Amorphization and conduc-
tivity of silicon and germanium induced by indentation. Phys Rev Lett 60:2156–2159
48. Patten JA, Jacob J (2008) Comparison between numerical simulations and experiments for
single-point diamond turning of single-crystal silicon carbide. J Manuf Process 10:28–33
49. Patten J, Fesperman R, Kumar S, McSpadden S, Qu J, Lance M, Nemanich R, Huening J (2003)
High-pressure phase transformation of silicon nitride. Appl Phys Lett 83:4740–4742
50. Wang H, Riemer O, Rickens K, Brinksmeier E (2016) On the mechanism of asymmetric
ductile–brittle transition in microcutting of (111) CaF2 single crystals. Scr Mater 114:21–26
51. Mizumoto Y, Kangawa H, Itobe H, Tanabe T, Kakinuma Y (2017) Influence of crystal
anisotropy on subsurface damage in ultra-precision cylindrical turning of CaF2 . Precis Eng
49:104–114
52. Muñoz A, Domínguez-Rodríguez A, Castaing J (1994) Slip systems and plastic anisotropy in
CaF2 . J Mater Sci 29:6207–6211
53. Chaudhari A, Lee YJ, Wang H, Kumar AS (2017) Thermal effect on brittle–ductile transition
in CaF2 single crystals. In: 17th international conference of the European Society for Precision
Engineering and Nanotechnology (euspen)
54. Shahinian H, Navare J, Zaytsev D, Kode S, Azimi F (2018) Effect of laser-assisted diamond
turning (Micro-LAM) on form and finish of selected IR crystals. In: 33rd annual meeting of
American Society for Precision Engineering, Las Vegas, Nevada, pp 1–5
55. Shahinian H, Navare J, Zaytsev D, Ravindra D (2019) Microlaser assisted diamond turning of
precision silicon optics. Opt Eng 58:1–8
56. Mohammadi H, Ravindra D, Kode SK, Patten JA (2015) Experimental work on micro laser-
assisted diamond turning of silicon (111). J Manuf Process 19:125–128
57. Ravindra D, Kode SK, Stroshine C, Morrison D, Mitchell M (2017) Micro-laser assisted
machining: the future of manufacturing silicon optics. In: Proceedings of the SPIE
58. Suthar K, Patten JA, Dong L, Abdel-aal H (2008) Estimation of temperature distribution in sili-
con during micro laser assisted machining. In: ASME 2008 international manufacturing science
and engineering conference collocated with the 3rd JSME/ASME international conference on
materials and processing, Evanston, Illinois, USA
59. Patten J, Ghantasala M, Shayan AR, Bogac Poyraz H, Ravindra D (2009) Micro-laser assisted
machining (μ-LAM): scratch tests on 4H-SiC. In: Proceedings of 2009 NSF engineering
research and innovation conference, Honolulu, Hawaii
60. Leung TP, Lee WB, Lu XM (1998) Diamond turning of silicon substrates in ductile-regime.
J Mater Process Technol 73:42–48
61. Yan J, Yoshino M, Kuriagawa T, Shirakashi T, Syoji K, Komanduri R (2001) On the duc-
tile machining of silicon for micro electro-mechanical systems (MEMS), opto-electronic and
optical applications. Mater Sci Eng A 297:230–234
62. Ravindra D, Poyraz HB, Patten J (2010) The effect of laser heating on the ductile to brittle tran-
sition of silicon. In: The 5th international conference in micromanufacturing (ICOMM/4M),
Wisconsin, USA
63. Ravindra D, Poyraz HB, Patten J (2010) The effect of laser heating on the ductile to brittle
transition of silicon carbide. Manufacturing Research Center, Western Michigan University
64. Yan J, Tamaki J, Syoji K, Kuriyagawa T (2004) Single-point diamond turning of CaF2 for
nanometric surface. Int J Adv Manuf Technol 24:640–646
References 285
65. Ravindra D, Ponthapalli SC, Patten J (2013) Micro-laser assisted single point diamond turning
feasibility tests of single crystal silicon. In: American Society for Precision Engineering (ASPE)
28th annual meeting, St. Paul, Minnesota
66. Mohammadi H, Patten JA (2016) Laser augmented diamond drilling: a new technique to drill
hard and brittle materials. Procedia Manuf 5:1337–1347
67. Shimada S, Tanaka H, Higuchi M, Yamaguchi T, Honda S, Obata K (2004) Thermo-chemical
wear mechanism of diamond tool in machining of ferrous metals. CIRP Ann Manuf Technol
