e
1;... I; ,
2.! . 1);: 0.
IL. 1 P). r

.3,

1‘
1.3.34
. £130. .
12“.
t It...

I. :
1."..13.

 

32.7. . ‘ . .
..:n,_, V . «#32:...
. 2.5.... e...

V . . .1 .,.

 

 

 

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

A New Methodology for Predictive Tool Wear

presented by

Won-Sik Kim

has been accepted towards fulfillment
of the requirements for

Doctor of Philosophy

3 Major professor

Date JanuanLJLZflQl—

MS U is an Affirmative Action/Equal Opportunity Institution

degreein Materials Science and

Mechanics

0-12T71

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DatoDInfi-nifi

 

 

A New Methodology for Predictive Tool Wear
By

Won-Sik Kim

AN ABSTRACT OF A DISSERTATION

Submitted to Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Materials Science and Mechanics

2001

Professor Patrick Kwon

ABSTRACT

A New Methodology for Predictive Tool Wear
By Won-Sik Kim

An empirical approach to tool wear, which requires a series of machining tests for
each combination of insert and work material, has been a standard practice for industries
since early part of the twentieth century. With many varieties of inserts and work
materials available for machining, the empirical approach is too experiment-intensive that
the demand for the development of a model-based approach is increasing. With a model-
based approach, the developed wear equation can be extended without additional
machining experiments. The main idea is that the temperatures on the primary wear
areas are increasing such that the physical properties of the tool material degrade
substantially and consequently tool wear increases. Dissolution and abrasion are
identified to be the main mechanisms for tool wear. Flank wear is predominantly a
phenomenon of abrasion as evident by the presence of a scoring mark on the flank
surface. Based on this statement, it is reasonable to expect that the flank-wear rate would
increase with the content of hard inclusions. However, experimental flank wear results
did not necessary correspond to the content of cementite phase present in the steels.
Hence, other phenomena are believed to significantly affect wear behavior under certain
conditions. When the cutting temperature in the flank interface is subjected to high
enough temperatures, pearlitic structure austenizes. During the formation of a new
austenitic phase, the existing carbon is dissolved into the ferrite matrix, which will reduce

the abrasive action. To verify the austenitic transformation, turning tests were conducted

m’rh plai
(XRDI.

Microsc

wear. '
uncoatc‘
compar
hate b
additic
Wear.

tensile

hardne

inCl’e;
Wear!

crater

with plain carbon steels. The machined surface areas are imaged using X-ray diffraction
(XRD), the Scanning Electron Microscope (SEM) and the Transmission Electron
Microscope (TEM).

On the other hand, crater wear occurs as a result of dissolution wear and abrasive
wear. To verify the wear mechanisms of crater wear, various coating inserts as well as
uncoated inserts were turned with various cutting conditions and the results were
compared with the proposed analytical wear models. The crater surfaces afier machining
have been carefully studied to shed light on the physics behind the crater wear. In
addition, the abrasive wear mechanism plays a major role in the development of crater
wear. Laser shock processing (LSP) has been applied to locally relieve the deleterious
tensile residual stresses on the crater surface of a coated tool, thus to improve the
hardness of the coating. This thesis shows that LSP has indeed improve wear resistance
of CVD coated alumina tool inserts, which has residual stress due to high processing
temperature. LSP utilizes a very short laser pulse with high energy density, which
induces high-pressure stress wave propagation. The residual stresses are relieved by
incident shock waves on the coating surface. Residual stress levels of LSP CVD
alumina-coated carbide insert were evaluated by the X-ray diffractometer. Based on
these results, LSP parameters such as number of laser pulses and laser energy density can
be controlled to reduce residual stress. Crater wear shows that the wear resistance
increase with LSP treated tool inserts. Because the hardness data are used to predict the
wear, the improvement in hardness and wear resistance shows that the mechanism of

crater wear also involves abrasive wear.

Cow/right by
Won-Sik Kim
200 l

ACKNOWLEDGEMENTS

I wish to thank my dissertation advisor, Professor Patrick Kwon, for his
instruction and guidance in my doctorial studies, for the financial support in this course
of my reseach, and for his constructive suggestion and careful modification which
contribute greatly to the completion of this dissertation.

I should also extend my gratitude my dissertation committee members: Professor
Brian Feeny, Professor Dahsin, Liu, and Professor Stanley Flegler for their invaluable
advice and suggestion in the completion of this dissertation. My thanks also go to my
friend J. K. Park, T. Wong, 1. Do, and H. Shin for their fiiendship and help.

My sincere thanks are owed to my parents for their encouragement, financial
support, and unsparing love. My special thanks go to my wife, Jeongjoo Moon, who
always has encouraged and supports me, and who has been there at all times when help is

needed.

‘JJ

TABLE OF CONTENS

ABSTRACT

LIST OF TABLES

LIST OF FIGURES

1.

2.

INTRODUCTION

LITERATURE REVIEW OF METAL MACHINING
2-1 Physics of Metal Machining

2-2 Mechanics of metal cutting

2-3. Mechanism of tool wear

2-3-1 Characteristics of tool wear

2.3.2 Wear mechanisms

2.3.3 Coated Tool Wear

2.4 Cutting Temperature

2.4.1 Analytical methods for determining cutting
temperatures

2.4.2 Experimental methods for determining cutting
temperatures

FLANK WEAR BEHAVIOR IN TURNING PROCESS
3.1 . Introduction

3.2. Implication of the phase transformation on the
machined surface in other works

3.3 Experiments

3.3.1 Turning experiments

3.3.2 Wear measurement

3.4.Results and discussions

3.4.1 Tool wear measurement

3.4.2. Examination of microstructure after cutting test
3.4.3 Phase identification

3.4.4 Microstructure of machined area

3.4.5 Phase identification with X-ray diffraction method
and TEM analysis

3.5. Discussion of experimental data

UNDERSTANDING CRATER WEAR OF CUTTING
INSERTS

4.1. IntroductionExperimental crater wear results

4.2. Comparison of the relative crater wear

4.3. Experiments

4.3.1 Turning tests

vi

6

APPENDIX

4.3.2. Characterizations of Crater Wear

4.3.3. Temperature Measurement

4.3.4 Inverse Estimation Scheme for the Chip-Tool
Interface Temperature

4.4 Experimental crater wear results

4.5. Microstructural Analysis on Crater Area

SURFACE TREAMENT ON THE TOOL INSERTS
5.1 . Introduction

5.2. CVD alumina-coated carbide insert

5.3. Laser Shock Processing

5.4. Residual stress measurement

5.5. Crater Wear test

5.6. Discussion of results

CONCLUSIONS

A1. Abrasive model: two body and three body model
A.2 Temperature at the sliding interface

A.3 Computation of the free energy of formation of
TiCN

A3. 1. Balinit B coating of Balzers

A.3.2. Final Form for the Free energy of formation for
TiC(1-x)Nx

A.3.3. Molar Volume of TiCl-xNx

A.4. Estimation of the molar excess free energy

REFERENCES

vii

......... 71
......... 71
......... 74

......... 76
......... 83

......... 97
......... 97
......... 97
......... 99
......... 101
........ 103
........ 115

........ 116
.........119
........ 119
........ 125

........ 130
........ 130

........ 131
........ 131
........ 132

........ 134

LIST

labia
Chara.

1 3316

12b“
u 1C

LIST OF TABLES

Table 3-1: Recent Machining Experiments and Finish Surface

Characterization. .......... 34
Table 3-2. Composition of the hot-rolled steels (all in wt%). .......... 39
Table 3-3. The specific geometry of inserts for turning test. .......... 39
Table 4-1. Composition of the hot-rolled steels (all in wt%). .......... 70
Table 4-2. The specific geometry of inserts for turning test. .......... 70

.......... 79

Table 4-3: the results of cutting coefficient for abrasive and dissolution.
Table 5-1. Composition of the insert substrate. .......... 100

Table A1: The estimated values of excess free energy of solution of tool
component in y-iron (austenite) ........ 131

viii

LIE

Fig‘
1001

[31].

F391;!
dunn

Figur

HELL-”r
finite
0f 15:
at 3 SI
[46].

FERHe
mEQU

Opiic E

LIST OF FIGURES

Figure 2-1: (a) A schematic diagram of chip formation [2] and (b) a laser
scanning microscopy micrograph of a chip after cutting A181 1045 steel
with a cutting speed of 125 m/min, a depth of cut of 0.635 mm, and a
feed rate of 1.956 mm/rev.

Figure 2-2: Photomicrographs of the 3 types of chip: (a) discontinuous
chip (type I), (b) continuous chip (type II), (c) continuous chip with built-

up edge (type III) [7].

Figure 2—3: Schematic diagram of cutting forces for orthogonal cutting
(Merchant’s circle) [15].

Figure 2-4: Schematic diagram of wear in a carbide insert during a
turning test, (a) flank and crater wear, (b) rake face, (0) flank face [23,
24].

Figure 2-5: the schematic diagrams of wear mechanisms

Figure 2-6: A schematic diagram of the various coating techniques for

tool inserts, with typical coating thickness and processing temperature
[31].

Figure 2-7: A schematic diagram of heat affected-temperatures areas
during cutting process.

Figure 2-8: A comparison of the measured vs. calculated tool surface
temperature for A181 1113 steel [6].

Figure 2-9: Temperature distribution in a cutting test, as predicted from a
finite element analysis (a) steel cut with a carbide tool at a cutting speed
of 155.4 m/min and a feed rate of 0.274 mm/rev [47] (b) 12L14 steel cut

at a speed of 106 m/min. The broken lines are the measured temperatures
[46].

Figure 2-10. A schematic diagram of (a) the thermocouple setup for the
measurement of the tool-chip interface temperature [50] and (b) a fiber
optic pyrometer used to measure the rake face temperatures [5]].

ix

.......... 11

.......... 13

.......... 17

......... 19

....21

.......... 23

........... 25

.......... 27

Figure 2-11: Cutting temperature distribution in a turning process for a
0.3%C and 1%Mn steel: (a) sharp tool (b) worn tool with a cutting speed
of 180 m/min and a feed rate of 0.74 mm/rev [6].

Figure 3-1. Micrographs of pealitic steels of A181 1018, 1045, 1070 and
4340.

Figure 3-2: Flank wear volume per sliding distance for plain carbon
steels with (a) TiN, (b) Alumina (c) TiCN.

Figure 3-3: Experimentally measured flank wear and calculated flank
wear (solid lines) from Eq. 5.

Figure 3-4: Sample preparation for microstructure to characterize the
machined surfaces.

Figure 3-5: Scanning electron micrographs of near surface areas after cut.

Figure 3-6: X-ray diffraction patterns of the surfaces after machining.

Figure 3-7: TEM images and diffraction patterns of machined surface
area.

Figure 3-8: Hardenability curves for A181 1045, 1060, 52100 and 4340
Figure 3-9: Similar heat penetration volume (solid lines) and different
phase transformation volume during cooling (dotted lines).

Figure 4-1: The calculated relative abrasive wear rates

Figure 4-2: The calculated relative dissolution wear rates for TiN,
alumina, and TiCN.

Figure 4-3: Illustration of the chip-breaker set-up over the insert (Top and
Isometric Views) [59].

Figure 4—4: Temperature distribution in a cutting tool for the inverse
temperature estimation.

Figure 4-5: Experimentally measured crater wear rate verses temperature
for various coating materials.

.......... 28

.......... 39

......... 42

.......... 45

.......... 48

.......... 50

.......... 53

.......... 55

.......... 59

.......... 59

.......... 66

......... 68

.......... 73

........... 75

........... 79

.v‘“
1 r

18““
5

,.

Ltd:

C0a:

Inez}
50m

Figure 4-6: Experimentally Measured Crater Wear Rate verses
temperature for various coatings and work materials.

Figure 4-7: Microstructures of the crater area for TiN, and TiCN
coatings.

Figure 4-8: Microstructures of the crater area for alumina coatings.
Figure 4-9: The evolution of crater wear with cutting time

Figure 4-10: The evolution of crater wear as a function of cutting time in
seconds elapsed with WC tool.

Figure 4-11: The crater depth (KB) with elapsed time

Figure 4-12: The Quantitative comparison of element on the crater area.
The scanning areas of A, B and C are indicated in Figure 4-9.

Figure 4-13: Micrographs taken on the crater areas (a) inner crater edge
(b) near cutting edge.

Figure 4-14: A schematic diagram showing the mechanism of iron
deposition.

Figure 4-15: Comparison of crater area after etching WC tool inserts.
These images are taken areas on Area 2 and 3 indicated in Figure 4-13.

Figure 4-16: X-ray diffraction analysis on the crater area.

Figure 5-1: a) CVD alumina-coated carbide insert, b) A1203 coating
surface morphology by SEM, c) cross sectional view for coating
thickness measurement by LSM.

Figure 5—2: Schematic diagram showing mechanism of inducing
compressive stress field by LSP, a) upon the plasma expansion showing
application of tensile strain field, b) after the interaction, arrows showing
induced compressive stress [105], c) LSP on the CVD processed alumina
coated carbide insert using excimer laser.

Figure 5-3: Measurement of tensile residual stresses by X-ray diffraction
method on A1203 coated carbide inserts (20 pulse irradiation with
constant laser energy density at 1.3 J/cm2)

xi

.......... 80

.......... 81

.......... 82

.......... 85

.......... 86

.......... 87

.......... 89

......... 91

.......... 93

.......... 94

.......... 96

.......... 100

.......... 102

......... 105

Figure

Il‘rlmki
11' cm:
31 const

Fzgire
craters.
IDt‘Tpht
morph

httre

figure
craters
morp'r

morp-f

Figur:
(N de

Fig.3:

‘\
afld_-

F701"-
‘Kdil
5

0H >6

Figure 54: Residual stress reduction by LSP, a) residual stress vs.
number of laser pulses (0 ~ 140 pulses at constant LPD of 6.0x107
W/cm2), b) residual stress vs. laser energy density (from 0.6 to 2.3 J/cm2
at constant 30 pulses).

Figure 5-5: Surface morphology and corresponding 3D topography of the
craters, for LSP with various number of laser pulses, a) surface
morphology, non-treated, b) 3D topography, non-treated, c) surface
morphology, 120 pulse LSP, and (1) 3D topography, 120 pulse LSP.

Figure 5—6: Plots for LSP showing number of laser pulses vs.(a) width
and(b depth of craters at constant LPD of 6.0x 107 W/cm2), .

Figure 5-7: Surface morphology and corresponding 3D topography of the
craters, for LSP with various laser energy densities, a) surface
morphology, 0.6 J/cm2, b) 3D topography, 0.6 J/cm2, 0) surface
morphology, 2.3 J/cm2, and d) 3D topography, 2.3 J/cm.

Figure 5-8: Plots for LSP showing laser energy density vs. (a) width and
(b) depth of craters at constant 30 pulses.

Figure A-l. The schematic diagrams of the difference between 3-body
and 2-body abrasive wear.

Figure A-2: The theoretical the 2-body and 3-body abrasive wear models
on selected coated inserts using the computer algorithm [56].

Figure A-3:Geometry of band source [3].
Figure A4: The temperature of the sliding surface from equation A33

and 4(solid line, cutting energy is available in the literature [111] and the
temperature of crater area measured by the pyrometer systems(dots).

xii

......... 107

.......... 109

......... 111

......... 113

......... 114

......... 120

.......... 124

......... 127

......... 129

11

._..
_ [In

ex

”135:7

1. INTRODUCTION

An advanced scientific methodology for predicting tool wear is needed in order to
automate machining processes in various industrial applications. A scientific method
would allow us to directly apply the developed wear model to different types of work
materials such as alloys and composites and many advanced coated-inserts such as multi-
layer and tertiary coatings. Consequently, it would greatly reduce the number of
experimental investigations and associated cost. This thesis will concentrate on coated
tools because they are especially important in recent years due to the drastic improvement
in the machinability. Cutting fluids were not used in this thesis to preclude any complex
heat and fluid boundary conditions on inserts and contact surfaces. In addition, dry
machining is becoming more popular due to the environmental friendliness of the
processe. Thus, the results of this research can be directly applied to “Green
Manufacturing”.

The intent of this study is to develop a new scientific methodology for predicting
the wear behaviors of inserts in machining. The resulting comprehensive wear model
must be able to describe the wear behavior with a set of fundamental parameters such as
interface temperatures, microstructure, and properties of a work material. Therefore, this
study will certainly provide a new paradigm where a rigorous wear model can be applied
to predict tool wear and to select a proper tool for a specific machining application.

An understanding of tool wear behavior requires through understanding of the

physics of metal cutting. The fundamentals of physics are briefly reviewed in Chapter 2.

Tr
the predir
show son
phase tr:
transfom'
dissociari
discussed

C
combinat
Pitdictiu
cutting 1:
Addition.
Various c
The emit
“’33! area

In
refinance

Lem136mm
Shocking
residual sr

the 1091 cr

The observed flank wear rates in machining spherodized steelsmatch well with
the prediction with the developed models, while the observed wear rates in pearlitic steels
show some deviation from the predictive wear behavior. Based on our observations, the
phase transformation of pearlites during cutting is proposed and proved. A phase
transformation occurrence in the cutting process diminishes the flank wear due to the
dissociation of hard inclusions in the work materials. This phase transformation is
discussed in Chapter 3.

Crater wear is discussed in Chapter 4. The main mechanisms of crater are the
combination of dissolution and abrasive wear mechanisms. Under this premise, the
predictive dissolution wear model for crater wear was developed as a function of the
cutting temperature and solubility of the coating material into a work material. In
addition, abrasive wear in the crater wear is discussed. The crater wear behaviors of
various coatings were studied experimentally and compared with the calculated values.
The evolution of crater wear is discussed with the observed microstructures on the crater
wear area in Chapter 4.

In Chapter 5, the use of surface treatment on the tool surface to improve wear
resistance is discussed. CVD coated tools have a large residual stresses due to high
temperature processes. A compressive wave was applied on the coated surface by a laser
shocking process (LSP) which reduced the residual stress. With the reduction of the
residual stress, the hardness of coating increases. This minimized the abrasive wear and

the total crater wear is diminished as shown in chapter 5.

 

Firefly
based (

afresh

Finally, the results and discussions of this research are summarized and the conclusions
based on this research are provided in Chapter 6. The details of calculation of the

abrasive and dissolution wear models are presented in the Appendix.

QOHE
§'\-
16-0] f
T‘INI
E‘SQC‘

10m-

2. LITERATURE REVIEW OF METAL MACHINING

In this chapter, the fundamentals related to metal cutting will be reviewed briefly.
The fimdamentals discussed here include physics and mechanics of metal cutting, tool

wear behavior, and cutting temperature behavior in the cutting process.

2-1 Physics of Metal Machining

The work materials are subjected to concentrated shear deformation along a shear
plane during metal cutting. The work material that passes through the shear plane
experiences a substantial amount of plastic deformation. To relieve the strain energy
caused by the shear deformation, chips are formed at the front of the cutting edge. The
study of chip formation has been a fundamental aimed toward understanding the basics of
metal cutting processes. Mallock [1] investigated the chip formation process with
various metals. He sketched a series of the chip formation events by observing the
deformation of the primary and secondary shear areas with an optical microscope.
Piispanen [2] proposed a simple chip formation model as shown Figure 2-1. In this
model, chips are formed from a recurring process of cut and slide along the shear plane at
an angle known as the shear angle.

