FUNDAMENTAL TOOL WEAR STUDY IN TURNING OF Ti-6Al-4V ALLOY (Ti64) AND
NANO-ENHANCED MINIMUM QUANTITY LUBRICATION (MQL) MILLING

By

Trung Kien Nguyen

A DISSERTATION

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

Mechanical Engineering - Doctor of Philosophy

2015

ABSTRACT
FUNDAMENTAL TOOL WEAR STUDY IN TURNING OF Ti-6Al-4V ALLOY (Ti64) AND
NANO-ENHANCED MINIMUM QUANTITY LUBRICATION (MQL) MILLING

By

Trung Kien Nguyen

Titanium (Ti) alloy, in particular Ti-6Al-4V (Ti64), has been widely used in a variety of
industries such as automobile, aerospace, chemistry, biomedicine and other
manufacturing industries because of their desirable and unique mechanical properties.
The well-known properties of Ti alloys include light-weight, excellent strength even at
elevated temperatures, resistance to corrosion and biocompatibility, which cannot be
collectively and comprehensively satisfied by any other alloys in some applications. In
machining of Ti alloys, however, the low thermal conductivity and high hardness
exposes cutting tools to high temperatures and cutting forces, which often fracture the
cutting tools catastrophically. More importantly, the high chemical solubility of cutting
tools causes the high chemical wear leading to accelerated wear on cutting tools,
especially when cutting at high speeds. Polycrystalline diamond (PCD) and uncoated
carbide tools are the most widely used tool materials for machining Ti alloys. In order to
find the main reason for this puzzling behavior, this study revisits the fundamental wear
mechanisms in rake and flank faces using PCD and carbide tools in dry turning of Ti64
alloy. The original microstructure of work material was characterized using Orientation
Image Microscope (OIM) to explain the correlation of the wear pattern with the observed
microstructure. Based on the microstructure and the tool wear patterns, this study

claims that wear damages are caused primarily by the heterogeneity coming from not
only the presence of both α (hexagonal closed packed) and β (body centered cubic)
phases but also the hard orientation of the α−phases. In addition to the heterogeneities,
the adhesion layer detaching parts of the tool material also contributes to flank wear.
This thesis also considers improving tool life by adopting new lubrication techniques.
In particular, Minimum Quantity Lubrication (MQL)-based machining process was
chosen as it has many merits over not only conventional flood cooling machining but
also dry machining. However, few disadvantages make the MQL-based machining
process impractical to be adopted in many industrial production settings for more
aggressive cutting conditions. At high cutting speeds, for example, a minute amount of
oil used in MQL will simply evaporate or disintegrate as soon as the oil droplets strike
the tools already heated to high temperatures. Lamellar structured solid lubricants
(graphite and hexagonal boron nitride) in a platelet form have been mixed with a typical
vegetable MQL oils to mitigate this major deficiency of MQL process. When the mixture
of oil and these platelets are applied, the platelets are expected to provide additional
lubricity even after the oil droplets have been disintegrated at high temperature. Thus,
the enhancement achieved by adding these platelets allows us to expand the
processing envelope of MQL. In this research project, a comprehensive study on the
effect of the diameter and thickness of platelets was carried out. The results showed
that the presence of nano-platelets in the MQL oil decreased the tool wear and
improved the tool life compared to traditional MQL with pure oil as well as dry machining
1045 steel and Ti64 not only by providing lubricity at high temperature cutting condition
but also by reducing the micro-chipping and tool fracture.