53:57–60
68. Schuelke T, Grotjohn TA (2013) Diamond polishing. Diam Relat Mater 32:17–26
69. Fedoseev DV, Vnukov SP, Bukhovets VL, Anikin BA (1986) Surface graphitization of diamond
at high temperatures. Surf Coatings Technol 28:207–214
70. Yan J, Syoji K, Kuriyagawa T (2000) Ductile–brittle transition at large negative tool rake angles.
J Jpn Soc Precis Eng 66:1130–1134
71. Patten JA, Shayan AR, Poyraz HB, Ravindra D, Ghantasala M (2009) Scratch tests on 4H-
SiC using micro laser assisted machining (μ-LAM) system. In: Advance laser applications
conference and exposition, San Jose, California, USA
72. Douglas-Hamilton DH, Hoag ED, Seitz JRM (1974) Diamond as a high-power-laser window.
J Opt Soc Am 64:36–38
73. Zaitsev AM (2001) Optical properties of diamond. Springer, Berlin, Heidelberg
74. Collins AT (1993) Intrinsic and extrinsic absorption and luminescence in diamond. Phys B
Condens Matter 185:284–296
75. Mizumoto Y, Amano H, Kangawa H, Harano K, Sumiya H, Kakinuma Y (2018) On the
improvement of subsurface quality of CaF2 single crystal machined by boron-doped nano-
polycrystalline diamond tools. Precis Eng 52:73–83
76. Fontanella J, Johnston RL, Colwell JH, Andeen C (1977) Temperature and pressure variation
of the refractive index of diamond. Appl Opt 16:2949–2951
77. Chaudhari A, Soh ZY, Wang H, Kumar AS (2018) Rehbinder effect in ultraprecision machining
of ductile materials. Int J Mach Tools Manuf 133:47–60
78. Chaudhari A, Wang H (2019) Effect of surface-active media on chip formation in microma-
chining. J Mater Process Technol 271:325–335
79. Rozzi JC, Pfefferkorn FE, Shin YC, Incropera FP (2000) Experimental evaluation of the laser
assisted machining of silicon nitride ceramics. Trans ASME 122:666–670
80. Pfefferkorn FE (1997) Laser pyrometry: non-intrusive temperature measurement for laser-
assisted machining of ceramics. Purdue University
81. Dinc C, Lazoglu I, Serpenguzel A (2008) Analysis of thermal fields in orthogonal machining
with infrared imaging. J Mater Process Technol 198:147–154
82. Davies MA, Yoon H, Schmitz TL, Burns TJ, Kennedy MD (2003) Calibrated thermal
microscopy of the tool-chip interface in machining. Mach Sci Technol 7:167–190
83. Xiong HP, Kawasaki A, Kang YS, Watanabe R (2005) Experimental study on heat insulation
performance of functionally graded metal/ceramic coatings and their fracture behavior at high
surface temperatures. Surf Coatings Technol 194:203–214
84. Gavrunov A, Ravindra D (2013) Custom cooling stage to eliminate thermal expansion of
spindle during micro-laser assisted machining. In: Advance laser applications conference and
exposition, Chicago, Illinois, USA
85. Chaudhari A, Antwi EK, Wang H, Liu K, Kumar AS, Tu J, Zhang X (2017) Tool design and
implementation for thermally assisted ultraprecision diamond turning. In: The 7th international
conference of Asian Society for Precision Engineering and Nanotechnology, Seoul, Korea
Chapter 13
Summary
In this book, ductile mode cutting of brittle material is presented and discussed
systematically in terms of fundamentals, engineering applications and hybrid ductile
mode cutting techniques, which is summarized as the following aspects:
• Ductile chip formation mechanism in cutting of brittle material is developed and
analysed theoretically, which is the competition between dislocation emissions
and crack propagations caused by cutting stresses in chip formation zone based
upon an analysis of cutting geometry and cutting force, fracture mechanics and
yield strength enhancement. Ductile mode cutting of brittle material is a result of
large compressive stress and shear stress in cutting zone, which shields the growth
of pre-existing flaws in work material by suppressing its stress intensity factor K I .