Piispanen’s model is ideal, where the chips are formed by shear deformation
along the shear plane in the primary shear zone. In fact, the deformation zone ahead of a
tool for most materials is more complicated, and other machining parameters such as rake
angle and cutting speed.. have been introduced to explain the mechanism of the chip

formation during machining processes.

 

 

 

Workpiece }
(a)

 

Figure 2-1: (a) A schematic diagram of chip formation [2] and (b) a laser scanning
microscopy micrograph of a chip after cutting A181 1045 steel with a cutting speed of
125 m/min, a depth of cut of 0.635 mm, and a feed rate of 1.956 mm/rev.

In
cutting er
formed 1

huh-up

as steel.
this re;

corrtirrt

In actual machining processes, the chip takes on various shapes depending on the
cutting conditions. Chips can largely be classified into 3 categories based on the shapes
formed in metal cutting: discontinuous chips (type 1), continuous chips (type 2), and
built-up edge (type 3) [3, 4].

A continuous chip (Type 2) is commonly formed in cutting ductile materials such
as steel, copper and aluminum alloy. It can also be seen at high cutting speeds [3]. For
this reason, the chips in most metal cutting are continuous. The formation of a
continuous chip is caused by the shearing work materials along the primary shear zone.
Although, this type of chip generally produces a good surface, this is not necessarily
desirable. In automated machining systems, continuous chips can tangle with the tool
holder, thereby stopping machining process. Using a chip breaker, this problem can be
improved.

Discontinuous chips can be formed in brittle work materials, which are not able to
endure the high shear strains generated during machining. During cutting, microcracks
are generated and propagated across the chip in the primary shear band and discontinuous
chips are developed as shown Figure 2-2(b). Discontinuous chips can be formed under
strains of larger than 1 in the shear band [5].

A Built-up edge (BUE) may form at the tip of tool during the machining process
as shown Figure 2-2 (c). BUE’s are formed at the cutting edge of the tool, and grow
progressively. They may be carried away by the chip, or remain in the region ahead of
the cutting tool. If a BUE forms, the geometry of the effective cutting edge changes and
the resulting surface finish becomes poor. As the cutting speed increases, BUE’s either

decrease or disappear. Under normal turning conditions, BUE’s take on their largest

U

values at a cutting speed of 18 m/min and disappear at a cutting speeds of over 72 m/min
[6]. The other cutting conditions that reduce the size of BUE’s are: a smaller depth of

cut, a larger rake angle, and/or the usage of lubricants [4].

 

  

Figure 2-2: Photomicrographs of the 3 types of chip: (a) discontinuous chip (type I), (b)
continuous chip (type II), (c) continuous chip with built-up edge (type III) [7].

2.2 )1 ech ar

crater wear
cuttrrg for.
cutting for
and crater
Figure 2-2
material (
are requiy

T!
geometn.

Merchan

2-2 Mechanics of metal cutting

The cutting force and wear are related; cutting forces affect tool wear and vice
versa. Michetti et a1 [8] claimed the cutting forces are linearly related to flank wear and
crater wear. Cho and Komvopoulus [9] observed the relationship between wear and the
cutting force by performing a tuming test. They measured flank and crater wear and the
cutting force components. In the turning of A181 4340 steel with coated tools, as flank
and crater wear progressed, the axial cutting force and radial cutting force (Fc and Fr in
Figure 2-3) increased until the end of cutting. The mechanical properties of the work
material (such as shear strength) also affect on the cutting force. Higher cutting forces
are required for materials with high yield strengths [10].

The force components related to the machining have been developed using
geometric relationships and experimental data [1 1-14]. Among the available models,
Merchant’s [15, 16] description has been accepted as a basis of the mechanics in metal
cutting processes. Figure 2-3 shows Merchant’s circle diagram representing the
components of force together with the related angles in the cutting process. R is the force
between the tool face and the chip and can be resolved into 3 sets of components: the
forces along the cutting direction (FC) and perpendicular to the cutting direction (FT);
shear force along the shear plane (F5) and normal force on the shear plane (FN); and the

force along the tool face (F) and normal to the tool face (N).

derermir

fallout:

~r1
1|
:1 .

A dynamometer is used to measure the forces Fc and FT. Once these forces are
determined, the shear- and tool-face components are analytically determined by the

following relationships:

FS=FCcos¢—Frsin¢ FN=Frcos¢+FCsin¢ or Fstan(¢+r—a)

. . Equation 1
F=FC srna+FTcosa N=F(.cos,B—F,.sma

10

Billie 2.
C”Ciel [1

c

I

+49)

v

Chip

 

Workpiece

 

 

 

Figure 2-3: Schematic diagram of cutting forces for orthogonal cutting (Merchant’s
circle) [15].

11

2-3. Slecl
2-3-1 Ch:

Gr
wear (see
life testis,
process. C
near) (see
severe me
Crater we:
dissolutior

mechanisrr

2.3-2 \‘ieal

The
“tar and r
examples 0
ChEmiCa] m
bur a COmbi;

One
lfltiijnS 1r
rolling a] 011g
incur“ to b‘

phlpomona] t

2-3. Mechanism of tool wear
2-3-1 Characteristics of tool wear

Geometric parameters related to the cutting test are defined to characterize tool
wear (see Figure 2-4). ISO Standard 3685-1977 (B) gives a geometric standard for tool
life testing. This thesis limits its discussion to two regions of wear during the cutting
process. One is located on the rake face (crater wear), the other is on the relief face (flank
wear) (see Figure 2-4). Flank wear occurs near the tool tip nose and is caused by the
severe mechanical and temperature conditions between the tool and the work material.
Crater wear, on the other hand, is mainly generated by the combination of abrasion and
dissolution at the interface between the chip and the tool surfaces. The detailed

mechanisms behind flank wear and crater wear will be discussed in Chapter 3 and 4.

2.3.2 Wear mechanisms

The mechanisms of tool wear can be classified into two categories: mechanical
wear and chemical wear. Abrasive wear, delarnination wear, and adhesion wear are
examples of mechanical wear, while dissolution wear and diffusion wear are examples of
chemical wear. However, tool wear may not necessary result from one single mechanism
but a combination of these mechanisms may be involved in the cutting process.

One of the most important wear mechanisms is abrasive wear, where hard
inclusions in a work material indent and scrape the tool surface [17-20] by sliding or
rolling along the tool surface, resulting in the removal of tool material. The wear rate is
known to be inversely proportional to the hardness of a tool material but directly

proportional to the hardness of the inclusions [21, 22].

12

Xi =
KM =
VB
VB

3'“

Figure 2“l: S.
33121 CHIEI' We;

 

KT = crater wear depth

KM = crater center distance

VBrm = wear land width

VBM = maximum wear land width

Figure 2-4: Schematic diagram of wear in a carbide insert during a turning test, (a) flank
and crater wear, (b) rake face, (c) flank face [23, 24].

bemeer.
subsurf;
extende
thin she
ahrasix'c‘
wear is.
this Stu:

be cons

Normal and tangential loads that are applied induce shearing along the interface
between the tool and the work material. Continual loading leads to deformation of the
subsurface, which results in the initiation of cracks (Figure 2-5(a)). These cracks are
extended and propagated with further loading. When the cracks reach the surface, a long,
thin sheet of tool material is removed, resulting in delamination wear [25-27]. While
abrasive wear occurs gradually, delamination wear comes about abruptly. Delarnination
wear is, however, observed to occur only in high-speed steels (H88) [28]. The focus of
this study is on understanding the wear of coated inserts, and delamination wear will not
be considered as a primary wear mechanism.

Abrasive wear occurs as tool materials are removed by hard inclusions. When the
harder inclusions slides along the tool-work material interface, plowing and removal of
softer material takes place, resulting in grooves or scratches on the worn tool surface.
Further compounding the problem is the fact that as cutting temperature increases, the
hardness of tool material typically drops — a phenomenon known as thermal softening.

A distinction is ofien made between two-body and three-body abrasive wear
(Figure 2-5(b)). In 2-body wear, hard inclusions penetrate into the tool surface and
scrape out the tool material. In 3-body wear, the trapped, hard inclusions roll and slide on
the tool wear surface. In this study, abrasive wear will be taken as the primary
mechanism responsible for tool wear. Accordingly, an abrasive-wear model will be used
to predict wear in a coated tool system. This model explains wear in the flank area well,
and it is in good agreement with experimental results [21, 22, 29, 30]. The details of two-

body and three—body wear are described in Appendix Al.

14

materia
junction
fracture
as the

relation
Howex'e
under a

adhesior.

atoms. :1.
the tool 5
susceptib
ml be u
away b”
on the dif
the [00] s
Ill-‘Ically d
the higher
times an.
IEml-‘ftature

Cher

[00] “Ear.

When an asperity on the tool surface comes into contact with the surface of work
material, the tool and work material can adhere firmly to each other and create an asperity
junction. Further loading on the interface causes the weaker junction near the asperity to
fracture (Figure 2-5(c)). However, this adhesion wear mechanism has not been accepted
as the dominating wear mechanism because it does not take into consideration the
relation between friction and wear, and the wear of the harder material is not explained.
However, when sliding wear occurs in vacuum or in an inert atmosphere, or if it occurs
under a large asperity flash temperature, then asperity adhesion is enhanced, and the
adhesion mechanism can become dominant [31].

Diffusion wear occurs when components of the tool material, such as carbon
atoms, diffuses into the chips. This deteriorates the tool surface and results in wear on
the tool surface. Tool materials that have high-diffusivity constituents such as carbon are
susceptible to diffusion wear at high cutting temperatures. In addition, the tool material
may be worn by the constituent atoms from the tool diffusing into, and being carried
away by the stream of chip flowing over the tool’s surface. Diffusion wear is dependent
on the diffusivity of the tool constituent into the work materials in the chip flowing over
the tool surface. Diffusivity increases with temperature, and the rate of diffusion
typically doubles for an increment in temperature on the order of 20 °C [10]. However,
the higher diffusion coefficients of oxides compared to those of covalently-bonded
nitrides and carbides does not explain high wear resistance of oxide coatings at high
temperature cutting conditions.

Chemical reactions between a work material and the insert can also give rise to

tool wear. The chemical reaction occurs at interface of tool and work materials and

15

 

caus

tool

111361

for c
achie
errr

dissol

causes the reaction region to weaken. This region is readily worn out by the motion of
tool and work material (Figure 2-5 (d)). However, they have only been observed in the
machining of titanium alloys [32], where the cutting temperature is extremely high.
Dissolution wear has recently been accepted to be an important wear mechanism
for crater wear, especially at high cutting speed [33-36]. As the work-tool interface
achieves high temperatures during machining, the tool material becomes
thermodynamically unstable and dissolves into the chips. In this study, abrasive and
dissolution wear will be taken as the primary mechanisms responsible for crater wear

(Figure 2-5(e)). During the course of this work, this assumption will be verified with the

wear rate observed in the machining experiments.

16

FifUre 2.5: lire:

Work material

 

 

    

 

(a) delamination wear

Work material

   

(b) 2-body and 3-body wear

Work material

 

 

(c) adhesion wear

Work material

. . v .

    
  

 

(d) chemical reaction wear

tool

Work materia

(e) dissolution wear

Figure 2-5: the schematic diagrams of wear mechanisms

17

Accor
2-6 51
proces
derelo

good e

of frict:
lends it

the ahra

due to r}
are PTOdr

all-UNITE r

2.3.3 Coated Tool Wear

Tool inserts are coated with thin ceramic films to improve wear resistance.
Accordingly, deposition techniques for tool inserts have also attracted interest. Figure
2-6 shows various film deposition methods together with typical film thickness and
processing temperatures [37]. Numerous new coating materials have been studied,
deve10ped, and marketed for commercial use. Temery coating and mutilayer coating are
good examples of emerging coating materials.

Titanium nitride (TiN) coatings have high values of hardness and low coefficients
of friction. They also adhere well to the carbide substrate. In addition, its golden color
lends itself to decorative applications. With the high hardness of TiN coated tool inserts,
the abrasive wear resistance was significantly improved [38-40].

Alumina coated tools have good wear resistance and low coefficients of friction
due to the high value of its hardness values and good chemical stability. Coating layers
are produced by chemical vapor deposition on carbide substrate. Even a few microns of

alumina coating can increase tool life by a factor of as much as 10 [41].

18

Iillica] C03:

 

Coating Methods on the Inserts I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Vapor . Solution state ~ Mona" State
Chemical Vapor Deposition Chemical Solution Deposition Laser
Physical Vapor Deposition Electrochemical Deposition Thermal Spray
Ion Beam Assisted Deposition Sol-Gel Welding
50' gel Sol gel
Welding Welding
Electroplating Electroplating
CVD/PACVD CVD/PACVD
“mm“ spraying Thermal spraying
PVD/PAPVD PVD/PAPVD
[BAD IBAD
L l I l I j I I If fl 1 l
10" r 10 102 103 10‘ 0 200 400 600 800 1000
Thickness of coating (um) Processing temperature °C)

Figure 2-6: A schematic diagram of the various coating techniques for tool inserts, with
typical coating thickness and processing temperature [31].

19

deter
defer
temp
and f
mecha
the IOC

“63L

number
generall)
Chapter. I

m6€Xpen

2.4 Cutting Temperature

As will be discussed in later chapters, temperature plays an important role in
determining tool life. In cutting processes, energy is dissipated by the plastic
deformation of materials and the friction between the tool and work material. As a result,
temperature increases in the shear zones and the neighboring regions including the crater
and flank surfaces (Figure 2-7). These temperature changes have an effect on the
mechanical properties of the work material and thus tool life. As temperature increases,
the tool softens and exhibits higher solubility, which results in a rapid increase in tool
wear.

However, it is impossible to measure the cutting temperatures directly because a
number of indirect measurement techniques have been developed. Each approach
generally gives only limited information on the complete temperature distribution. In this
chapter, the evaluation of cutting temperatures will be described from two points of view:

the experimental approach and the analytical temperature approach.

20

Secondary ‘5’ Rake
surface

f
5

Shear zone

Primary"? ................ ‘3:
Shear zone ‘

 

 

Workpiece

 

Figure 2-7: A schematic diagram of heat affected-temperatures areas during cutting
process.

21

2.4

pro;

man.

Where .
conducti
measurer
AISI l l 1
Conducrer

mm.

2.4.1 Analytical methods for determining cutting temperatures

Numerous analytical models for predicting the cutting temperature have been
proposed. Due to the complexity of the cutting process [6, 42, 43], there are usually
many assumptions in the development of the models.

The temperature change along the tool face is the sum of the temperature rise

along the shear plane (ATS) and the temperature rise along the tool-chip contact region

(ATC). Shaw [6, 44] has been able to express analytically such a temperature change in
terms of material properties and the energy dissipation on the cutting area:

RSuS RruFV

+ 0.377 —— Equation 2

sps CpC

 

AT = A7; +AT(. =

where AT is the change of temperature, R is heat flow partition coefficient, C is
conductivity, and u is specific cutting energy. Figure 2-8 gives a comparison between the
measured cutting temperature and the cutting temperature calculated with Equation 2 for
A181 1113 steel. The tool material was a K28 cemented carbide and the cutting test is

conducted at a cutting speed of 139 m/min, a rake angle of 20 and a depth of cut of 0.06

mm.

22

Depth of cut, d (mm)
2

 

 

 

 

4 6
600 I F I
O
O
05
g 400
E
G)
0.
0E)
I- ZOOW ............. . . .

 

 

 

r r
0 6.1 0.2 03
Depth of cut, d (in)

Figure 2-8: A comparison of the measured vs. calculated tool surface temperature for
A181 1113 steel [6].

23

the It

[101%

QUITE

11161

1.116

16H

Finite element analysis has been applied to estimate the temperature change on
the tool surface during the cutting process [44—46]. Although this analysis method is very
powerful, its success depends on the accuracy of the inputs — namely, the measured
cutting forces, tool-chip contact length, and work material properties. These parameters
must be obtained from experiments of specific groupings of tool and work materials.
Figure 2-9 shows a calculated temperature distribution [47] using the finite element
method. The workpiece is a 12Ll4 steel that has been subjected to a cutting speed of
155.4 m/min and a feed rate of 0.274 mm/rev. The maximum temperature is located near
the cutting edge and close to the midpoint of the contact length. However, the
temperature distribution has a limit of accuracy because the friction between tool and
work material has not been included. In particular, a reliable constitutive equation for
the work material is difficult to obtain due to the difficulty in obtaining measured values.
Despite these drawbacks, temperature distributions from limited data can be obtained by
FEM and then compared with the measured partial tool temperature distribution [46]
(Figure 2-9). The figure shows a good match between the predicted temperature

distribution and the limited number of measured values [46].

24

(a)

 

 

 

(b)

 

025mm

 

22 so

 

 

Figure 2-9: Temperature distribution in a cutting test, as predicted from a finite element
analysis (a) steel cut with a carbide tool at a cutting speed of 155.4 m/min and a feed rate
of 0.274 nun/rev [47] (b) 12L14 steel cut at a speed of 106 rn/min. The broken lines are

the measured temperatures [46].
25

2.4,7 E.

tempera
and cor
tempera
tempera
the mea
thennoa
broad to
results ir
are embe
r61311\'el_\
An exam
5 $11va

diflance l

mined f
infidiatiol
for the It

Were COm

[Ill] 5h0“.5

2.4.2 Experimental methods for determining cutting temperatures

Thermocouple technique has been commonly used in measuring cutting
temperatures. A thermocouple is composed of two dissimilar metals joined at a junction
and connected to electrical circuits. The interface of the junction is held at different
temperatures, a small electromotive force (emf) is generated, which depends on the
temperature difference between the junctions. Figure 2-10 (a) shows a general method for
the measurement of the tool-chip interface using a thermocouple. The elements of the
thermocouple are the tool and workpiece, and the output is the mean temperature of the
broad tool-chip areas. Care must be taken when a built-up edge is present because it
results in misleading temperature readings. To circumvent this difficulty, thermocouples
are embedded at various locations in the tool or the workpiece. While this method gives
relatively accurate temperature distribution, it requires extensive specimen preparation.
An example of a cutting temperature distribution determined with the embedded method
is shown in Figure 2-11. The maximum temperature is located on the rake surface at a

distance from the cutting edge.

The infrared optical technique can be used to measure the infrared radiation
emitted from the cutting zone. Boothroyd [48, 49] took photographs of such infrared
irradiations in the cutting zone, where the workpiece was preheated to 600 °C to account
for the low sensitivity of the film. The observed intensities of the infrared irradiation
were converted into the temperature values using standard calibration charts. Figure 2-11

(b) shows the tool-chip interface temperature distribution found by this technique.

26

(a)

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

    
  
 
   
 

Mercury —T‘1 Hot junction
slip ring
- Spindle ,
Coil junctio
Amplifier
Recorder
(b) Fiber optic
I 8.
, tool holder
Workpiece

Pyrometer

Recorder

1
Thermal
monitor

.0
. I O
:::=:-.-:3 .......
ooooooooooooooooooooo
0000000000
' 'o.u.c:e:0:u
ooooooo

Figure 2-10. A schematic diagram of (a) the thermocouple setup for the measurement of
the tool-chip interface temperature [50] and (b) a fiber optic pyrometer used to measure
the rake face temperatures [51].