It is also a result of enhancement of material’s yielding strength in chip formation
zone, which in turn, directly enhances material’s fracture toughness K C .
• Large compressive stress and shear stress in the cutting zone are achieved by
having two conditions satisfied in cutting of brittle material:
1. The first condition is to have a very small undeformed chip thickness, such that
compressive stress in cutting zone is large enough to suppress stress intensity
factor K I , resulting in K I being smaller than fracture toughness K C .
2. The second condition is to have the ratio of tool cutting edge radius to unde-
formed chip thickness being larger than 1, such that material’s yielding strength
in cutting zone is enhanced by dislocation hardening and strain gradient at
mesoscale, resulting in an enhanced material fracture toughness.
These conditions are established by having a micrometer or nanometer scale
undeformed chip thickness in cutting of brittle material.
• Ductile mode cutting of brittle material is analysed theoretically in relation to its
temperature, elastic modulus, hardness and fracture toughness in cutting zone,
which indicates that in cutting of brittle material there is a transition between
ductile chip formation and brittle chip formation. The ductile-to-brittle transition
is dependent on the tool geometry, work material and cutting conditions used.
An energy model for ductile mode cutting of brittle material is developed using
work material fracture toughness and mechanical properties, of which temperature-
dependent hardness is estimated with the temperature rise in cutting region and
microstructural parameters of brittle material. Critical undeformed chip thickness
in ductile mode cutting of brittle material or critical depth of cut in grooving
of brittle material is predicted based on work material properties, cutting tool
geometry and cutting conditions used.
• Theoretical simulations of ductile mode cutting of brittle material are realised with
crystal plasticity finite element methods and molecular dynamics. The simulation
methods encompass the modelling length scales from nanometers to micrometres,
integrating the orientation-dependent effects of an anisotropic single crystal brit-
tle material. Chip flow and tool wear simulations are produced with the optimal
interatomic potentials, which determine the material plasticity and tool-workpiece
interaction. The Tersoff, Vashishta and Buckingham interatomic potentials are
most commonly used in modelling the material deformation process of brittle
materials.
• Experimental investigation using conventional grooving of tungsten carbide as
an example to verify the model for predicting critical undeformed chip thickness,
which shows a substantial agreement between the predicted value and experimental
results. There is a transition from ductile mode cutting to brittle mode cutting in
grooving of tungsten carbide when the depth of cut is increased. Ductile mode
cutting occurs in grooving of tungsten carbide only when the depth of cut is smaller
than a critical value. Once the depth of cut is increased to exceed the critical value,
the cutting mode is changed from ductile mode cutting to brittle mode cutting.
Similar experimental results for grooving of soda-lime glass, ZKN7 glass, fused
silica glass, BK7 glass and single crystalline silicon are achieved as well. Three
cutting modes are generated in grooving of tungsten carbide when the depth of cut
is increased, said ductile mode cutting, semi-brittle mode cutting and brittle mode
cutting. Crack propagation is the main reason to cause brittle fracture of brittle
material during the grooving process. Compressive overloads are likely to have a
more dominant effect on crack propagations.
• In ductile mode cutting of tungsten carbide, thrust force F t is much larger than
cutting force F c , which results in a large compressive stress in cutting zone. Large
compressive stress and shear stress could shield the growth of pre-existing flaws
in work material by suppressing its stress intensity factor K I , such that K I < K C
making work material is able to undertake a large cutting stress without fracturing
to achieve ductile mode cutting. In general, smooth surface generated and con-
tinuous chips formed are used to verify its material removal mode in cutting of
brittle material being ductile mode cutting. While fractured surface generated and
particle chips formed are used to verify its material removal in cutting of brittle
material being brittle mode cutting.
• Nanometric cutting of silicon wafers using an ultra-precision CNC lathe with
single crystal diamond cutters are conducted to investigate tool edge radius effect
on critical undeformed chip thickness and verify ductile mode cutting performance
of silicon wafer. Critical undeformed chip thickness differs when cutting of silicon
13 Summary 289
wafers using diamond cutters with different tool edge radius. Larger diamond
tool edge radius, larger critical undeformed chip thickness. But there is an upper
bound for diamond tool edge radius, above which chip formation is changed from
ductile mode to brittle mode even though undeformed chip thickness remains to be
smaller than tool edge radius. Experimental results are found to well substantiate
the analytical findings and nanometric ductile mode cutting of silicon wafer is
successfully achieved under certain cutting conditions.
• Nanometric cutting of tungsten carbide, soda-lime glass and single crystal silicon
wafer are carried out to evaluate their cutting performance under nanometer scale
undeformed chip thickness. Experimental results indicate that ductile mode cutting
is achieved when undeformed chip thickness being less than a critical value for
all above-mentioned materials. Examinations on machined workpiece surfaces
and chip formation indicate that ductile mode cutting is mainly determined by
undeformed chip thickness when tool cutting edge radius is fixed. Under different
cutting conditions three types of surfaces of machined workpiece are achieved:
ductile mode cutting surface, semi fractured surface and fractured surface. Exam-
inations on cutting tools indicate that tool wear mainly occurs on flank face and
tool wear mechanisms are dominated by diffusion, adhesion and abrasion wear in
cutting of brittle material.
• Ductile mode cutting of calcium fluoride (CaF2 ) single crystals is achieved with the
consideration of several cutting conditions and an evaluation of the large variation
in ductile-to-brittle transition along different cutting directions with respect to each
crystallographic plane. To date, the effects of cutting speed on the material response
have yet to reach a consensus but a largely negative rake angle would hinder the
ductile-to-brittle transition. The optimal rake angle to achieve the highest critical
uncut chip thickness is determined to be −15°. Cutting fluids also promote ductile
mode cutting. While a doped cutting tool has shown some minor improvements
in reducing the subsurface damage after micro-cutting, the elliptical vibration-
assisted machining is deemed one of the more promising techniques to improve
the machinability of CaF2 .
• Hybrid cutting techniques have been proposed to enhance the cutting performance
in terms of improving surface integrity, prolonging tool life and increasing critical
undeformed chip thickness for ductile-to-brittle transition in the cutting of brittle
material. Commonly, the development of hybrid ductile mode cutting techniques
are mainly focused on 1D/2D ultrasonic vibration assisted cutting, and thermally
assisted cutting with a general classification of globalised heating systems and
localised heating systems.
• An analytical model is developed to predict critical undeformed chip thickness in
ultrasonic vibration assisted cutting of brittle material, based on the variation of
specific cutting energy for prediction of ductile-to-brittle transition. Both predicted
and experimental results show that critical undeformed chip thickness nonlinearly
increases with the decrease of nominal cutting speed, and a carefully chosen nom-
inal cutting speed leads to a significantly larger critical undeformed chip thickness
compared to that in the conventional cutting of brittle material.
290 13 Summary
• Critical depth of cut for the transition from ductile mode cutting to brittle mode
cutting in 1D ultrasonic vibration assisted grooving of tungsten carbide and soda-
lime glass is several times larger than that in conventional grooving. It well agrees
with the theoretical analysis of the ductile mode cutting performance of brittle
material with ultrasonic vibration assistance. Comparing with material removal at
ductile-to-brittle transition in conventional grooving, material removal ratio in 1D
ultrasonic vibration assisted grooving is more close to 1, indicating a reduction
on ploughing effect by ultrasonic vibration. And the larger critical depth of cut
for 1D ultrasonic vibration assisted grooving implies that ultrasonic vibration
could be used to improve ductile mode cutting performance of brittle material.