27

Ewell]

5136]; (a) Si);
0.74 mmfe‘

1200 1000

 

400°C

x=-1.6mm

Gradient

 

(a)

01000, 9002rr00 3

Gradient
(b)

Figure 2-11: Cutting temperature distribution in a turning process for a 0.3%C and 1%Mn

steel: (a) sharp tool (b) worn tool with a cutting speed of 180 m/min and a feed rate of
0.74 mm/rev [6].

28

The more recent use of optical fibers in the infrared technique has enables a more
accurate estimation of the cutting temperatures, particularly at the tool-chip interface
regions [51-53]. A typical setup of such an infrared method is shown in Figure 2-10 (b).
It is important to note that the infrared technique of cutting-temperature measurement is
not without limitations. Temperature measurements are limited to exposed surfaces. Due
to accuracy requirements, the observed area is often limited to an area of a few square
millimeters. However, as more sensitive detectors are being developed, this limitation is
becoming less serious. In this study, infrared optic pyrometry technique is adopted to
estimate cutting temperatures. This method is deemed more reliable than the other
methods of cutting tool temperature measurement for the following reasons: (1) Setting
up thermocouples on the cutting tool is always a problem, given the dynamic nature of
the process, (2) the pyrometer has a fast response time which allows for rapid
temperature changes, (3) the pyrometer is more accurate, and its capability to measure
temperature without making contact means that it also saves time. This pyrometer

technique will be discussed in Chapter 4.

29

chose
isnthe
exper
irate
haul
the

1113i

3. FLANK WEAR BEHAVIOR IN TURNING PROCESS

In machining plain steels, the cementite phase abrades the flank surface of a tool
whose hardness value is much higher than that of the cementite phase under an
isothermal condition. During chip formation process, however, the cementite phase
experiences only a transient condition. Therefore, as the ‘hot’ flank wear surface
traverses across the surface of a work material, the ‘harder’ cementite phase abrades the
flank surface extending the flank wear land. In spherodized steel, the wear result shows
the flank wear rate is well matched with the 3 body wear behavior. The result of our
machining pealitic steels shows that flank wear rate has deviated from 2-body wear
behavior. Other imperative phenomena must be occurring during machining peralitic
steels, which deter the action of these abrasives. Although flank temperatures are on
average few hundreds degree lower than crater temperatures, the flank interface
temperatures can be high enough to austenize a very thin surface layer of the work
material when machining at high cutting speed. To verify this, additional cutting tests
were conducted with A181 1045 and 1070 steels at various cutting speeds. The machined
surfaces have been characterized using X-ray diffraction (XRD), scanning electron
microscope (SEM) and transmission electron microscope (TEM). The samples obtained
afier machining at high cutting speeds showed evidence of a pearlite-to-austenite phase
transformation. Also, this phase transformation in pealitic steels will be compared with
the observations made in other relevant works in the literature and, will be applied in the

characteristics of the resulting flank wear behavior.

30

 

with minor

being mach
becoming
Various to
Mar
reaction '
1001 We;
mathini
data [5:
knovm

How-cs

3.1. Introduction

A considerable amount of research has been done in an attempt to characterize
various mechanisms of tool wear during metal cutting. However, the physical phenomena
behind tool wear are so complex that only the empirical Taylor’s model has been adopted
with minor success for iron-based work materials. As other advanced materials are now
being machined in industry, understanding tool wear from a more scientific approach is
becoming essential. This will allow the developed tool wear model to be applied on
various combinations of work material and tool.

Many steady-state wear mechanisms, including abrasion, delamination, chemical
reaction [54], dissolution [32], and diffusion [55, 56] have been considered to explain
tool wear behavior [3]. Among them, chemical reactions were observed only for
machining of titanium alloy [54] and diffusion has not been supported with experimental
data [55]. Another argument against diffusion wear is that oxides such as alumina are
known to be wear-resistant especially at high temperature cutting conditions [56].
However, the high diffusion coefficients of oxides do not explain the excellent
performance of aluminum oxide at high temperature cutting conditions [57].
Delarnination wear typically occurs due to the sofiening of the tool and observed only on
H88 and alumina ceramic tools [3]. The dissolution wear is mainly responsible for crater
wear. In this present work, the main mechanism responsible for flank wear is taken to be
abrasive wear as claimed in our earlier work and others’ work [19, 21, 58].

Tool wear occurs in the flank and crater surfaces of a tool. In machining plain
carbon steels, flank wear is mainly caused by the abrasive inclusions present in the

cementite phase, which leads to the perception that the flank wear land should increase

31

with Lil
prer'iou
data m
cemenri'
the pred
been apr
consider:
and other
Flar
material 2
flank We;
conducted
steels. A15
The eXpen'
increases 1
howeyfl. f

DECESSaI-y II

with the cementite content as shown with metal matrix composite [20]. However, our
previous experiments [5 8] with the plain carbon steels showed that the flank wear rate
data were inconsistent with a monotonically increasing trend with respect to the
cementite content. In practice crater wear does not pose a greater challenge in terms of
the predictability as the combination of abrasive and dissolution wear mechanisms has
been applied to describe crater wear [33]. On the contrary, flank wear, in spite of its
considerably lower magnitude, is critical in meeting the required tolerance of dimensions
and other related factors such as the forces encountered, power consumed, chatter, etc.

Flank wear is especially known to cause by the hard second-phase in a work
material abrading the flank surface of a tool. This thesis is aimed toward understanding
flank wear in a quantitative and more scientific manner. Turning experiments were
conducted with various plain carbon steels at various machining conditions. Plain carbon
steels, AISI 1018,1045, and 1070 were chosen in order to vary the cementite content.
The experimental results showed that flank wear rates increase as the cementite content
increases for spherodized steels [59]. In our experiments with the pearlitic steels,
however, flank wear rates were consistently lower at high cutting speeds and did not
necessary increase with the content of cementite inclusions.

During machining, as pearlitic steel traverses across flank wear land at high speed,
the flank interface temperature can reach beyond the eutectoid temperature of the
material. A region near the surface of the steel experiences rapid heating and cooling. As
the steel turns, the region undergoes a temperature cycle. The temperature rises and falls
as the region passes through the primary shear zone and flank wear land. At high cutting

speeds, the pearlite-to-autenite transformation occurs during heating. During cooling, the

32

austenite phase transforms into martensite and/or martensite with retained austenite
depending on the alloy chemistry and the rate of cooling. Due to the short time involved
during the process, some of the austenite phase retains in a resultant microstructure. This
phenomenon has substantially obliterated the action of the cementite that extends flank
wear land. The cementite phase, which presumes to abrade the flank surface, dissociates
and no longer abrades flank surface. In fact, the phase transformation may have been the
reason for the scattered data in literature because the eutectoid temperatures and
composition of a steel are very sensitive to the alloying elements such as Ti, Mo, and Ni.
The presence of a phase transformation has been mainly focused in crater-wear
study due to the high temperature associated with chip flow zone. Although the effect of
phase transformation on crater wear has not been addressed fully, the observation of
pearlite-austenite phase transformation has been reported on the chip flow zone [60-63].
Crater wear is known involve a combination of abrasive and dissolution wear [32, 33, 64].
As the crater temperature increases, however, dissolution mechanism dominates tool wear
and the effect of abrasion diminishes.

Phase transformation at the tool-work material interface is more difficult to accept
due to the lower temperature involved. Nonetheless, a series of works has shown the
presence of ‘white layer’ on machined surfaces in drilling [65, 66], turning [65, 67, 68],
grinding [69, 70] and other types of surface finish [71]. Most of the reported works

concentrated on AISI 52100 and 4340 steels.

33

Table 3-1: Recent Machining Experiments and Finish Surface Characterization.

 

 

 

 

 

 

 

 

 

 

Studies Work Cutting Speed (V), Claims
Materials Feed (1’) and Depth (thickness)
of Cut (d)
Akcan et al. 52100 V=50-200m/min, UM (7.2 um)
(1999) 4340 f=0.05-0.1mm/rev, UM
M2 steels d=0.1-0.2mm, No Change
Chou and Evans 52100 V=45-450m/min, UM/RA
(1998 and 1999) f=0.05-0.75mm/rev, (3-10 pm)
d=0.025-0.75mm
Tonshoff et a1. 5115 V=145 m/min, RA
(1995a and f=0.lmm/rev, (3-6 pm)
1995b) d=0.2 mm
Masumoto et a1. 4340 V=9l.4 m/min, UM/RA (~ 5 pm)
(1987) f=0.89 mm/rev,
d=0.15 mm
Present Work 1045 V =75-275 m/min RA, UM
1070 f= 0.356 mm/rev (< 2 pm)
4340 (1 =1.905 mm UM

 

(UM-Untempered Martensite and RA-Retained Austenite)

34

 

As sh.
11.711) and 0
experiments
retained aus
and 1070 st-
phase was 0

other data in

 

3.2. lmplic:

Table
identificatior
depth of the
increases \s‘
aslmptote, ’
exI’erienced
“hi“ cOndiri
on the deVel
mOriel atITEe

Ali-ar
M2 Steels.
layers (\7
1c”Writing

Speed 1‘” 8 IS

As shown in Table 3-1, these works characterize the layer as untempered matensite
(UM) and/or retained austenitic (RA) phases. It is interesting to note that in all of the
experiments with 4340 steels, mainly UM has been reported. In our experiments, both
retained austenite and martensite were observed on the machined surfaces of AISI 1045
and 1070 steels at relatively high cutting speed (>200m/min), while, only the martesite
phase was observed in 4340 steel. The difference between our experimental data and the

other data in Table 3-1 will be explained later in this chapter.

3.2. Implication of the phase transformation on the machined surface in other works

Table 3-1 lists the more recent works that have concentrated around the
identification and characterization of a white layer. Chou and Evans [72] showed that the
depth of the white layer (on machined surfaces) attained afier machining 52100 steels
increases with flank wear land and increases with speed but seems to approach an
asymptote. They claimed that it was a mixture of retained austenite and martensite that
experienced a large amount of shear deformation. However, they did not indicate under
what condition these layers are formed. Subsequently, Chou and Evans [73] concentrated
on the development of the thermal model to predict the layer thickness and show that their

model agrees well with the experimental results.

Alkan et a1. [74] have performed similar machining study on AISI 4340, 52100 and
M2 steels. Their results provide evidence of martensitic phase transformation on thin
layers (~7 um) on the work materials of AISI 4340 and 52100 steels. The eutectoid
temperature of M2 steels was too high for the transformation to occur even at high cutting
speed. Matsumoto et al. [65] obtained a similar conclusion that their 4340 steels attained a

35

mixr
auste
Baser'
sear

transit

alumina
have sin
from mm
in flank
interfacia
Wear mec.
At r1
already be:
filctiOn be
Cons:‘fquenr
eutfi‘CIOId {I
Bracamom3
aimCitizes a
Wiring Speec
Mstnnan-Ol

a“SteniLiIIOn ‘

 

mixture layer of transformation consisting of untempered martensite and retained
austenite. These works presented the conflicting results on the nature of white layers.
Based on these works, the eutectoid temperature of the work material, the extent of flank
wear land and the cutting speed are important variables that augment the phase
transformation.

The work by Kim and Durham [75] presented the difference in flank wear on
alumina tools when machining AISI 1045 and 4340 steels. At the low cutting speeds, they
have similar wear behavior. However, at the high cutting speeds, the flank wear resulting
from machining 1045 steels was twice that of 4340 steels. They explained the difference
in flank wear at high cutting speeds due to different wear mechanisms and/or possible
interfacial oxide layer. Any convincing argument was not presented as to how different
wear mechanisms cause the difference in the flank wear rate.

At the crater interface, the temperatures are much higher. Thin layer of the chip,
already heated by severe shear deformation in primary shear zone, is being heated by the
friction between chip and tool as it traverses through the rake surface of a tool.
Consequently, the crater temperatures can easily reach the temperature beyond the
eutectoid temperature. Optiz and Gappish [60] and Hau-Bracamonte [63], Hau-
Bracamonte and Wise [62]and Shelbourn et al., [61] have shown that pearlitic steels
austenizes and some of the austenite phase are retained when machining at moderate
cutting speed. In this chapter, two main questions on flank wear in relation to phase
transformation are investigated: austenization during machining and the effect of

austenization on the tool.

36

3.3 Experiments

3.3.1 Turning experiments

The lathe used was a Clausing-Manchester mediurn-sized lathe with the
provisions for infinitely variable speed and programmable feed and depth. Dry cutting
experiments were performed at a constant feed rate of 0.3 56 mm/rev and depth of 1.905
mm. The cutting speeds were varied between 90 and 225 m/min to examine the different
flank wear behaviors. Cutting speeds were chosen after referring to the insert
manufacturer’s recommendations. The chosen machining conditions ensure that the
majority of wear takes place in steady-state temperature conditions [5 8].

Using the machining condition described above, hot rolled AISI 1018, 1045 and
1070 steels were turned at various cutting speeds to examine the effects of cementite
concentration on the flank wear. These work materials have the initial diameter of 15.24
cm and length of 58.8 cm. The microstructures of these steels were pearlite and ferrite as
shown in Figure 3-1. Major inclusions of these plain carbon steels are the cementite
phase. The detail compositional elements of the steels used in the experiments are
provided by the supplier and shown in Table 3-2.

Square tool inserts (ISO designation SPGN 19 04 12) were used for this study.
The geometry of tool inserts is presented in Table 3. For all inserts, the substrate is a
K420 Kennametal grade. Three different coating materials were used. The TiN-coated
inserts were commercially available from Kennametal as KC710. It consisted of the
K420 substrate also provided by Kennametal with a 4m coating of PVD TiN over it.
The A1203-coated inserts were prepared by CVD coating method, which was custom-

made for this study by Valenite Inc., Troy, Michigan. It has a 3m thick alumina

37

coating; however, due to the well-known problem of the A1203 coating not adhering to
carbide substrate, a lum intermediary layer of TiN was deposited between the A1203

coating and the K420 carbide substrate. Uncoated K420 inserts were sent to Balzers, Inc,

Lansing, MI, where a 3.5m coating of TiCN was performed. The PVD process used for

this purpose was Reactive Ion Plating.

38

Table 3-2. Composition of the hot-rolled steels (all in wt%).

 

Table 3-3. The specific geometry of inserts for turning test.

 

 

 

 

 

 

 

 

 

Back Side rake End Side End cutting Side cutting Nose
rake angled, relief relief edge edge angle. radius.
angle,orb angle,0c angle, 0, angle. Cc Cs 7'
0° 4°42' 4°42' 0° 15° 15° 1.19mm

 

 

1018

1070

 

50 pm

1045

4340

 

Figure 3-1. Micrographs of pealitic steels of AISI 1018, 1045, 1070 and 4340.

39

 

3.3.2 Wear measurement

After machining, flank wear land on an insert was measured using the Mitutoyo TM-505
toolmaker’s Microscope at a magnification of 200. The microscope was equipped with
digimatic heads enabling measurements accurately down to 1pm. Wear lands on flank

surface are measured using the microscope and converted to the wear volume using the

formula provided by Shaw [69];

2
V =bw ganfl (1)

where V is wear volume, b is a depth of cut, w is a length of wear land, and 0 is the relief
angle. Kim and Durham [75] argued that the measured wear land divided by the total
distance of machining is better than the measured wear land divided by the total time of

machining. Accordingly, the flank wear rate used here is defined as the flank wear

volume per sliding distance.

40

3.4.Results and discussions

3.4.1 Tool wear measurement

Flank wear rates should increase with increasing cutting speed [76]. In addition,
because flank wear is mainly caused by the abrasive inclusions, the flank wear land
should increase with an increase in the concentration of abrasive particles [20].
However, the flank wear land measurements from our experiments with 1018, 1045 and
1070 steels did not agree. When the flank wear rates between AISI 1018 and 1045 steels
were compared, the flank wear rates increased with the increase in cementite content.
However, when the flank wear rate between AISI 1045 and 1070 steels are considered,
they did not increase with higher cementite content. In fact, the cementite effect is less

obvious at high cutting speeds. As shown in Figure 3-2, this phenomenon occurred on

the coated inserts used in the experiment.

41

1000+va

 

 

 

A I 90 mlmin
”E 800 c O 150 mlmin
5,, ML 0 225 mlmin

D

’5

a 400 - .
g , o

"é 200 - . ' O -
u—. . g I -

 

 

 

0 A a A A
0.1 0.2 0.3 0.4 0.5 0.6 0.7
Concentration of C (%)

 

 

 

 

 

 

 

 

 

(a) TiN
1000 a T
A , I 90 mlmin
"E 000 » o 150 mlmin
g o 225 mlmin
r soc ~ '
9 . I I ‘
E 400 " . -4
g 200 1’ I “i
E
o A 1 4_ 4 k l 4 1 4_ I . 1
01 02 03 0 4 0.5 06 O 7
Concentration of C (%)
(b) Alumina
1000
. I 90 mlmin
g 000 » o 150 mlmin
nE ’ O 225 mlmin
3 600 - -
$3 .
S
h 400 ’
i... .
a . - .
E 0 . .l 1 g . 1

 

 

 

0.1 0.2 A 03 L04 0.5 0.6 J 0.7
Concentration of C (%)

Figure 3-2: Flank wear volume per sliding distance for plain carbon steels with (a) TiN,
(b) Alumina (c) T iCN.

42

As described in Sec. 3.1 and 3.2, this may be due to the pearlite-austenite phase
transformation at the flank interface when machining at high cutting speed. As our
experiments indicate in Figure 3-2, flank wear diminishes at the higher concentration of
cementite. To determine if austenization truly occurred, the machined surfaces of AISI
1045 and 1070 steels were characterized using SEM, XRD and TEM.

In our previous work on the spherodized steels [59], the flank wear rate decreased
with the temperature and showed good correlation with 3-body wear model (Figure 3-3).
Based on Rabinowicz’s (1961) model, 3-body wear rate per sliding distance was

developed and is given in Eq. 2.

V = KAtP.‘""’)/(Pr") (2)

where K is a cutting coefficient, A is the area fraction of cementite, P is the hardness of
tool and abrasive. The area fraction of cemente (A) is obtained by volume fraction which
is converted from the weight fraction of cementite. Temperature dependent hardness (hot
hardness) is used because thermal softening of the inserts and abrasive. The relationship

between hardness and temperature is express by an exponential function:

P(T) = Poe:aT (3)

where Po and a is obtained by curve fitting the hot hardness data.
In this analysis, P, is assumed to be the hardness of the temperature of the interface which

was measured by pyrometer system. While, the temperature of the abrasive is the

43

temperature of flank area (Tflank) which is determined analytically using J aeger’s equation
for moving sources of heat and temperature at sliding contacts. The relationship between

measured crater temperature and flank temperature is calculated and given (see Appendix

A2):

Tflm=r —(177+0.25T ) (4)

CNN?!“ CMIPI'

The cutting coefficient is determined from the experimental wear volumes. To apply this
cutting coefficient for any tool materials, this value was determined from the alumina and
TiN coated tool data. The general form of the flank wear rate (mm3/m) is given as

v = 1.166A(Pa("'”)/(P,“) (5)

The solid lines in Figure 3-3 illustrate the flank wear volume. The calculated wear

behaviors are well matched with the experimental data for alumina and TiN coated tool.

44

 

 

 

6000 *x. T I T r I I
A r‘.‘ ‘. I 1010—
.5 5000 - ‘-\ o 1045- - -
S , -\ A 1005 - ~ --
4000) ‘\ O 1095-----
2 . o \
E l \ \
$3000 - x \ \_ -
. \ \. .
\ \_
Age 2000 A \ . \ O ‘
\ \
> ‘ \ - '\. 4
— x \.
u-1000 - . *-;-;:--. °-
’ I ‘i‘t—‘ZT-‘f‘r-- ‘
.x ’ A L k l L ‘1 ‘- -‘

 

 

 

goo ‘ 8504 900 950‘100010501100115012001250
Temperature. (°C)

(a)TiN

 

1015——
1045- - -
1095- -----

OD.-

 

 

 

 

I
0...
§ -"'~
~ ~‘ ..........
"‘ ~ ~ .A" '7.

A A

 

004350‘900‘950 1000
Temperature. (°C)

 

 

 

 

 

 

(b)alumina

6m " I l I I j
ӣ5000, I 1045"-
E A mes-~-
5 ’ 1095- -----
24000“- 0
g .
$3000- .
"c‘2000~ -
.9
u_ .

1000- -

200 850 900 950 1000 1050 1100 1150 1200

Temperature. (°C)
(c)TiCN

Figure 3-3: Experimentally measured flank wear and calculated flank wear (solid lines)
from Eq. 5.

45

3.4.2. Examination of microstructure after cutting test

The flank wear land measured in our experiments with various pearlitic steels did
not conform to our current understanding that the inclusions in work materials mainly
contribute to flank wear. The flank wear measurements indicate that flank wear did not
increase with the increase in cementite phase. Therefore, additional cutting experiments
were conducted to examine any change in microstructure before and after machining.
AISI 1045 and 1070 steels were chosen due to the unexpected behavior on these steels
from the previous experiment. In addition, AISI 4340 steel was turned to be compared
with other works as mentioned in Table 3-1.

To examine any microstructure change on the machined surface, the bar samples
were sectioned to evaluate the microstructure changes that occur during cutting test.
After the samples were polished to mirror-like surfaces with diamond paste (30, 6, and 1
pm) and colloidal silica (0.3 pm), which are then etched by 2 % nital solution. The
cutting surface was analyzed with x-ray diffraction to determine the phases. The
microstructure change was evaluated using SEM (J SM- 6400V) with 20 kV accelerating

voltage.

3.4.3 Phase identification

To determine the microstructure on the machined surface, x-ray diffraction
(XRD) method was used on machined surface. Due to the short time involved for heating
and cooling, the transformed austenite phase should be retained after machining. Before
XRD tests, the analyzed surface was cleaned by plunging the samples into deionized

46

water in the ultrasonic cleaning system (L&R DC3) for 10 minutes and then washed with
ethanol. The machined surface and the sample attained before machining were examined
using Rigaku 200D dilfractometer with Cu-Kor radiation. 45 kV of accelerating voltage
and 100 mA of current were selected. The specimens were scanned in the range of 40° to
110° of 20 values. The phase on the surface was identified by matching the 20 values
from the standard files on x-ray powder diffraction patterns. The phase between
martensite (BCT) and fenite (BCC) was difficult to determine due to the closeness of
their characterized peaks. Consequently, a transmission electron microscope (Hitachi
H800) was used to determine the microstructure. The samples were prepared for TEM

analysis as shown in Figure 3-4.

47

200 um disk (low

speed diamond 50-60 pm thick disk (polished by
EB _, 600-2400 grid sandpaper)

/ \a

9
’7 One side dimpling (5-10
um) and Jet thining for
TEM analysis

 

Figure 3-4: Sample preparation for microstructure to characterize the machined surfaces.

48

3.4.4 Microstructure of machined area

The microstructures of AISI 1045, 4340 and 1070 before machining are shown in
Figure 3-1. In the hypoeutectoid 1045 steel, the pearlite and ferrite phases coexist
whereas only the pearlite phase can be observed in the near eutectoid 1070 steel. AISI
4340 steels include more alloying ingredients of Ni, Cr and Mo (see Table 3-2). The
presence of Mo and Ni reduces eutectoid composition in steels [77]. Shear deformation
and heat flux from a tool change the characteristic microstructures of these steels during
machining. Especially with the phase transformation, the change in the microstructure
will be significant. Figure 3-5 shows the microstructures of the samples after machining

the 1045 and 1070 steels at low and high speeds.

49

  
  

  
  

1pm

1045 steel (at speed of 75 m/sec and 275 mlmin)

Iriui

‘_ ]

1070 steel (at speed of 75 m/sec and 275m/min)

Figure 3-5: Scanning electron micrographs of near surface areas after cut.

50

As observed by Chou and Evans [72], the thickness of the white layers increased
with the increase in cutting speed. In the both microstructures of machined area at low
speed cutting, two distinct regions exist. The region A has a severe plastic deformation
layer with 3—5 pm thickness. On the contrary, the samples after machining at high cutting
speed, had three different regions. The region A has a similar microstructure as the
machined surface at low cutting speed. However, a new layer was observed in region B

close to machined surface, which may be the results of phase transformation.

3.4.5 Phase identification with X-ray diffraction method and TEM analysis

X-ray diffraction patterns were obtained to identify the phases on machined
surfaces. Figure 3-6 showed the plots of X-ray intensity and diffraction angle, 20, for
AISI 1045 and 1070 steels at various cutting speeds.

On the machined surfaces of the 1045 and 1070 steels machined at high cutting

speeds (225m/m and 275 m/min respectively), the X-ray diffraction patterns have the
peaks at 20 of 43, 50, 74 and 90° which exactly match the typical austenite peaks. The

austenite peaks are represented with downward arrows in Figure 3-6.

51

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

600 _ _3 500 ’ ti I V f V I ' I f I
550 r 1 450 _- I
500 '- j 400 r i '1
:33 ’ 1 350 - l -
e . . , .
g‘ 350 ’- { 3°° T 1
30°; 1 '5 25° T 1
c 250 h 1 200 T ‘1 7
_ $23: } 1 E 150 .. I I .
‘ 1 100 1 . l ‘
100 ’- 1i ‘ p ‘ * I * 4
L ' . J 50 ' l V i -
5: Ln E .JL . JL . A- - ‘ 0 MA .A . MIWL ‘
40 50 80 70 80 90 100 110 40 50 60 70 80 90 100 110
20 20
1070 at 75 m/min at 225 mlmin
50° ' v ' r v w a 1 v T
m r I V I I I I I ‘50: .:
5001. ‘00.. i
» 350 '- .1
40° - - 300 r- -
a * '95 250 - -
E 300 P 200: l
. c > 4
E 200- ii . - 150.
. [r 100
roar ii . so
P 1': A .A i 0 ‘
0 L ‘ ‘ *- ' M“ Le ' 4O 50 80 7O 80 90 100 110
40 50 60 70 80 90 100 110 20
20
1045 at 75 mlmin at 275 m/min
90° 7 I ' I V I ' I ' I V I V 900 ’ ‘r r V I Y ' ' ' v ' I I ‘
800 " -1 800: -:
700? .i 700 t- -
500. -1 600 - a
.2" 500 r- - g 500 . .
400 a 5 400: j
5 300 . . g 3°°i .1
200 - l - 200 - .
100 - J 100 - i .
o i L‘.‘ 4‘ L l _. JL 1 4A LA . 0 v I , 1 A, I v A r v—‘Ajk w ‘
40 50 80 70 80 90 100 110 40 50 60 70 80 90 100 110
4340 at 75 m/min at 275 m/min

Figure 3-6: X-ray diffraction patterns of the surfaces afier machining.

52

The observation from X-ray diffraction patterns showed that the retained austenite
phases were found on the 1045 steel when cutting at 275 m/min and the 1070 steel at
225m/min. The phase transformation from pearlite to austenite was shown to occur as
the surface temperature increased beyond the austenization temperature of each steel.
Because the cementite phase is the main abrasive in the steels we used, austenization had
direct consequence on the development of flank wear lands. The characteristic peaks of
martensite (BCT structure) are too closed to the ferrite (BCC structure) to distinguish
them. For example, 20 of the characteristic peaks of (110) planes located at 44.6 and 44.8
respectively. To identify this phase of the machined surface, TEM analysis were done.
Figure 3-7 shows the micrographs of the 1070 (225 mlmin) and 1045 (275 m/min) steel
with the diffraction patterns. There exist tangled dislocations which were caused by the
high deformation during the turning test. To verify the phase, the diffraction patterns
were obtained. With the software package for electron diffraction analysis
(http://cimesgl.epfl.ch/CIOL/ems.html), the obtained diffraction patterns turn out to be
characteristic martensite phase patterns (Figure 3-7). Interestingly, no austenite peak is
present in the machined surface in AISI 4340 even at the cutting speed of 275 m/min.
Figure 3-7 shows the micrograph of machined surface area of AISI 4340 with the
diffraction pattern. It shows the machined surface has martensite structure with grains
size of 0.2-3 um. Similarly, Akcan et a1. [73] observed the martensite phase in AISI

4340 with TEM image and a similar microstructure was obtained.

53

 

4340 at 275 mlmin

Figure 3-7: TEM images and diffraction patterns of machined surface area.

54

3.5. Discussion of experimental data

The fact that higher cutting speeds are needed to retain austenite phase in AISI
1045 steel than 1070 steel can be explained using phase diagram. AISI 1070 steel is near
eutectoid and the austenization temperature is much lower than that of AISI 1045 steel.
During machining, because the deformation and friction behaviors of AISI 1045 and
1070 are different, the flank temperatures may be different. However, it is very difficult
to accurately measure flank temperature. When no austenite peaks was observed on 1045
steels at the cutting speed of 225 mlmin, simply the higher cutting speed of 275 m/min
was used to promote higher temperature. This must have pushed the flank temperature
beyond the autenization temperature of the 1045 steels.

The retained microstructure is significantly manipulated during cooling by alloy
chemistry. Our experiments with AISI 1045 and 1070 steels indicate the presence of
retained austenite (RA) and other works with AISI 4340 steels reveal the presence of
untempered martensites (UM). AISI 4340 steels include typical alloying ingredients of
0.7% Mn, 1.85% Ni, 0.8% Cr, 0.25% Mo and minor trace of P and S. The presence of
Mo, Mn, Ni and Cr reduces eutectoid composition of a steel [77]. Especially any
addition of Mo drastically reduces the eutectoid compositions. In the particular 4340
steels used by Kim and Durham [75], 0.25% Mo reduces the eutectoid composition to
about 0.5w%C. The other alloy ingredients, Mn, Ni and Cr, lower the eutectoid
composition even further. Consequently, the 4340 steels used in the machining
experiments were very close to the eutectoid composition. AISI 1045 steels only had a
similar composition of Mn were present. The eutectoid temperature would not be

drastically affected because while Mn and Ni reduce the eutectoid temperature, Mo and

55

Cr increase the eutectoid temperature. Therefore, the 4340 steels have near eutectoid
composition while the 1045 steels have hypoeutetoid composition. The microstructure
(Figure 3-1) attests to the argument presented here. As Kim and Durham [75] observed
the wear behavior with AISI 1045 and 4340 steels, the similar wear behavior was
observed in our experiment with 1045 steels and 1070 steels. Another important
variable not presented in Kim and Durham [75] is the effect of flank temperature.

As a tool traverses across the surface of a work material, a small region of the
material rapidly heated. Because of the small volume of the material involved and the
large volume of the surrounding material, the region cools off relatively fast after a tool
passes through the region. If the heating and cooling rates can be exactly calculated, the
resulting microstructure can be predicted by refening to continuous cooling
transformation (CCT) diagram for each work material.

The hardenability curves were replotted for various steels in Figure 3-8. The
hardenability curve shows the cability of a steel alloy to transform into martensite and is
usually obtained with the J ominy End-Quench test where the distance from the quench-
end was plotted against the change in hardness. The hardenability curves indicate the
nature of retained microstructure afier a material has been cooled from their austenization
temperature. In our experiments, the retained austenite and martensite phase are
observed in machined surface. Based on the results, martensitic phase can be formed due
to rapid cooling rate. The cooling rate is so fast and some of the austenite phase can be
retained. In accordance with hardenability curves, we can speculate that the cooling rates

would be faster than 180 °C/sec.

56

Even though the heated volume is most likely similar in the cutting process, the
heat flux due to the deformation of each material may be different. From the
hardenability data and the experimental data, the heat-affected zone is obtained as shown
in Figure 3-9. Solid lines represent heat penetration region, which are generally the
same. Dotted lines represent the region undergoing a phase transformation during
cooling. A larger region of microstructure is affected in the 4340 steel than the 1045 and
1070 steels. Table 3-1 supports this trend. The thickness of the transformation area is
not truly compared with other works listed in Table 3-1 due to various cutting conditions
and friction at the inserts used in each of the works. However, the trend implied by Figure

3-9 cannot be disputed.

57

 

Figure 3-8: Hardenability curves for AISI 1045, 1060, 52100 and 4340

    

    

    

1045 Ste 1070 Ste 4340 Ste

Figure 3-9: Similar heat penetration volume (solid lines) and different phase
transformation volume during cooling (dotted lines).

58

4. UNDERSTANDING CRATER WEAR OF CUTTING
INSERTS

Cutting temperature has been known to play an important role in the generation of
tool wear. However, the literature is typically lacking in the tool wear studies that bring
together the temperature fields and the wear mechanisms. The major portion of this
chapter adopts a dual approach to the problem by empirical quantifying tool wear and
semi-analytically modeling tool wear. The rest of chapter is to address the underlying
physics behind the crater wear. Abrasive and chemical dissolution models are presented
in an attempt to explain the crater wear. This chapter presents the results of our
experimental studies to understand the wear mechanisms involved in the development of
crater on the rake face of a cutting tool. We have conducted the machining experiments
on various coated inserts as well as uncoated inserts. The resulting interfaces have been
carefully studied to shed light on the physics behind the crater wear and the reasoning

behind the development of crater wear, typically observed in turning steels.

4.1. Introduction

Predicting tool life has been one of the most difficult obstacles in automating
machining process. In a typical machining condition, tool life is the culmination of
gradual wear on a cutting tool. In the view of the worn areas, the gradual wear is
typically classified into flank wear and crater wear. Flank wear occurs at the relief face

of tool, which is caused mainly by abrasion of the hard second phase in a work material

59

[19, 78, 79]. Brun et a1. [79] and Kwon [58] found that tool materials harder than the
abrasives performed much well than the others. This dependence on hardness
represented a complex abrasion behavior depending on the interface temperature and the
temperature degradation of properties of the cutting tool. The hardness ratio between the
reinforcing phase and the cutting tool material seemed to quantify the abrasive process in
relation to flank wear [5 8].

Crater wear is typically a combination of various mechanisms such as adhesion [80,
81], abrasion, dissolution [82] and diffusion [28, 83]. There are still many debatable
issues when each or a combination of these mechanisms dominate. At high cutting
speeds, crater wear is believed to be dominated by either diffusion [28, 83] or dissolution
wears[32, 33]. Many wear studies have been concentrated on the flank wear due to its
immediate impact on the dimensional inaccuracy and other detrimental consequences
such as the forces encountered and power consumed in a typical machining process. On
the other hand, the crater wear does not limit the tool life directly. Crater wear increases
the effective rake angle of a tool and thus leads to decrease in cutting forces. Still
excessive crater wear produces weaker cutting edge, which eventually fractures a tool.

For many years, the diffusion wear mechanisms have claimed for the crater wear on
uncoated tools by Trent [83], Naerhein and Trent [84] and Cook and Nayak [28]. Later
Kramer and Sub [32] developed the dissolution wear mechanism to describe crater wear.
As the chip traverse across the rake surface, the chip flow becomes slower in the
horizontal component and leads to a considerable convection of dissolved tool material.
This convection mechanism transports the tool material into chip and brings about the

tool wear on the rake face of tool. With the assumption of the dissolution wear in the

60

moving stream of the chip material, the wear process was modeled as a diluted solid
solution formation and modeled as a regular solution. The chemical stability of an insert
is related to its dissolution [32]. They showed that the free energy of formation of a
ceramic coating determined the effectiveness of the coating in relation of its crater wear
resistance.

There are still many debates as to which mechanism is more appropriate for given
insert and cutting condition. However, Subramanian et al. [36] studied the wear behavior
of carbide inserts while cutting AISI 1045 steel. Using neutron activation analysis on the
chips generated during machining, they were able to distinguish the amount of tungsten
dissolved in the chip as a consequence of dissolution and the amount of tungsten carbide
particle detached from the tool as a result of abrasion. They also found coatings on the
tool with HfN, whose solubility is six orders of magnitude less than tungsten carbides,
significantly reducing the crater wear. Based on this experiment, dissolution wear
mechanism is believed to be dominant in the crater wear. Chubb and Billingham [85]
performed the machining experiment on TiC coated tool inserts with EN24 steel
(equivalent to AISI 4340 steel in US). They characterized crater wear with a measurement
of crater depth, KT, as the time elapsed and observed that the crater wear increased slowly
and then rapidly increased. The removal of coatings allowed the chip to be in a direct
contact with carbide substrate, which is much more prone to dissolution.

Dixon et a1. [80] studied initial crater wear with tungsten carbide tool inserts for
short cutting time (1-11 seconds). They found thin iron layer on the crater area and
claimed that the adhesion of iron degrades the WC. This causes intragranular

micromechanical fractures of the WC particles. Ham and Narutaki [86] claimed that the

61

crater wear is caused by chemical or adhesion at the tool chip interface. Akasawa and
Hishiguti [81] observed that work materials adhered to crater area. With ion probe
analysis, they found tool inserts material exited in the chips and claimed that atomic
diffusion of tool material into chips leads to the carter wear. Brandt [87] observed
dissolution/diffusion wear when addition of Ti (N, C) into alumina tool instigated an
increased crater wear rate. They also asserted that crater wear of alumina based ceramic
tool is mainly caused by superficial plastic deformation.

As evident by the previous works on crater wear, there is no conclusive results of
the exact mechanisms involved in crater wear. In this study, we have studied the crater
wear on various coated inserts as well as uncoated carbide inserts to shed light on
understanding of the mechanisms involved on crater. In addition, we have studied the
evolution of crater wear on carbide inserts and made a careful mapping of elements on
various locations of the crater to elucidate the details of the action taking place at the

crater interface during chip formation.

4.2. Comparison of the relative crater wear

We have developed the computer program that predicts the crater wear as well as
abrasive wear. The input to the program is the hot-hardness and the free energy of
formation of the respective coatings. The hardness data of the coatings was also collected
from the literature and a variety of sources [88-91]. The hot hardness data was then
curve fitted for an exponential relation and then used in evaluating the wear rate
expressions. The free energy of formation data for the TiN and A1203 binary coatings was

directly obtained from Kramer and Kwon [33] whereas the procedure for the tertiary

62

TiCN coating data is obtained in Appendix A. Though cementite is softer than the
coating materials at all temperatures, the variation of the abrasive wear-rate with
temperatures was more keenly seen. The theoretical data for the raw carbide indicated a
high rate of dissolution and very low rate of abrasion, in accordance with the conclusions
of others [32, 76].

Among the various abrasive wear models in tribology literature, the ones
appropriate to modeling abrasive tool wear are the three-body and two-body abrasive
wear models. Rabinowicz et al. [22] performed experiments wherein two surfaces slid
against each other with the abrasives introduced in-between. They drew conclusions
related to wear rates, sliding conditions and material hardness. A number of materials and
abrasives were chosen and a general empirical relation was found to fit the data. The final

form of the equations, as applicable to tool wear [33], is as shown in Equations 4.1.