Lower thrust directional amplitude in 2D ultrasonic vibration leads to less brittle
fracture generated on the machined surface of tungsten carbide, and 1D ultrasonic
vibration with no thrust directional vibration leads to minimum brittle fracture
and less diamond tool wear. Increasing tool nose radius and reducing nominal
cutting speed leads to less diamond tool wear and suppress the catastrophic crack
of cutting edges in face turning of tungsten carbide. Nano-polycrystalline diamond
with isotropic mechanical properties does not perform better than single crystal
diamond as tool material in terms of tool flank wear in ultrasonic vibration assisted
turning of tungsten carbide.
• For thermally assisted cutting of brittle material, two phenomena are reported on
top of fundamental material softening as a result of thermal assistance. The phase
transition of silicon occurs to effectively form an amorphous layer during cutting.
In calcium fluoride single crystals, the elevated temperatures activate secondary
slip systems to encourage dislocation slip and plastic deformation in place of brit-
tle failure by cleavage. Micro-laser assisted cutting is the most advanced thermal
assistance technology to date to promote the ductile mode cutting of brittle mate-
rial. The successful implementation of the technique is governed by the integration
of laser beam diameter, laser power, and effective cutting speed and feed, as well as
material’s thermal properties. The benefits of the system include reduced cutting
forces and enhanced resistance to tool wear.
All brittle materials would experience a ductile-to-brittle transition when cutting
with an increasing depth of cut from zero. But, the critical undeformed chip thick-
ness value obtained from both theoretical prediction and experimental results for
ductile mode cutting of brittle material is very small, at micron, sub-micron or even
nanometer level, which largely constrains their actual industry application. Natu-
rally, questions on how to overcome this problem are surfacing up. Noted that the
key parameter is critical undeformed chip thickness as ductile mode cutting can be
achieved on brittle materials when cutting below the critical value. In the practical
application aspect, the value of critical undeformed chip thickness is expected to
be as large as possible. Thus, are we able to increase the critical undeformed chip
thickness value for a given material? And to what extent?
The effect of tool sharpness (or called cutting edge radius) on ductile mode cut-
ting has been studied using different diamond tool edge radii to some extent. The
undeformed chip thickness for ductile mode cutting of silicon has not to be larger
13 Summary 291
than its cutting tool edge radius. Increasing cutting tool edge radius could derive a
larger critical undeformed chip thickness. But there is an upper bond of cutting edge
radius for ductile mode cutting of silicon wafers. When cutting of silicon using a tool
edge radius beyond the limit, ductile mode cutting would not be achievable. Likely,
the compressive stress in the cutting region decreases with increasing of cutting edge
radius. As such, any way to increase compressive stress in the cutting zone?
Extensive experimental studies have proved that in ductile mode cutting of brit-
tle material, there are three key apparent features: Continuous chips formed, smooth
and crack-free surface, and no subsurface damage obtained. There are also two unap-
parent features derived from ductile mode cutting: Residual stress and subsurface
microstructure, which most likely would change the materials’ mechanical, optical,
physical and chemical properties. Thereafter, it would largely limit their promising
industrial applications. But the research on residual stress obtained in ductile mode
cutting of brittle materials is not sufficient and need to be further explored.
Although some models and cutting mechanisms have been developed to illustrate
the brittle-to-ductile transition in cutting of brittle material, ductile mode cutting
mechanism needs to be broadly investigated through extensive theoretical, experi-
mental and simulation studies to have a comprehensive understanding of the chip for-
mation and damage-free surface generation. Hybrid manufacturing and/or machin-
ing processes have evidenced the somehow improvement in ductile mode cutting
of brittle material. Novel and breakthrough technologies on hybrid manufactur-
ing/machining processes need to be innovated and developed to largely improve
ductile mode cutting performance and brittle material’s machinability. Also, the
research on ductile mode cutting should be extended to more advanced and new
emerging brittle material. It will help to eliminate manufacturing barriers effectively
and bloom industrial applications significantly.