 

P
Vm : XLtang ’ —"<08
3P, Pa
L a P ‘2” P '
V," = 1&[_] , 1.25 > —' > 0.3 Equations 4.1
5.3Pr P, Pa

 

—6
P
Vm=than6£ , —'>l.25
2.431% P, P

a

where tan 6 is the average tangent of roughness angle of the abrasive grains (a measure of
the particle shape or sharpness), x is a sliding distance, L is the normal force of
interaction between the surfaces, P, is the hardness of tool and P, is the hardness of the

abrasive. Equations 4.1 calculate the abrasive wear volume as a function of a sliding

distance, x. Three-body conditions exist when two bodies slide against each other with

63

the simultaneous rolling of a hard abrasive particle in-between. As will be shown in

Sec.4. 3, this is the proper condition at the crater.

 

1.E-04 — * ‘TiN

--u--»TiCN
S E 1.E-06 —A—Alumina
5 C ‘x
8’ l ‘xg
35 g.__ “\
§e1E‘08 ' -..-.'~._-.. \‘o
as.
> ‘1‘

1.E-10 - - 4
700 800 900 1000 1100 1200
Temperature(°C)

Figure 4-1: The calculated relative abrasive wear rates

65

At a relatively higher interface temperature, the chemical dissolution wear dominates the
wear process. A quantitative model developed by Kramer and Sub [32] has relatively
well predicted the crater wear when the chemical stability controls the wear process. To
simplify the calculation, a given binary (or ternary) tool material behave as regular

solutions then chemical solubility can be determined as [82]

 

A021,”. -xAGf,‘“ — AG}; —zAG(".' —RT(xlnx+ylny+zlnz)
CAM} =ex
' (x+y+z)RT

where C431: is the chemical solubility of the coating material in the workpiece (mole
fraction), AG A: B.('. is the free energy of formation of the coating materials AxByCz,

AG 3;" ,AG j,“ and A65“ are the relative partial molar excess free energy of solution of

component A, B and C in the workpiece, respectively, R is the gas constant and T is the
absolute interfacial temperature. Then, the dissolution wear rate may then be calculated
as [82]

BMVO'5 C A, ”7‘ .1 Equation 4.2

where B is the dissolution wear coefficient to be determined from cutting tests, M is the
molar volume of the coating material in cm3/mol and V is the cutting speed (mlmin). The
square root dependence on cutting speed comes from the mass transfer coefficient
expressed as square root of the Reynolds number that is proportional to the free stream
velocity [92]. Figure 4-2 illustrates the calculated dissolution wear rate. As expected, the
wear rate increases as cutting temperature increases. Also, the wear rate of TiCN coated

tool is highest while alumina has low wear rate.

66

 

 

 

 

 

 

 

 

0.007 . I V I I

—l— TIN ,
‘3 0.005 + Alumina f;
'D ...... . ...... TiCN .
h I
g 0005 '— "I, -
3 5.? '

C

5 3 0.004 -
:5 «E -
° ’= 0003 -
8 e '
‘6 3
g 0.002 -
.3
I!
g 0.001 _.

l- ........ V

0000‘ I 1—-—l——¥ """"" / a
500 800 1000

Cutting Temperatuer (°C)

Figure 4-2: The calculated relative dissolution wear rates for TiN, alumina, and TiCN.

67

4.3. Experiments

4.3.1 Turning tests

Cutting tests were done with a Milltronics Manufacturing 20 HP medium sized
lathe, which is capable of infinitely variable speed and programmable feed and depth. It
had a rigid tailstock, important for ensuring minimal chatter while turning long bars used
in this experiment. Dry cutting experiments were performed at a constant feed of 0.356
mm/rev and depth of 1.905 mm with various cutting speeds between 75 to 275 mlmin.
With the cutting times of at least 2 min, these selected cutting conditions ensure that tool
wear occurred mostly under steady state cutting condition. The work material are AISI
1018, 1045, and 1070. The compositions of steels are given in Table 4-1. The geometry
of the inserts had the ISO designation SPGN 19 04 12. The specific geometries for this
study are given in Table 4-2.

To observe the crater behavior on the coated tools, TiN, TiCN, and A120; were
coated on the WC substrate tools. The coating thickness of inserts is 4 urn, 3.5 um and 3
pm respectively. To observe the evolution of crater wear, additional cutting tests were

carried out using the work material of AISI 1045 steel with the cutting speed of 275

m/min of, the feed rate of 0.356 mm/rev and the depth of cut of 0.98 m using KC420

carbide insert.

68

Table 4-1. Composition of the hot-rolled steels (all in wt%).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

C Mn P S Si Ni Cr Mo Cu Sn AI V

1018 0.218 0.70 0.02 0.03 0.21 0.07 0.13 0.02 0.26 0.01 0.02 -
1045 0.48 0.74 0.01 0.04 0.27 0.05 0.08 0.02 0.11 0.01 0.04 0.004
1070 0.68 0.78 0.01 0.02 0.22 0.04 0.17 0.02 0.05 0.01 0.02 -
Table 4-2. The specific geometry of inserts for turning test.

Back Side rake End Side End cutting Side cutting Nose

rake angle,0rs relief relief edge edge angle, radius,

angle—,ab angle,0c angle, 0S angle. Cc Cs r
0° 4°42' 4°42' 0° 15° 15° 1.19mm

 

 

 

 

 

 

 

 

 

69

4.3.2. Characterizations of Crater Wear

The evolution of crater wear on the carbide tools was characterized in terms of the
crater depth as well as the worn area. The profilometer, Sloan Dektak IIA, was used to
measure the crater depth by gently dragging a mechanical stylus across a surface.
Scanning electron microscopy with 20 kV accelerating voltage was used to observe the
crater area as it evolves during the course of 2 min. To determine the elements on the
crater area after machining, energy disperse spectrometer (EDS) was used. The x-ray
signal was detected from the sample with thin Be-window and Li-drifted detector to
provide EDS spectrum for the information of elements. JEOL J SM- 6400V was used for
EDS analysis with 20 kV accelerating voltage and 200 mA current. To identify the phase
on the crater surface, Rigaku 200D diffractometer was used with Cu-Kor radiation, 45 kV
of accelerating voltage and 100 mA of current. The specimens were scanned in the range

of 40° to 110° of 20 values. The phase on the samples can be identified by matching the

20 values from the standard x-ray powder diffraction pattern files. Before tests, the

analyzed surface was cleaned by plunging the samples into deionized water in the

ultrasonic cleaning system (L&R DC3) for 10 minutes and then washed with ethanol.

4.3.3. Temperature Measurement

Due to undulating chips on rake face, the l-D ellipsoidal mapping model was
introduced, which simplifies the inverse problem of estimating the interfacial temperature

[93]. To use this model, an infrared pyrometer system was adopted to estimate the

70

cutting temperature. The coupling of fiber optics to infrared detectors can be beneficially
used in cutting tool temperature measurement since it allows the probe to travel along
with the cutting tool during the feed. The detector was 081513 sensor was connected to
Omega® Model 3026 single channel thermal monitor. The response time is 10-usec and
the range of measurement temperature was 84-300°C. The absolute accuracy of the
pyrometer was found to be 3°C. It possessed a response time of 10-usec and was
calibrated for the temperature range of 84-300°C. The end probe of the fiber optic cable
was a glass tipped steel probe of 7.50m in length and the spot size was found to be 0.785
mm2. The calibration of the pyrometer was checked using the BB-41 black body
calibration source. The non-linearity was found to be less than 1%. To resolve the
different emissivities of the coating insert surfaces, a thin coating of black high
temperature paint (Flat Black) was sprayed on the inserts prior to the tests, which set the
emissivity at 0.92.

A fixture to hold the pyrometer in place was designed using one of the same inserts.
Figure 4-3 shows the schematic of the set-up shown to illustrate this point. It should be
noted that there might be some disturbance in the temperature field due to this device.
But in all cases it was found that it was sufficiently far away from the interface as shown
in Figure 4-4. Temperature data was collected in an automated data collection system,

which consisted of a data acquisition board, signal conditioning and the software.

71

Hole for
end

   
 
 

C ip breaker

Coated
ins rt

nd probe

  
 
   

Optical fiber to
infrared sensor

Chip breaker

holder

Figure 4-3: Illustration of the chip-breaker set-up over the insert (Top and Isometric
Views) [59].

72

4.3.4 Inverse Estimation Scheme for the Chip-Tool Interface Temperature

The wear models discussed in Sect. 4.1 require the determination of the average
cutting temperatures. The inverse estimation of the chip-tool interface temperature
developed by Yen and Wright [93] will be used. For the steady state temperature
distribution, an assumption that the one-eighth ellipsoid to be an isothermal surface
shown in Figure 4-4 leads one to solve the steady-state Laplace equation in ellipsoidal
coordinates. However, it can be simplified for the special case of the one-dimensional (l-
D) steady-state problem, where the temperature distribution in the tool body is a function

O, only of the radial coordinate g as in equation 4.3 [51].

—d—[R¢ 99) = 0 Equation 4.3
6’6 d6

 

with R, =\/(02 4.3-(52 +5); and 9: T: ”at where g is the radial coordinate in the l-D
R _ T1:

 

ellipsoidal model (m2), a, b are the parameters describing the base ellipse with a>b
(mm), 9 is the relative steady state temperature, TR is the steady-state chip-tool interface
temperature (°C), T; is the temperature at the location determined by E (°C), and T00 is the
ambient temperature (°C). The boundary conditions specified are; all other faces are

insulated, therrnophysical properties of the tool material are constant, the tool is rigid and

tool wear is negligible.

73

 

 

 

Figure 4-4: Temperature distribution in a cutting tool for the inverse temperature

estimation.

74

The modeling and the governing differential equations are discussed in the
referred paper. The final form of the inverse solution used is the one used in [51] can be

expressed as

 

T—Too (2) _,[ (x) Equation 4.4
=r- - tan — -r
7;.—-T It a

where T r is the steady-state cutting interface temperature, and T is the remotely measured
rake face temperature, T.o is the ambient temperature (Taken to be 25°C), x is the distance

of the point of measurement from the origin of the axes and a is the radius of the circular

tool-chip contact area.
4.4 Experimental crater wear results

The main difference between diffusion and dissolution is in the assumption on the
nature of the interacting materials. Dissolution assumes the interacting occurs between
the solid tool material and the flowing chip which act like a fluid. Typically the crater
depth characterizes crater wear. The crater wear occurs by the contacting the rake
surface and the chip form workpiece. As cutting the progresses, the crater wear depth and
the contacting time of chip and tool increases. Hence, the crater wear have been
characterized by the crater depth with cutting time [9, 33, 34, 85, 87, 94]. Crater depth
per time (minute) was used for the crater wear rate in the study.

Figure 4-5 shows the crater wear rate with cutting temperature. The crater wear rate
increases with the cutting temperature for all cases. The crater wear rate of TiCN has

largest value of crater wear rate at high temperature as shown in Figure 4-5

75

The crater wear rate is a combination of the abrasive wear rate and the dissolution
wear rate. Hence, the general form of crater wear is the combination of Equation 4.1 and

4.2 and given as
Crater Wear Rate = Kab AV (Pam'm/(P,n ) + Kdiss(1 - A)MV°'5 C AxByCz Equation 4.5

where Kab and Kdiss are for the dissolution wear and abrasive wear and A is area fraction
of cementite. The coefficient Kdiss term is obtained from crater wear results of TiCN and
TiN at high temperature where the controlling crater wear rate is mainly due to
dissolution wear. After the value of Kdrss is estimated, the abrasive wear coefficient, Kab,
can be obtained. The values of the coefficients, Kab and Kdiss, are given in Table 4-3.
Based on Figure 4-5, the wear rate of alumina is deviated from the predicted
dissolution wear rate. This indicates that other wear mechanism may be involved in
crater wear of alumina coating. To examine this effect, the microstructures of the wear
area are investigated. As expected from the wear results of alumina, the microstructure
of alumina is quite different from the other two coatings (Figure 4-7 and Figure 4-8).
TiN and TiCN coatings have the smooth surface typically observed with dissolution
wear. While alumina coating shows very rough surface and some pullout of grains. This
phenomenon in the wear of alumina coating causes the deviation from the above wear
model. Therefore, the developed crater wear model is well matched with carbide and
nitride coating materials where the dissolution wear is dominant controlling wear
mechanisms. Figure 4-6 shows the crater wear rate behavior with respect to the
concentration for various cutting speeds and coating materials. In all cases, the crater
wear rate increases as a function of the cutting speeds. While, crater wear do not have a

relationship with the concentration of cementite as cutting temperature increases as

76

shown 11
tempera‘
disarm
Chapter
abrasive

temper.

shown in Figure 4-6. The cutting temperature on the rake area is subject to high cutting
temperature with increasing cutting speed and the effect of concentration of cementite
disappears due to the austenization on the surface of the work material discussed in
Chapter 3. Obviously, the dissolution wear has no relation with concentration of
abrasive. In addition, the dissolution wear mechanism is more dominant as the

temperature increases as shown in Figure 4-5.

77

Table 4-3: the results of cutting coefficient for abrasive and dissolution.

 

Cutting coefficient, Kab

Cutting coefficient, Kdiss

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

145 31
Table 4-4: Area factor of cementite
AISI1018 AISI 1045 AISI 1070
0.07 0.16 0.22
35 f - 1 A as , - , - , i ,
A I rw—o— .E I TiN—°—
E 30 O Alumina-~- 4 E 10 O Alumina----
E A TICN--+-- . E A TiCN-~-
g 5,
v ’ .' o
o 2.0- .- -1 ‘5
ti » .' 0:
(E 15* .r'A . 5
. .' Q)
8 1n. ‘ .'. .1 3
E s ‘.' I a
3 0.5 ~ 0* ------------ 4 E
800*---"T”‘ - 1' . o
O .00 1000 1100 120° 1300 900 050 1000 1050 1100 1150 1200 1250 1300
Cutting Temperature I0 C) Cutting Temperature (° C)
(a)lOl8 (b)1045
A3-5ITiN+-- ....r.,.‘
.5 O Alumina - ‘
E 1° A TiCN-H'- 1
E ‘1
2.5- ..
f l 1
vat-us 2.0» A. .1
E 1.5: J: .1
cu . I
E W 3‘ 1': ~
‘- s ' I ‘51," H 1
2 0.5 h ‘ "’t- :‘C‘ T
8 0° I’V-V.“;T..f':. ; A l A l A l L l A 1
U 000 950 1000 1050 1100 1150 1200 1250 1300
Cutting temperature (°C)
(0)1070

Figure 4-5: Experimentally measured crater wear rate verses temperature for various
coating materials.

78

1.. 1 v 1 I ‘ ‘1' I T l
E I 2751111111111 . A I 2251111111111
E u A 125mm 4 .E 1.4 A 125mm
‘ + 75 mlmin ‘ E + 75mm
2 1.4 ’ j E ‘2 _
o 1.2 - l '4 a ’ I
a P . 9 1.0 -
m 1.0 '- q a r
b ' ‘ m 0.0 -
8 0.0 - ‘ {i >
g ’ I A A ‘ 0) 0.0 .
b 05 - 4 g ’ I ‘
2 04 - . a 0.: - l
8 A + . *5 . g
02 rIt.t r ' I t r f t V v o 03 V Y ' I ' T ' I f I v I
0.1 02 0.3 0.4 0.0 0.0 0.7 0.1 0.2 0.3 0.4 0.5 0.0 0.7
Concentration of C (%) Concentration of C (%)
(b) alumina
A u I 225 mlmin 1 ' W ' ' 4
.E H A 1251111111111 _‘
E ' + 751mm 4
g 1.4 . .
v ' I 4
o ‘2 ’ I "
g 1.0 1- . -1
:6 0.0:- A -
o . .
g 0.0 .— 4
B 0.4:- ‘ § :1
E 02’ + + ‘
o 0.! 0:2 0:3 0T4 0.5 03 0']
Concentration of C (%)
(c) TiCN

Figure 4-6: Experimentally Measured Crater Wear Rate verses temperature for various

 

 

 

 

 

 

(a) TiN

 

 

 

 

 

 

 

 

 

 

 

coatings and work materials.

79

 

   

TiN (at 1203 °C) TiN (uncut)

         

TiCN (at 1213 °C) TiCN (uncut)

Figure 4-7: Microstructures of the crater area for TiN, and TiCN coatings.

8O

   

alumina (at 1203 °C) alumina (uncut)

    

Magnified micrographs of area A

Figure 4-8: Microstructures of the crater area for alumina coatings.

81

As shown in Figure 4-5, the ob served crater wear shows the close relationship
with temperature. Similarly, Cook [28] observed that crater wear increases with the
cutting temperature. He examined crater wear behavior of M2 high speed steel (H88) and
tungsten carbide tool inserts at various cutting speeds using AISI 1018, 1045 and 4140
steels as work materials. He claimed that under “high cutting speed conditions” where
cutting temperature reached over 427 °C for H88 and 704 °C for carbide tools, crater
wear rates are primarily function of temperature and not of stress, dimension, speed,
atmosphere and etc. Ingle [95] examined crater wear quantitatively with instrumental
neutron activation analysis (NAA) method for AISI 1045 steel and tungsten carbide. He
observed that the total amount of crater wear consisted of the abrasive wear and
dissolution wear component. He claimed that crater wear was dominated by dissolution
wear at high cutting speed. As the cutting speed increases from 150 to 240 m/min, the
contribution of the dissolution wear was elevated from 66 to 93 %. This wear behavior
agrees with our developed wear model. This developed model (Equation 4.5) shows that
the cutting temperature (or cutting speed) increases the portion of dissolution wear

increases rapidly as described in previous chapter.

4.5. Microstructural Analysis on Crater Area

A series of cutting tests were conducted on 1045 steels with KC420 carbide

inserts all at the same cutting condition; the cutting speed of 275 m/min, the feed rate of

0.356 mm/rev and the depth of 0.98 mm.

82

By using new cutting edges of the inserts, the SEM images of gradual changes on
crater were generated after machining for 15, 30, 45, 60, 75, 90 and 120 seconds. The
crater area of each cutting edge is examined with scanning electron microscope. A
gradual increase of crater wear areas is observed by superimposing the contours of crater
edges at various elapsed cutting times as shown in Figure 4-9 and Figure 4-10. Figure
4-11 shows the change in crater depth with elapsed cutting time. As Czicho’s wear
model [96] divides the wear-verses-time curve into 3 characteristic regions: a high wear
rate region at the beginning of cutting for short duration, a steady wear rate region which
has low wear rate and finally very high wear rate to failure. Our results show the similar

trends shown in this model as the wear rate is initially very high initially then slowly

increases.

83

2m (1m

 

Figure 4-9: The evolution of crater wear with cutting time

84

 

Figure 4-10: The evolution of crater wear as a function of cutting time in seconds elapsed
with WC tool.

85

 

 

 

 

Teraterdepth I ' j ' 1 1
2° ’ A after etching
. + Thickness A .
16 1- ‘
E ‘ '
V 0
g 12 '- ‘
11> r A '
u o
b
2 8 r- l o
E 0
U r
4 - + +
+ +
+ +
+
o r l A l A l A I A 1 A l
o 20 40 so so 100 120
Cutting time (seconds)

Figure 4-11: The crater depth (KB) with elapsed time

86

The microstructural observations of crater area show the existence of a new layer
covering the crater. The EDS method was used to determine the nature of the elements on
the new layer. Iron elements (> 98%) were mainly detected on the area A in Figure 4-9
while W and Fe elements were observed in the area B and C in Figure 4-9. W, Co and Ti
elements, which consist of tool inserts, were mostly detected and very small amount of Fe
(< 1.2 °/o) existed in area D. The quantitative comparison between W (tool insert) and Fe
(workpiece) are made on at least three local points and the average values were presented

in Figure 4-12. As shown in Figure 4-12, the amount of tungsten in Area A and B is

similar and decreases with the cutting time.

87

 

 

 

 

 

 

 

10 . , 1 ,
A Crater(edge. A)
r 0 Crater(inside. B)
3 L- __ 3K NearEdge(C)
; * 3% ~
'2 ° - I i ‘
5
8 4 _ _
.9
E
2 i
< 2 _ _ q
I I I I t
0 r l fi t V I I I
o 20 40 so so 100 120

Elapsed time (seconds)

Figure 4-12: The Quantitative comparison of element on the crater area. The scanning
areas of A, B and C are indicated in Figure 4-9.

88

During the chip generation process, stagnation region or dead metal zonel forms in front
of the edge radius. By replenished by additional work material coming into the
‘stagnation’ region, some of the materials must extrude out of the ‘stagnation’ zone
(shown in Figure 4-13 and Figure 4-14 (a)), which is carried out by the flowing chip. As
the chip traverse across the crater surface, the material from the stagnation zone creates
the convective (circulatory) motion within it. This is depicted in Figure 4-14. Near the
cutting edge of the tool, high normal and shear components of the traction make this layer
from the stagnation region to flow between the chip and the cutting tool while locally
creating the convective motion of the material. The local convective motion continues as
the layer moves away from the cutting edge. Both normal and shear components of the
traction are relieved at the region where the chip is detached from the rake face. Each
layer that moved across the crater coalesces to form thicker and wider layers. At this
point, how the detachment of this layer occurs is not evident. The important question in
the context of this thesis is the implication of this phenomenon on crater wear. The
deposition of thicker layers at the trailing edge of the crater does not imply that the tool is

completely protected from tool wear.

 

' Because the term, stagnation or dead metal zone, means ‘not moving’, it is misleading.

89

100 m.. m

 

Figure 4-13: Micrographs taken on the crater areas (a) inner crater edge (b) near cutting
edge.

90

This phenomenon implies that the local convective motion of this layer creates the
supporting evidence for dissolution wear. The implication of this phenomenon in term of
abrasive wear is interesting in terms of the action of cementite phase. Even though the
original microstructure of pearlitic phase implies 2-body abrasion, the crater will
experience 3-body abrasion during the motion of the layer if all the phases involved do
not undergo any transformation.

Figure 4-15 shows the microstructure of the crater after removing iron layers with
HCl. The observed area is near cutting edge and inside edge of crater area (the locations
denoted 2 and 3 in Figure 4-13). It has similar structures in crater areas. The EDS
analysis was done to check the elements underneath the thin layer and the results shows
tungsten and titanium which together with carbon constitute the major ingredient for the
carbide inserts used in this study. Even though crater surface is known be smooth, crater
surface appears rough because the cobalt was removed during etching the iron layer. As
shown in previous section in Figure 4-7 and Figure 4-8, the SEM micrographs of

alumina, TiN and TiCN coated inserts show smoother surface.

91

Stagnation zone

Chip

   

Iron debris occur Extrude out

Chip 7

Local Convective Chip detached from
Motion of a layer the crater area

Chip

      

Figure 4-14: A schematic diagram showing the mechanism of iron deposition.

92

 

Figure 4-15: Comparison of crater area after etching WC tool inserts. These images are
taken areas on Area 2 and 3 indicated in Figure 4-13.

93

To iden
#16 sh
speeds.
exactl)
austen
and 9(
Base:
auste
tool

84.?

To identify the phases on crater area, X-ray diffraction patterns were obtained. Figure
4-16 showed the plots of X-ray intensity and diffraction angle, 20 at various cutting
speeds. A series of peaks were observed at 29 of 44, 65, 82, and 99°. These peaks
exactly match the peaks observed with a ferrite or martensite phase. More importantly
austenite phase exists on the crater area with a series of peaks located at 20 of 43, 50, 74
and 90°, which is a proof of the phase transformation during the cutting test in Chatper 4.
Based on the existence of austenite phase, the previous peaks must be those from
austenize phase. Cobalt and tungsten carbide phase, which are the main ingredients of

tool inserts, were also observed with peaks located at 29 of 41, 44, 45, 60, 65, 72, 73, 75,

84, 98, 100 and 104.

94

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

700 T I . I I I I I I
6m - v—
500 - _
400 H ...
5:" .
(D
g 300 - -
200 - N
, V i
100 -
V l
0| I I I I ' I ' I ' I . I fl T
40 50 60 70 80 90 100 110
29

Figure 4-16: X-ray diffraction analysis on the crater area.

95

[CBS

puls

Res
X-r;
puls
rela
equ
con

TESL

C03
ther
incr
mac

The

time

5. SURFACE TREAMENT ON THE TOOL INSERTS

Laser shock processing (LSP) has been applied to locally relieve the deleterious
tensile residual stresses on the wear area of a coated tool. LSP utilizes a very short laser
pulse with high energy density, which induces high-pressure stress wave propagation.
The residual stresses are relieved by incident shock waves on the coating surface.
Residual stress levels of LSP CVD alumina-coated carbide insert were evaluated by the
X-ray diffractometer. Based on these results, LSP parameters such as number of laser
pulses and laser energy density can be controlled to reduce residual stress. From the
relationship between LSP parameters and the residual stress behavior, the empirical
equation is derived. Crater wear on the coated carbide inserts with various LSP
conditions was examined afier the turning tests. We have analyzed the preliminary

results to verify the effects of residual stress in reducing crater wear.

5.1. Introduction

Coated carbides have been widely applied for commercial metal cutting inserts
[33, 97-99]. The major advantage of coated carbide inserts is to reduce tool wear: TiN as
coating has low friction coefficient with high hardness, and A1203 as coating has a good
thermomechanical resistance to high temperature. Coated tools have been taken
increasing interest in machining new engineered materials since they can improve the
machinability by adding wear-resistant coating material on the various grades of carbides.
The improved wear resistance, at a higher cutting speed with the reduction in cutting

time, are the major benefits of coated inserts. One of the important deposition methods

96

for alum
deposit:
of a cut
in pren

operati

with b
substrz
crystal
Diamc
depos
low ]
morpl
16ml):
Inism
YEmaj

delan

reSidi
SITES 5

prOdL

for alumina coating is chemical vapor deposition (CVD). However, the CVD process
deposits films with high tensile residual stresses, which limit the abrasive wear resistance
of a cutting tool. When tensile residual stresses remain on the surface, these tend to result
in premature crack initiation and delamination, ultimately reducing tool life. In a typical
operation, the life of the inserts can be extended by a factor of at least two to three.

The physical vapor deposition (PVD) process deposits wear resistant coatings,
with beneficial compressive residual stress, due to the physically impacting plasma on a
substrate material. However, the growth rate by PVD is very slow and the deposition of
crystalline films has been not successful for the coating materials such as A1203 and
Diamond [3, 4]. These coatings are produced successfully by the chemical vapor
deposition (CVD) process, which involves vapor phase reaction at high temperature and
low pressure. CVD process deposits a film with uniform thickness, consistent
morphology and excellent adherence. However, the inherently high processing
temperature (700 °C ~1300 °C) raises residual stress level of a coated layer due to the
mismatch in coefficient of thermal expansion (CTE) [97]. If tensile residual stresses
remain on the coating surface, they are subject to the premature crack initiation and
delamination, ultimately reducing tool life [98].

Laser shock processing (LSP) is a unique method to alter the deleterious tensile
residual stress level. LSP, on metallic target, have produced compressive residual
stresses, resulting in improved fatigue and hardness properties [100-105]. Laser surface
treatment, with a very short laser pulse (~25 ns) and high intensity (~GW/cm2), can

produce surface plasma, which induces propagation of high-pressure stress waves in

97

about on:
surface b

T
stress b
measure
stress vs
energy .
experin
LSP cc
interprc

lESlS OI

5.2. C‘

Square
C VD-
reactc

S‘Ibi

about one m depth range [6, 7]. The compressive residual stresses were formed on the
surface by incident shock waves due to the recoil pressure of the surface plasma.

The CVD alumina-coated carbide inserts were used to show reduction of residual
stress by LSP. Localized relief of residual stresses was detected based on the
measurements by the X-ray diffractometer method. In addition, the change in residual
stress was observed by varying LSP parameters, such as number of laser pulses and laser
energy density. A simple relationship among these parameters is modeled based on the
experimental observations. Turning tests were conducted using the inserts with various
LSP conditions to examine how the LSP processing affected crater wear. Analytic

interpretation was accessed on the preliminary results for full extension of machining

tests on LSP of coated carbide inserts.

5.2. CVD alumina-coated carbide insert

Materials investigated in this study were A1203 coated carbide inserts, which have
square positive-rake geometry with honed comer (Figure 5-1a). Alumina thin film was

CVD-coated on C-6 grade tungsten carbide substrates in a commercial industry scale
reactor from Valenite® Inc. The microstructure of the A1203 coatings is shown in Figure

5-1b with average grain size of ~2 um.

98

 

 

Figu
by 8

Tab

 

‘,v5vna

~

Figure 5-1: a) CVD alumina-coated carbide insert, b) A1203 coating surface morphology
by SEM, c) cross sectional view for coating thickness measurement by LSM.

Table 5-1. Composition of the insert substrate.
W Element 1 wc 1 TaC I Co [ TiC l

IComposition(Wt%)[ 73.2 T 10.9 I 8.5 i 7.4 |

 

 

99

A
shows the
binding 1:
substrate
bonding
carbide s

in Table

Nucleat
substrat
growth

size ter

5-3. L:

that g
The s
the Si
Sires:

IS 113

A confocal laser scanning microscope (LSM, Zeiss 210 model) measurement
shows the average A1203 coating thickness about 1 pm (Figure S-lc). TiN interface
binding layer (average thickness is about 1 pm) was coated by PVD in between the
substrate and A120; coating. The TiN interface coating was deposited to improve the
bonding strength of A1203 coating due to the poor nucleation of A1203 on cemented
carbide substrates [98]. The detail composition of the substrate (WC-Co alloy) is listed
in Table 5-1.

For the CVD coating, the nucleation and growth processes are complex.
Nucleation is formed by the collision of atoms from the concentrated vapor phases on the
substrate [106]. Fine heterogeneous nucleation initiates on a substrate, and then the grain

growth takes place by subsequent lateral and vertical development. In general the grain

size tends to increase with the coating thickness [98].

5.3. Laser Shock Processing

The mechanism behind LSP is explained by the propagation of laser shock wave
that generates from a pulsed laser irradiation with a short duration time and high intensity.
The shock wave is induced from highly localized thermal expansion of plasma formed on
the surface of a target by laser-material interaction [104]. The shock wave (compressive
stress wave) propagates into the interior of the elastic solid. The propagation mechanism

is usually described in two step processes as shown in Figure 5-2 [105].

100

 

F1 gm
bs Ls
the ir

Proce

(1:)

plastic strain
a) pulse laser irradiation

 

 

 

 

effected volume
/_

~111—

—.._—

residual stress

b) after the interaction

 

 

 

 

 

reflecto% . . ,_- excimer laser -

C: o—planar convex lens

water \m/ N20, coated insert

I Q

x-y-z table

 

c) excimer laser LSP

He-Ne Ias r

Figure 5-2: Schematic diagram showing mechanism of inducing compressive stress field
by LSP, a) upon the plasma expansion showing application of tensile strain field, b) after
the interaction, arrows showing induced compressive stress [105], c) LSP on the CVD

processed alumina coated carbide insert using excimer laser.

101

compICE
(normal
surrour
comprc
water
plasrn:
shout
at KrF
excin
time
To c
EXCr
laye
suel

Plas

the

an

In the first step in Figure 5-2(a), the localized thermal expansion of plasma
compresses the surface of a target material, which subsequently experiences tensile strain
(normal to the stress waves) surrounding the impacted area. After the rapid interaction,
surrounding area of the interaction zone recovers the volume change by inducing a
compressive stress into the interaction zone (Figure 5-2(b)). In case of the LSP with
water confinement, it has been suggested that peak pressure (Gpa) of the expanding
plasma is proportional to the square root of the laser power intensity (GW/cmz) [105]. As
shown in Figure 5-2(c), laser shock processing on the insert surfaces was performed using
a KrF excimer laser (Lambda Physik®, Compax 301 model, wavelength 248 nm). A KrF
excimer laser pulse has 1.5 joule of maximum energy from its resonance with the duration
time of 25 ns.

To observe the residual stress change by various densities, laser energy density of the
excimer laser was used from 0.6 to 2.3 joule/cmz. A spot size was 0.5 x 1.3 cm2. A thin
layer of black paint (~ 100 um) was applied over the inserts to protect the A1203 coating

sueface from laser ablation. The inserts were submerged in deionized water to confine

plasma in the laser irradiation zone as shown in Figure 5-2(c).

5.4. Residual stress measurement

Generally, CVD reaction requires high processing temperature to carry out
thermodynamic mass transport. The processing temperature was about 1000°C for CVD
alumina coating on the WC substrate. Upon subsequent cooling, tensile residual stresses

are produced on the alumina coating layer due to the CTE mismatch between A1203

102

 

 

coating and the carbide substrate, which have CTE of 8.3 um/mK and 6.8 um/mK
respectively [97]. The tensile residual stresses on the coating surface tend to result in
crack initiation and delamination, and ultimately cause a reduction in tool life.

Alumina coating, whose residual stress is altered by LSP, is analyzed by X-ray
diffractometer (Sintag® XDS 2000, K=l.54) from which the difference of diffraction
angle (A29) can be measured by serial tilting of the specimens. The A29 of the alumina
coating was measured from (124) planes (d=l.404 A and 29=64°) with tilting angles of
0°, 10° and 20° degrees. The residual strain was calculated from the A29 caused by the
changes in interplanar spacing due to the residual stresses. To minimize possible error,
the calculation of the residual stresses requires a special elastic modulus and a Poisson’s
ratio for the highly textured and strained A1203 thin film. However, most of these special
values are not determined yet [97]. Hence, an elastic modulus and Poisson’s ratio of bulk
a-A1203 were assumed to be about 380 GPa and 0.26 respectively. We attempted to
determine the extent of reduction in tensile residual stresses by LSP. An example of
residual stress measurement/calculation is shown Figure 5-3. The peak points are
determined by obtaining full-width-half-maximum (FWHM). Then the average peak

shift (A29) is obtained from plotting FWHM value of each tiling angle.

103

6799111111111 111111111

 

 

0 005

l
l l
l l
l l

l
l

(0‘00
[[111]
f
11
5
ll
ID
0.

(
11
>3
1

1
11
Q
1111.111

llllTl“1

l

 

 

 

 

 

 

111111 111111111°°°5
sin’v

0‘
c»
O
O
0‘
0“
O
O
O
O
O
O
O
H
p
\l

Figure 5-3: Measurement of tensile residual stresses by X-ray diffraction method on

A12023 coated carbide inserts (20 pulse irradiation with constant laser energy density at 1.3
J/cm )

104

In Figure 5-4, residual stress changes were observed by varying two major LSP
parameters, the number of laser pulses and laser energy density. The changes in residual
stress values are recorded by varying the number of laser pulses from O to 140 pulses at

constant energy density of 1.30 J/cm2

. The results are plotted in Figure 5-4(a), which
shows gradual decrease in residual stress as the number of laser pulses increases where
the residual stress shows an ‘exponential decay’ type reduction with respect to the
number of laser pulses. The change in residual stress after30 pulses of varying laser
energy density is shown in Figure 5-4(b). The increase in energy density results in
gradual decrease in residual stresses with the similar trend to Figure 5-4(a).

From observed results of residual stress measurements, a simple empirical equation can

be induced for the residual stress (0,3) reduction by the LSP.

Where of, is residual stress level in as-coated condition (MPa), n is the number of

laser pulses, a is the laser energy density (Joule/cmz), and R is the residual stress factor
(MPa cm). The residual stress factor accounts the properties of coating materials, such as
elastic modulus (MPa) and coating thickness (cm). Even though this equation represents
the experimental trend in this study, further investigations are required to convert it to a
general model. The further investigations include systematic studies on coating material

properties (modulus, coating thickness, melting temperature, crystal structure, etc.) in

relation to the residual stresses.

105

 

 

5 .

E73

Residual Stress (MPa)

 

 

 

 

O'I'ITITI'I'ITT'
0 20406080100120140160

 

 

 

 

Number of Pulses
(a)
’65 .
n.
g 5200 -
d) r
(I)
g 5000 -
(0 1
g 4000 - -
:2 .
'0 I
o . .
m 4600
0.0 0.5 1.0 1.5 2.0 2.5
Energy Density (J/cmz) per Pulse
0))

Figure 5-4: Residual stress reduction by LSP, a) residual stress vs. number of laser pulses
(0 ~ 140 pulses at constant LPD of 6.0x107 W/cmz), b) residual stress vs. laser energy
density (from 0.6 to 2.3 J/cm2 at constant 30 pulses).

106

 

5.5. Crater Wear test

The wear in carbide inserts occurs in two major regions: flank surface and rake
surface. Geometric observation of flank wear land indicates that the hard phase abrasion
of the surface is the main source of wear during machining [78]. Crater wear is caused by
the contact of chip and the tool. Tool/chip interface is subjected to high temperature;
therefore, both abrasive and dissolution are the primary mechanisms responsible for crater
wear. Crater wear is formed on the rake face during machining.

Crater wear is one of the methods to determine tool life. It measures harmful
influences on cutting process, such as deteriorated tool/chip interface geometry and
degraded surface finish [34, 64]. AISI 1045 steel bar (diameter 15.24 cm, length 50.8 cm)
was used as a workpiece material for the turning test. Alumina-coated carbide inserts, in
various LSP conditions, were subjected to crater wear. Dry cutting experiments were
performed at a constant feed rate of 0.356 mm/rev and a machining depth of 1.905 mm.
The cutting speed selected was at 350 rpm (~165 mlmin), and total cutting surface area
was 0.238 m2 for each test. The crater wear on the inserts was measured using the
confocal laser scanning microscope. The examples of surface morphology and the

corresponding 3D topography after LSP are shown in Figure 5—5.

107

1422-94811m

1422 - 948 11m

 

Figure 5-5: Surface morphology and corresponding 3D topography of the craters, for LSP
with various number of laser pulses, a) surface morphology, non-treated, b) 3D
topography, non-treated, c) surface morphology, 120 pulse LSP, and (1) 3D topography,
120 pulse LSP.

108

The average width and length of the craters were measured fiom the surface morphology.
The average depth was measured from the 3-D topography, which is obtained by optical
z-sectioning of the craters. It is ideal to measure the removal value to evaluate crater
wear. Unfortunately, it is impossible to accurately measure crater wear volume directly;
therefore, the width and depth were measured as given in Figure 5-6. As the number of
laser pulse increased, the gradual decrease in the dimension of the crater was observed
from the measurements. The size of the crater was reduced in average approximately by

21 %, 10 % in width and depth which were measured 5 times, respectively.

109

‘m T v I r I V

 

 

 

 

 

 

 

 

I I I I I
1
350 1- -
L 1
E 300 - ..
5 I
E . E i
250 - E I E-l
2m 1 A l A l A l A l A l A l A l
0 20 4O 60 so 100 120 140
Number of Pulses
25.0 I I I I I I I I I
22.5 -' q
E 20.0- E E E4
5
31 I
17.5 d .1
J
15-0 T I I f I I I I r I I r I I I
0 20 40 60 80 100 120 140
Number of Pulse

Figure 5-6: Plots for LSP showing number of laser pulses vs.(a) width and(b depth of
craters at constant LPD of 6.0x107 W/cmz), .

110

 

For the LSP with various laser energy densities at a constant pulses of 30,
examples of surface morphology and the corresponding 3D topography are shown in
Figure 5-7. As compared to the non-treated condition in Figure 5-7(a) and Figure 5-7(b),
reduction in the crater dimensions is found at the highest laser energy density in Figure
5-7(c) and 7(d). The measured width and depth are given in Figure 5-8. Even though the
measured values of crater dimensions show some noticeable fluctuation, the reduction of
crater sizes was observed at LSP with high laser energy densities. The gradually
increased laser energy density results in reduction of average crater wear, approximately

18 % and 10 % in width and depth, respectively.

111

 

1422-94811m

250 pm

 

Figure 5- 7: Surface morphology and corresponding 3D topography of the craters, for
LSP with various laser energy densities, a) surface mo hology, 0. 6 J/cm2 ,b) 3D
topography, 0.6 J/cm2 , c) surface morphology, 2.3 J/cm , and d) 3D topography, 2.3 J/cm.

112

 

 

 

 

 

 

 

 

 

350 r . , . , r , . T .
300 -E _
E 250- .4
I E I i
200 - .
150 , . . T , e , . . .
0.0 0.5 1.0 1.5 2.0 2.5
Energy density (J/cm’) per Pulse
25.0 . , . , . , 1
225 r- .1
I 1
E 200 P- E u
i
a E .
17.5 ’- .1
150 L A l A l A l A l A
0.0 0.5 1.0 1.5 2.0 2.5
Energy Density (J/cmz) per Pulse

Figure 5-8: Plots for LSP showing laser energy density vs. (a) width and (b) depth of
craters at constant 30 pulses.

113

 

5.6. Discussion of results

The alumina coating layer was worn out at the initial stage of the turning test due
to the thin coating thickness (~lum) and high insert loading stress. Consequently, wear
in the substrate controlled rest of the crater wear. The effects of LSP on the alumina
coating layer disappeared when the substrate wear dominated the turning test. The
dispersion of the measured crater dimensions is presumably caused by the substrate wear,
which has 6 times faster in wear rate than that of coated wear [36, 76]. In future
investigations, these factors will be considered to minimize the deviation in wear process.
The processing parameters of the turning test will be adjusted by optimization of cutting
speed and insert loading thickness as well as the thickness of coating layer.

Despite of the improvement in residual stress, LSP offers another important
benefit. During LSP, the alumina surface that makes contact with the laser pulse ablates
even with the protective coating. As suggested in [76], the friction between tool and work
materials can be so intense that the rough surfaces can be sheared and lifted off the tool
material during chip formation. This is typically considered to be ‘cut-in’ wear, which
occurs at the initiation of cutting. The smoother surface after LSP improve the wear resistance

with less friction between tool and work materials.

114

 

6. CONCLUSIONS

The central idea of this work was to predict wear on the two major wear surfaces
of an insert: the flank and crater surfaces. In the flank wear, abrasion is the predominant
wear mechanism and developed the flank wear rate model which is the functions of
temperature, concentration of abrasive, and temperature dependent hardness of tool and
work materials. To verify the developed wear model experimentally, turning experiments
were done with various plain carbon steels at various cutting speeds using a variety of
coated tools. However, the experimental flank wear did not correspond to the wear
models. The flank wear data shows that the flank wear rate is not a function of the
abrasive (cementite in plain steel) concentration especially at high cutting speed. This is
caused by the phase transformation during cutting process. When the cutting temperature
in the flank interface is subjected to high enough temperatures, the pearlitic structure
(ferrite and cementite phases) austenizes. The machined surfaces after turning tests with
AISI 1045, 1070, and 4340 steels were experimentally observed using XRD, SEM and
TEM analysis. These surfaces showed evidence of austenitic transformation as the work

materials traverse across the flank wear land.

The exact location of the phase transformation cannot be determined easily because
it requires accurately modeling'the extent of flank wear land, the interactions such as the
friction between tool and work material and the effect of the primary shear zone. Only

about a IO-micron thick section of the work material undergoes austenitic transformation.

115

 

Even though phase transformation may not completely eliminate any abrasion action, it
directly affects the extent of flank wear.

On the other hand, crater wear occurs predominantly due to dissolution wear and
abrasive wear. Abrasive and chemical dissolution models were presented in an attempt to
explain the crater wear. TiN, alumina, and TiCN coating tools were turned with various
cutting conditions to verify the wear mechanisms of crater wear and the results were
compared with the analytical wear model. The experimental results with TiN and TiCN
coating tools agree well with the developed model. However, the results with alumina
coated inserts deviated from the model. This was due to other wear mechanism affected
in the turning test. With the microstructure of crater area, there exits pull-out grains in
CVD alumina coating tools and this phenomenon reduces the wear resistance of them.
However, the developed crate wear model can be readily applied to nitride and carbide
coated materials as the dissolution wear controls the wear process.

Based on the observation we made using SEM and x-ray diffraction, we have
elucidated the physics of chip flow as it traverses the rake face of an insert. The
phenomenon beneath the chip has been explained with the layers extruding out from the
stagnation zone in front of the cutting edge radius. These layers traverse the crater by
local convection between the chip and the rake face of the inserts. This provides the
supporting evidence that crater wear is directly related to abrasion and chemical
dissolution.

To improve the wear resistance of CVD coated alumina tool inserts, laser shock
processing (LSP) has been applied to locally relieve the deleterious tensile residual

stresses on the wear area of a coated tool. Residual stress levels of LSP CVD alumina-

116

 

coated carbide insert were evaluated by the X-ray diffractometer. Based on these results,
the residual stress decrease exponentially with LSP parameters such as a number of laser
pulses and laser energy density. From the relationship between LSP parameters and the
residual stress behavior, an empirical equation was derived. When residual stresses are
relieved by LSP, it improved the hardness value of the coating. Consequently the wear
behavior of alumina coated insert after LSP has improved. LSP achieved the reduction in
residual stresses and the better surface finish of the coating. Crater wear results show that

the wear resistance increases with LSP treated tool inserts by up to 21 %.

117

 

 

APPENDIX

A1. Abrasive model: two body and three body model

The development of an abrasive model requires the quantitative wear model to
describe the wear process, and the identification of parameters to be included in the
model. Both 2-body and 3-body models have been generally accepted as mechanical
abrasive wear mechanisms. In 2-body wear, the hard inclusions penetrate into the tool
surface and scrap out the materials. While the trapped hard inclusions roll and slide on

the tool wear surface in 3-body wear (Figure A-l).

118

 

 

Work 2

        

_ 1 :' Material
~t—-\
Chip
Tool
Surface
(a) 2-body Wear (b) 3-body Wear

Figure A-l. The schematic diagrams of the difference between 3-body and 2-body
abrasive wear.

119

 

The empirical quantitative 3-body model was firstly introduced by Rabinowicz
[21, 22]. It can bring out the parameter dependencies and make quantitative prediction on

the abrasive wear rate (Equation A. 1 . 1 ).

 

than19 P
= , -L < 0.8
3hody 3P! 0
L 19 P ’2" P
3W, = 3139— —'— , 0.8 < ——’— < 1.25 Equation A. 1.1
' 531’, Pa Pa
_ thant9 _13_ "’ 1’; >1 25
3b0dy 2.43Pl Pa 9 Pa °

where V is the volume wear rate, I is a sliding distance, L is the applied normal force
between the surfaces, tan6 is the average tangent of roughness angle of the abrasive
grains (a measure of the particle shape or sharpness), P1 is the hardness of tool and P8 is
the hardness of inclusions (measured on a bulk sample).

The implication of the 3-body wear model is that the second—phase (abrasives) in
a work material can have a relative motion against matrix phase (Figure A-l).
Rabinowicz’s model has been supported by a great number of experimental results [58,
79, 91, 107]. In the 3-body wear model, the second-phase (inclusions) in a work material
can have a relative motion against matrix phase in machining aluminum matrix
composites [108]. While turning aluminum matrix composites (40 vol% SiC particles in
an A1 matrix) with various tools with varying degree of hardness, the hardness ratio
between reinforcing phase and insert is found to be the most important factor in flank

wear [79]. In pearlitic steels, due to the complex morphology, the second phases are

120

 

 

constrained within the matrix of ferrite. The wear-rate governing relation for two-body
abrasion [22, 109] is

Ltant9
IzP

I

 

V2_,,m,y = x Equation A.1.2

where x is sliding length, L is the load between interacting surfaces, 6 is the half angle of
the conical abrasive particles (Figure A-la), Pt is the hardness of the abraded surface. In
the ferrous materials, the hardness ratio between cementite and ferrite ranges from a
value of 10.8 at room temperature to 4.5 at 700°C [91]. Even though the hardness of the
coatings is always higher than the hardness of the bulk cementite in an isothermal
condition, the temperature of the cementite during machining does not quite reach the
temperature at the flank surface because the cementite phase undergo only transient

heating [19]. Therefore, it is critical issue to measure the transient temperature during the

cutting process.
The hardness of ceramic coatings is degraded with the elevating temperature during

cutting process. The hardness as a function of temperature can be represented using an

exponential form;

Ha) =H0e'“T Equation A. 1 .3
where T in °C. Ho and or are constants obtained from a curve fitting process on the data.
The hardness data of the coatings and cementite were also collected from the literature
[82, 88]. This temperature-dependent hardness will be applied to three coatings, TiN,

TiCN and Alumina to verify the 3-body and 2-body wear models experimentally.

Because the cutting L and 9 values are not known for machining, the relative volumetric

121

 

wear rates are calculated using computer algorithm (Figure A-2).

122

 

Theoretical z-Body Abusive Wear Rate
61503 V——”—‘—*‘——r

 

 

700 900 1 100 1300
Temperature (°C)

Figure A-2: The theoretical the 2-body and 3-body abrasive wear models on selected
coated inserts using the computer algorithm [56].

 

123

 

This relative wear results can give a general ranking of the coating tools. As
shown in Figure A- 2, the wear rates were shown to decrease with the temperature for 3-
body abrasive wear model while the rates increase with the flank temperature for 2-body

wear model.

A.2 Temperature at the sliding interface

The temperature at the surface of workpiece is elevated during the cutting process
due to the work done by the external force: plastic deformation and tool-flank rubbing.
The temperature change at a machined area is a function of the depth of cut, cutting
speed, and material properties (specific heat and thermal conductivity). When the depth
of cut is large and the applied normal stress at the interface is large, the heat transfer
along the surface can be assumed to be one-dimensional [3]. From Jaeger’s general
solution for the temperature at sliding contact with moving heat sources, the change of
temperature can be determined on the assumption of the sudden release of heat Q in an
infinite solid [110]. Let the heat source the band of tool-workpiece contact as shown
Figure A-3, the steady-state temperature due to a band heat source in the surface (2 = 0)

of the semi-infinite solid (2 < 0) is given by Equation A2] with the assumption of no

heat loss from the plane 2 = O.

 

Equation A2]

I r 2 2
Tzqudtexp _(x—x +Vt) +2
27zkt 4a!

where T is temperature, t is time, k is thermal conductivity, and or is thermal diffusivity.

To determine the temperature at origin of heat source (t = 0) for a contact length 21 (see

Figure A-3) which has been moving for an infrnite time (steady state), Equation may be

124

 

 

integrated with respect to x' from —1 to l and the solution may is given by Equation A22

[3].

T = 2_qg_ +14K0(Z2 + n2)exp(—n)dn Equation A.2.2

flkV X—L

where K0 is the modified Bessel ftmction of the second kind and the dimensionless

quantities are defined as

Vx V2 V1

X=— - —
2a

‘5 a

 

125

 

4 i ’v

. ------ M3
t=0

Source of origin

 

 

 

      
   

 

 

Figure A-3:Geomet13I of band source [3].

126

To simplify equation A.2.2, approximate equations for the mean temperature rise at
high sliding speeds are given in Equation A23 [1]

E

T z %I.6q—I( )7”2 Equation A23
(1

k

The temperature on the surface of workpiece is not possible during the cutting process,
instead the analytical estimation obtained from the above will can be used. The heat

source is defined as [6]
q = 13W Equation A.2.4
where B is the ration of the nominal heat into the work materials to the cutting energy

(~0.5) and u is the specific cutting energy which is assumed to be a property of the

material and contact conditions.

Using Equation A23 and 4, the temperature of contact region can be evaluated and

given in Figure A-4.

127

 

 

1300

 

1200
1100 1
1000 I

900 .

Temperature

800 .

700 .

600

 

 

500

 

75 100 125 150 175 200 225 250 275 300
Cutting speed

Figure A-4: The temperature of the sliding surface from equation A.3.3 and 4(solid line,
cutting energy is available in the literature [111] and the temperature of crater area
measured by the pyrometer systems(dots).

128

 

The flank temperature cannot be estimated in direct method. However, the flank
temperature was assumed to be the sliding temperature and will used to predict the flank
wear rate. With the relationship of the temperature of crater area and the sliding surface
(Figure A-4), the following relationship between flank and crater temperature can be

obtained.

Tflank : 7“crater - (1 77 + O'ZSZrarer) Equation A3 5

A.3 Computation of the free energy of formation of TiCN

A3.]. Balinit B2 coating of Balzers

The TiCN coating was a commercial coating designated Balinit B. Though it is common
to represent the compound as TiCN, in reality, it exists as TiCanx depending on the
process parameters such as the N2 partial pressure in the PVD coating process. Hence, the
free energy of formation is a separate calculation for the purpose of computing the
dissolution wear. It can be seen in [112] that a mixture of TiC and TiN in the required
proportions can form TiCN. It is also known that they exhibit very good miscibility in
one another and the resulting mixture shows a color according to the final stoichiometry.
At a composition of TiCo,5No,5 the compound is gray with a bluish tinge, as was the case
for the coating in this work. One can therefore assume reaction in equation A41 for the

formation of TiC1-xNx. As a result, the free energy of formation of TiC1-xNx is given by

equation a.4.2.

(1 —x)<TiC> +xmv”<n-c._,1v,> A.4.l
(_.

 

2 Coutesy Balzers, Inc.

129

AG °mc N > = (1- x) AG,°<T1CN> + mafiams
f H x + RT(anx + (l — x)Ln(1 — x))

A.4.2
Essentially, this would mean a neglect of the excess flee energy of solution of the
respective components in the formation of the final compound. Since, the exact
composition itself was speculative, as known from Balzers, Inc., this simple assumption

would suffice for the computation in this context. For a more detailed study one is

referred to the literatures [1 12-1 14].

A.3.2. Final Form for the Free energy of formation for TiC(l-x)Nx

As described above, the correct composition of TiC(1.,,)Nx cannot be available. For
this study, x = 0.5 is acceptable because the color shows gray with a bluish tint whose
composition is closed to TiCo_5No,5. Using Equation A42 and forrnaition energy of TiN

and TiC [115], the formula for the free energy of formation of TiCo_5No,5 can be written

as:

A031.“ = -62000 + 10.931“ T < 1155 °K
= —62725 +11.81T 1155 °K < T <1900°K

A.3.3. Molar Volume of TiC1.,‘Nx

The molar volume can be obtained as a linear interpolation between the two
values. TiN and TiC have the same crystal structure (cubic) and also have comparable
molar volume values and have similar thermodynamic property such that they have same
transition temperature of formation at 1155 °K [115]. Thus this linear interpolation value

can be acceptable and the molar volume of TiCN is given as

130

MVTK]_,N, = (1' x)MVTi(‘ + xMVTiN

A.4.3
= 0.5(11.49)+ 0.5(12.20)=11.85

In this study, this molar volume of 11.85 (cm3/mol) is used to calculate the dissolution

wear for TiCo_5No,5 coating inserts.

A.4. Estimation of the molar excess free energy

The partial excess free energy of solution of a tool material should be estimated to
obtain the solubility in steel. All tool materials are assumed to obey Henry’s law because
an amount of tool material dissolved into chip during the cutting process is small enough
to be considered as dilute solution. Therefore, the partial molar free energy of solution of

the component i can be written by

A6,.“ = RTln y, + RTln c, = 1:15,.” + RTlnc, Equation A4.1
where y is activity coefficient and c is the concentration of i and ACT,” is a constant in the
low limit of solubility.
From equation A4. 1, the excess free energy of component i in steel, AC7,” , can be
estimated. For example, the solubility of the Ti in y—iron at 1373 is

A5,? = A5,? + RT ln c,.,. = —14970 = A5,? + 1.938 x 1373ln 0.0076
A5,? = -1657.7 cal/mole

where the molar partial free energy of Ti is 14970 cal/mole [115] and the solubility of Ti

in y—iron (0.76 atomic percent) is obtained from the literature [116-118].

The estimated values of excess free energy of solution of tool component in y-iron

(austenite) are listed in Table A1.

131

 

Table A1: The estimated values of excess free energy of solution of tool component in 7-

iron (austenite).

 

Tool

 

 

 

 

 

 

 

 

 

 

Temperature Solubility partial molar free A5,.“ (cal/mole)
material (K) [1 16- [116-1 18] energy of solution

118] oftool [115]
Ti 1373 0.0076 -14970 -1657.695197
N 869 0.0235 -35013.7 -28537.26033
C 1523 0.0267 -3460 7504.203 767
A1 1423 0.02 -26800 -15738.75104
O 1673 0.032 -23568.85 -12126.71365

 

132

 

REFERENCES

>19

10.

11.

12.

13.

A. Mallock, “The Action of Cutting Tools,” Proc Royal Soc (Lon), vol. 33, pp.
127-139, 1881.

V. Piispanen, “Theory of Formation of Metal Chips,” Journal of Applied Physics,
vol. 19, pp. 876-881, 1948.

NR Suh, Tribophysics. New Jersey: Prentice-Hall, Inc, 1986.

S. Kalpakjian, Manufacturing Engineering and Technology, 3rd ed. New York:
Addison-Wesley Publishing Company, 1995.

D.M. Turley and ED. Doyle, Microstrucural Behanvior-Its Influence on
Machining, vol. 7: ASME PED, 1982.

MC. Shaw, Metal Cutting Principles. Oxford: Clarendon Press, 1992.
E.J.A. Annarego and RH. Brown, The Machining of Metals. New Jersey:
Prentice-Hall, Inc., 1969.

GP. Michetti, A. DeFilippi, and R. Ippolito, “Tool Wear and Cutting Force in
Steel Turning,” Proceedings of the CIRP International Conference on
Manufacture Technology, pp. 513-524, 1967.

S.-S. Cho and K. Komvopoulos, “Cutting Force Variation due to Wear of Multi-
Laye Ceramic Coated Tools,” Journal of Tribology, vol. 120, pp. 75-81, 1998.

ED. Trent and PK. Wright, Metal Cutting. Boston: Butterworth Heinemann,
2000.

R. Komanduri, “Machining and Grinding: A Historical Revied of the Classical
Papers,” Appl. Mech Rev, vol. 46, no. 3, pp. 80-132, 1993.

DE. Buryta and R. Sowerby, “Experimental Determination of Rake Face Stress
Distributions,” Key Engineering Materials, vol. 138-140, pp. 57-125, 1998.

H. Ernst and ME. Merchant, “Chip foramtion, Friction, Surface Treatment of

Metals,” Transactions of the American Society for Metals, vol. 29, pp. 299-378,
1941.

133

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

EC. Thomsen, “Critical Composition of Metal Cutting Theories with New
Experimental Data,” Journal of Engineering for Industry, 1960.

ME. Merchant, “Mechanics of the Metal Cutting Process. II.Plasticity Conditions
in Orthogonal Cutting,” Journal of Applied Physics, vol. 16, pp. 318-324, 1945.

ME. Merchant, “Mechanics of the Metal Cutting Process. I Orthogonal Cutting
and a Type 2 Chip,” Journal of Applied Physics, vol. 16, pp. 267-275, 1945.

S. Ramalingham, Y.I. Peng, and J .D. Watson, “Tool Life distribution, Part 3.
Mechanisms of Single Injury Tool Failure and Tool Life Distribution in

Interrupted Cutting,” Journal of Engineering For Industry, Transactions of the
ASME, vol. 100, pp. 193-200, 1978.

S. Ramalingham and J. D. Watson, “Tool Life distribution, Part4: Minor Phases
in Work Material and Multiple-Injury Tool Failure,” Journal of Engineering For
Industry, Transactions of the ASME, vol. 100, pp. 201-209, 1978.

S. Ramalingham and W. P. K., “Abrasive Wear in Machining: Experiments With
Materials of Controlled Microstructure,” Journal of Engineering Materials and
Technology, vol. 103, pp. 151-156, 1981.

V. Marinov, “Experimental Study of the Abrasive Wear in Metal Cutting,” Wear,
vol. 197, pp. 242 - 247, 1996.

E. Rabinowicz, “Abrasive Wear Resistance as a Materials Test,” Journal of
American Society of Lubrication Engineering, vol. 33, no. 7, pp. ‘378 - 381, 1977.

E. Rabinowicz, L.A. Dunn, and PG. Russell, “A Study of Abrasive Wear Under
Three-Body Conditions,” Wear, vol. 4, pp. 345-355, 1961.

ISO Standard 3 685-1 977 (E): International Organization for Standardization,
1977.

NH. Cook and B. Bhushan, “Sliding Surface Interface Temperatures,” Journal of
Lubrication Technology, vol. 95, no. 1, pp. 59-64, 1973.

N.P. Suh, “The Delamination Theory of Wear,” Wear, vol. 25, pp. 111-124, 1973.

J .J . Pamies-Teixeira, N. Saka, and NP. Suh, “Wear of Copper-Based Solid
Solutions,” Wear, vol. 44, pp. 65-75, 1977.

S. J ahanmir, N.P. Suh, and ER Abrahamson, “Microscopic Observations of the
Wear Sheet Formation by Delamination,” Wear, vol. 28, pp. 235-249, 1974.

134

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

NH. Cook and RN. Nayak, “Development of Improved Cutting Tool Materials,”
Technical Report of AFML-TR-69-I 85, U. S. Air F orce Materials Laboratory,
1969.

P. Hartung and BM. Kramer, “Tool Wear in Titanium Machining,” CIRP Annals,

International Institution for Production Engineering Research, vol. 31, pp. 75-80,
1982.

BM. Kramer and NP. Suh, “Tool Wear by Solution: A Quantitative

Understanding,” Journal of Engineering for Industry, vol. 102, pp. 303-339,
1980.

BM. Kramer and PK. P. Kwon (formally Judd, “Computational Design of
Coating Materials,” Journal of Vacuum Science and Technology A, vol. 3, no. 6,
pp. 2349-2444, 1985.

BM. Kramer, “Predicted Wear Resistance of Binary Coatings,” Journal of
Vacuum Science and Technology A, vol. 4, no. 6, pp. 2870 -2873, 1986.

BM. Kramer, “Tribological Aspects of Metal Cutting,” Journal of Engineering
for Industries, vol. 115, pp. 372-376, 1993.

S.V. Subramanian, S.S. Ingle, and D.A.R. Kay, “Design of Coatings to Minimize
Tool Crater Wear,” Surface Coating Technology, vol. 61, pp. 293-299, 1993.

D.S. Rickerby and A. Matthews, Advanced Surface Coatings. UK: Glagow,
1991.

E. Sirvio, M. Sulonen, and H. Sundquist, “Abrasive Wear of Ion-Plated Titanium
Nitride Coatings on Plasma-Nitrided Steel Surface,” Thin Solid Films, vol. 96, p.
93, 1982.

D.S. Rickerby and RI. Burnett, “The Wear and Corrosion Resistane of Hard PVD
Coatings,” Surface and Coatings Technology, vol. 33, pp. 191-211, 1987.

S.K. Ghosh and MS. Hohler, “Study of the Relative Wear and Abrasion
Resistance Ti(C,N) and TiN Coatings,” Surface and Coatings Technology, vol.
54, pp. 466-469, 1992.

B. Lux, C. Colombier, H. Altena, and K. Stjemberg, “Preparation of Alumina

Coatins by Chemical Vapour Deposition,” Thin Solid Films, no. 138, pp. 49-64,
1986.

K. Holmberg and A. Mathews, Coatings Tribology. The Netherlands: Elsebier,
1994.

135

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

E.G. Loewen and MC. Shaw, “On the Analysis of Cutting-Tool Temperatures,”
ASME Transactions, vol. 76, pp. 217-223, 1954.

K.J. Trigger and B.T. Chao, “An Analytical Evaluation of Metal Cutting
Temperatures,” ASME Transactioins, vol. 73, pp. 57-68, 1951.

AD. Tay, M.G. Stevenson, G.d.V. Davis, and P.L.B. Oxiey, “A Numerical
Method for Calculating Temperature Distributions in Machining, from Force and
Shear Angle Measurements,” Int. J. Mach. Tool Des. Res, vol. 16, pp. 335-349,
1976.

PD. Muraka, G. Barrow, and S. Hinduja, “Influence of the Process Variables on
the Temperature Distribution in Orthogonal Machining Using the Finite Element
Method,” Int. J. Mech. Sci, vol. 21, pp. 445-456, 1979.

MG. Stevenson, P.K. Wright, and J .G. Chow, “Further Developments in
Applying the Finite Element Method to the Calculation of Temperature
Distributions in Machining and Comparisons with Experiment,” ASME J. Eng.
Ind, vol. 105, pp. 149-154, 1983.

AD. Tay, M.G. Stevenson, and G.d.V. Davis, “Using the Finite Element Method
to Determine Temperature Distributions in Orthogonal Machining,” Proc. Instn.
Mech. Engrs., vol. 188, pp. 627-638, 1974.

G. Boothroyd, “Temperatures in Orthogonal Metal Cutting,” Proc. Inst. Mech.
Eng, vol. 177, pp. 789-802, 1963.

G. Boothroyd, “Photographic Technique for the Determination of Metal Cutting
Temperatures,” Brit. J. Appl. Phys, vol. 12, pp. 238-242, 1961.

H. Shore, “Thermoelectric Measurement of Cutting Tool Temperature,” J.
Washington Acad. Sci, vol. 15, pp. 85-88, 1925.

J. Lin, S.-L. Lee, and C.-I. Wang, “Estimation of Cutting Temperature in High
Speed Machining,” ASME J. Eng. Mat. T ech., vol. 114, pp. 289-296, 1992.

T. Ueda, A. Hosokawa, and A. Yamamoto, “Measurement of
GrindingTemperature Using Infrared Radiation Pyrometer With Optical Fiber,”
ASME J. Eng. Ind, vol. 108, pp. 247-251, 1986.

J .P. Kottenstette, “Measuring Tool-Chip Interface Temperatures,” ASMEJ. Eng.
Ind, vol. 108, pp. 101-104, 1986.

BM. Kramer, “An Analytic Approach to Tool Wear,” Ph. D. thesis,, MIT,
Cambridge, Mass, 1979.

136

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

NP. Suh and PD. Fillion, “Optimization of Cutting Tool Properties Through the
Development of Alumina Cerrnet,” Wear, vol. 62, pp. 132-137, 1980.

ED. Case, “Private Communication,” , 1996.

P. Kwon, 2000, ",", “Predictive Models for Flank Wear on Coated Inserts,”
Journal of Tribology, vol. 122, no. 1, pp. 340-347, 2000.

H. Opitz and M. Gappisch, “Some Recent Research on the Wear Behavior of
Carbide Cutting Tools,” Int. J. Mach. Tool Des. Res, vol. 2, pp. 43-73, 1962.

AM. Shelbourn, W.T. Roberts, and EM. Trent, “Structures of Machined Steel
Chip,” Materials Science and Technology, vol. 1, pp. 220-226, 1985.

J .L. Hau-Bracamonte and M.L.H. Wise, “Austenization of Steel during Chip
Formation,” Metallurgical Transactions, vol. 14A, pp. 1743-1745, 1983.

J .L. Hau-Bracamonte, “Partial Austenization within Flow Zone When Cutting a
Low-carbon Steel,” Metals Technology, vol. 8, pp. 447-450, 1981.

K. Ramanujachar and S.V. Subramanian, “Micromechanics of Tool Wear in
machining Free Cutting Steels,” Wear, vol. 197, pp. 45-55, 1996.

Y. Matsumoto, M.M. Barash, and CR. Liu, “Cutting Mechanism During
Machining of Hardened Steel,” Materials Science and Technology, vol. 3, pp.
299-305, 1987.

B.J. Griffiths, “White Layer Formations at Machined Surfaces and Their
Relationship to White Layer Formation at Worn Surfaces,” Journal of Tribology,
vol. 107, pp. 165-171, 1985.

H.K. Tbnshoff, H.-G. Wobsker, and D. Brandt, “Hard Turning - Influences on the

Workpiece Properties,” Transactions of NAMRI/SME, vol. XXIII, pp. 215-220,
1995.

H.K. Tbnshoff, H-G. Wobsker, and D. Brandt, “Tribological Aspect of Hard

Turning with ceramic Tools,” Lubrication Engineering, vol. 51, no. 2, pp. 163-
168, 1995.

MC. Shaw and A. Vyas, “Heat-Affected Zones in Grinding Steel,” Annals of the
CIRP, vol. 43, no. 1, pp. 279-282, 1994.

R. Snoeys, K.U. Leuven, M. Maris, W. N.F., P. J ., and K.U. Leuven, “Thermally

Induced Damage in Grinding,” Annals of the CIRP, vol. 27, no. 2, pp. 571-581,
197 8.

137

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

P. Kruth, L. Stevens, L. F royen, and B. Lauwers, “Study of the White layer of a
Surface Machined by Die-Sinking Electro-Discharge Machining,” Annals of the
CIRP, vol. 44, no. 1, pp. 165-171, 1995.

Y.K. Chou and G]. Evans, “Process Effects on White Layer Formation in Hard
Turning,” Transactions of NAMRI/SME, vol. 26, pp. 117-122, 1998.

Y.K. Chou and G]. Evans, “White layer and thermal modeling of hard turned

surface,” International Journal of Machine Tools and Manufacture, vol. 39, pp.
1863-1881, 1999.

S. Akcan, S. Shah, S.P. Moylan, P.N. Chhabra, S. Chandrasekar, and TN. Farris,
“Characteristic of White Layers Formed in Steels by Machining,” Proceedings of
ASME, IMECE, Nashville, TN, 1999.

S. Kim and DR. Durham, “Microscopic Studies of Flank Wear on Alumina
Tools,” Journal of T ribology, vol. 113, pp. 204-209, 1991.

PA. Deamley, “Rake and Flank Wear Mechanisms of Coated and Uncoated
Cemented Carbides,” Transactions of the ASME Journal of Engineering
Materials and Technology, vol. 107, pp. 68-82, 1985.

P. Kwon and R.K. Kountaya, “Experimental Observation on Machining
Spherodized Plain Carbon Steels,” , vol. 42, pp. 265-272, 1999.

W.D.J. Callister, Materials Science and Engineering: An Introduction, , 3rd ed.
New York, NY: John Wiley & Sons, Inc., 1994.

J .D. Byrd and BL. Ferguson, “A study of the influence of hard inclusions on
carbide tool wear utilizing a powder metal technique,” Proceedings of the VIth
NAMRC, pp. 310-315, 1978.

MK. Brun, Lee, M. and F. Gorsler, “Wear Characteristics of Various Hard
Materials for Machining SiC-Reinforced Aluminum Alloy,” Wear, vol. 104, pp.
21-29, 1985.

G. Dixon, W.R. N, and M. Lee, “Process Involved in the Wear of Cemented
Carbide Tools,” Wear, vol. 104, pp. 157-171, 1985.

T. Akasawa and H. Hishiguti, “Crater Wear Mechanism of WC-Co Tools at High
Cutting Speeds,” Wear, vol. 65, pp. 141-150, 1980.

P.K. Kwon, “Theoretical Foundations for the Optimization of Ceramic Coated
Tools,” Thesis” MIT, 1985.

138

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

EM. Trent, “Cutting steel and iron with cemented carbide cutting tools Parts I, II
and 111,,” Journal of the Iron and Steel Institute, vol. 201 , pp. 847-855, 1963.

Y. Naerheim and EM. Trent, “Diffusion Wear of Cemented Carbide Tools When
Cutting Steel at High Speeds,” Metals Technology, vol. 4, pp. 548-556, 1977.

J .P. Chubb and J. Billingham, “Coated Cutting Tools- A Study of Wear in the
Turing Process,” Wear, vol. 61, pp. 283-293, 1980.

I.Y. Ham and Y. Narutaki, “Wear Charateristics of Ceramic Tools,” Transactions
of the ASME, pp. 951-959, 1973.

G. Brandt, “Flank and Crater Wear Mechanisms of Alumina-Based Cutting Tools
When Machining Steel,” Wear, vol. 112, pp. 39-56, 1986.

A.T. Santhanam and D.T. Quinto, Surface Engineering of Carbides, C ermet and
Ceramic Cutting Tools, vol. 5, 1994.

A.T. Santhanam, Private Communication: Kennametall., 1997.

P.K. Mehrotra and D.T. Quinto, “High Temperature Microhardness Profiles of
Hard CVD Coatings,” , pp. 639-657., 1985.

KB. Gove and J.A. Charles, “The High Temperature Hardness of Phases of
Steels,” Metals Technology, vol. 95, pp. 279-283, 1974.

BM. Kramer, “On tool materials for high speed machining,” Journal of
Engineering for Industry, vol. 109, pp. 87-91, 1987.

D.W. Yen and PK. Wright, “A remote temperature sensing technique for

estimating the cutting interface temperature distribution,” Journal of Engineering
for Industry, vol. 108, pp. 252-263, 1986.

N. Cook, “Tool Wear and Tool Life,” Journal of Engineering for Industry, vol.
95, pp. 931-938, 1973.

SS. Ingle, “The Micromecharrisms of Cemeted Carbide Cutting Tool Wear,” ,
MaMaster University, 1993.

H. Czichos, Tribology. Amsterdam: Eleseviser, 1978.

SR. Lampman, ASM Handbook. Materials Park, OH: American Society of Metals
International, 1994.

139

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.
106.

107.

108.

D.T. Quinto, G.J. Wolfe, and RC. Jindal, “High Temperature Microhardness of
Hard Coatings Produced by Physical and Chemical Vapor Deposition,” Thin Solid
Films, vol. 153, pp. 19-36, 1987.

D.T. Quinto, “Mechanical Property and Structure Relationships in Hard Coatings

for Cutting Tools,” Journal of vacuum science and technology, vol. A6, no. 3, pp.
2149-2157, 1988.

D. Devaux, R. Fabbro, and L. Tollier, “Generation of Shock Waves by Laser-
Induced Plasma in Confined Geometry,” Journal of Applied Physics, vol. 74, no.
4, pp. 2268-2273, 1993.

M. Gerland, M. Hallouin, and H. Presles, “Comparison of Two New Surface
Treatment Processes, Laser-Induced Shock Waves and Primary Explosive:

Application to Fatigue Behavior,” Materials Science and Engineering, vol. A156,
pp. 175-182, 1992.

R. Fabbro, J. Foumier, P. Ballard, and D. Devaux, “Physical Study of Laser-

Produced Plasma in Confined Geometry,” Journal of Applied Physics, vol. 68, no.
2, pp. 775-784, 1990.

K. Mukherj ee, T.H. Kim, and WT. Walter, “Shock Deformation and
Microstructural Effects Associated with Pulse Laser-Induced Damage in Metals,”
The Metallurgical Society of AIME, 1981, (K. Mukherjee and J. Mazumder, Eds.),
pp. 137-150, Chicago, IL: , 1981.

P. Peyre, R. F abbro, P. Merrien, and HP. Lieurade, “Laser Shock Processing of
Aluminum Alloys. Application to High Cycle Fatigue Behavior,” Materials
Science and Engineering, vol. A210, pp. 102-113, 1996.

P. Peyre, R. Fabbro, L. Berthe, and C. Cubouchet, “Laser Shock Processing of
Materials, Physical Processes Involved and Examples of Applications,” Journal
of Laser Applications, vol. 8, pp. 135-141, 1996.

CE. Morosanu, Nucleation and Growth of C VD F ilms. Amsterdam-Oxford-New
York-Tokyo: Elsevier, 1990.

P.L.B. Oxiey, The Mechanics of Machining. Chicester: Ellis Horwood, 1989.

L. Iuliano, L. Settineri, and A. Gatto, “Composites Part A,” , pp. 1501-1509: ,
1998.

MA. Moore, “A Review of Two-Body Abrasive Wear,” Wear, vol. 27, pp. 1 - 17,
1974.

J .C. J aeger, “Moving Sources of Heat and The Temperature at Sliding Contacts,”
Proceedings of the Royal Society, N. S. W., vol. 56, pp. 203-224, 1942.

140

109. W.R. Devires, Analysis of Material Removal Processes. New York: Springer-
Verlag, 1991.

110. R. Kiffer, E. P., and M. Freudhofrneier, About Nitrides and Carbonitride-based
Cemented Hard Alloys, in Mordern Developments in Powder Metallurgy. New
York: Plenum Press, 1971.

11 l. H. Pastor, “Titanium Carbonitride-Based Alloys for Cutting Tools,” Materials
Science and Engineering, vol. A105/106, pp. 401-409, 1988.

112. K.I. Portoni and Y.V. Levinskii, “High Temperature Equilibrium of the Reaction
Between Titanium Nitride and Carbon,” Russian Journal of Physical Chemistry,
vol. 37, pp. 1424-1428, 1963.

113. O. Kubaschewski, E.L. Evans, and CB. Alcock, Metallurgical Thermochemistry,
4 ed. Oxford: Pergamon Press, 1967.

114. R. Elliot, Constitution of the Binary Alloys, First Supplement. New York:
McGraw-Hill, 1965.
115. M. Hansen, Constitution of the Binary Alloys. New York: McGraw-Hill, 1965.

116. F. Shunk, Constitution of the Binary Alloys, First Supplement. New York:
McGraw-Hill, 1969.

141

 

1113111111111111‘111‘111111