To my wife, Ha Dao,
my children, Chi Nguyen, and An Nguyen,
and my beloved family

iv

ACKNOWLEDGMENTS

I would like to gratefully and sincerely thank my advisor, Dr. Patrick Kwon for his
expertise, generous support, valuable encouragement, and excellent guidance for me to
proceed through my doctoral studies and the completion of this dissertation. My deepest
gratitude also goes for his patience, enlightening advice and help in improving my
scientific and writing skills. I also wish to thank my committee members, Dr. Bieler, Dr.
Feeny and Dr. Baek, for their support, valuable comments, and guidance to improve
and complete this dissertation. My appreciation especially goes to Dr. Bieler for
providing equipment for experiment that supported for analyzing data. A special thanks
to Mr. Lars Haubold at Franhoufer CCL, USA for his generous help in doing the
experiments and making equipment available for use. I would like to thank Brian Hoefler
at Sandvik Coromant for providing tools for experiments. Many thanks go to Mr. Steve
Allen at West Michigan Precision Machining for help in experiments.
A special acknowledgement goes to Dr. Tim K. Wong, and Dr. Kyung H. Park for
their encouragement and instruction at the beginning of my graduate studies. Many
thanks go to my colleagues, Xin Wang, Wang Mingang, David Schrock, Truong Do,
Sirisak Tooptong, Dinh Nguyen for valuable discussion and general support as friends
and co-workers. Furthermore, I am appreciative of my co-workers Di Kang for
assistance with conducting some of the experiments contained in this work.
Most importantly, I wish to thank my wife, my children, my parents and my parents in
law for their love, support and encouragement that provided my inspiration and was my
driving force during my PhD study.

v

TABLE OF CONTENTS
Chapter 1: Tool wear of carbide and PCD inserts in turning of Ti64 ........................ 1
I.1
INTRODUCTION................................................................................................ 1
I.1.1
Machining Titanium overview ...................................................................... 1
I.1.2
Motivation .................................................................................................... 6
I.1.3
Tool materials and tool wear mechanisms reported in literature.................. 7
I.1.3.1
Dominant tool wear mechanisms .......................................................... 9
I.1.3.2
TiC protection layer ............................................................................... 9
I.1.3.3
Cobalt diffusion ................................................................................... 10
I.1.4
Phases and microstructure of Ti alloys ...................................................... 10
I.1.4.1
Phases in Titanium alloys ................................................................... 10
I.1.4.2
Phase transformation in Titanium alloys. ............................................ 17
I.1.4.3
Microstructure of α + β Ti alloys .......................................................... 18
I.2
EXPERIMENTAL SETUP AND PROCEDURES .............................................. 20
I.2.1
Turning experiments of Ti64 alloy.............................................................. 20
I.2.2
Confocal Microscopy ................................................................................. 23
I.2.3
SEM picture and element mapping............................................................ 27
I.3
EXPERIMENTAL RESULTS AND DISCUSSION ............................................ 28
I.3.1
Wear Characteristics of tools inserts at low DOC ...................................... 28
I.3.1.1
Crater wear at low DOC ...................................................................... 28
I.3.1.2
Nose wear at low DOC........................................................................ 30
I.3.2
Wear characteristics of tools inserts at high DOC ..................................... 41
I.3.2.1
Crater wear at high DOC..................................................................... 41
I.3.2.2
Nose wear and flank wear at high DOC .............................................. 52
I.3.3
SEM images and element mapping results ............................................... 59
I.3.4
Chip morphology ....................................................................................... 63
I.4
CUTTING TEMPERATURE PROFILES WITH FEM SIMULATION ................. 69
I.4.3
2D-FEM with the Johnson-Cook (JC) model ............................................. 69
I.4.4
2D-FEM JC model with heat transfer ......................................................... 71
I.4.5
Cutting temperature profiles with 2D-FEM simulation ................................ 73
Chapter 2: Evidence of phase change and root causes flank wear and scoring
marks with orientation imaging microscopy (OIM) .................................................. 80
II.1 OIM SETUP FOR CRYSTAL ORIENTATION IN TI64 ALLOY AND CHIPS .... 80
II.2 OIM RESULTS AND DISCUSSION ................................................................. 83
II.2.1 Evidence of phase change (α →β) in machining of Ti64 ........................... 83
II.2.2 Root causes flank wear and scoring marks in Ti alloys machining ............ 90
Chapter 3: Driven process of thermochemical wear in Ti alloys machining ....... 105
III.1 BACKGROUND ON WEAR MECHANISMS .................................................. 105
III.1.1 Mechanical wear...................................................................................... 106
III.1.2 Independent of dissolution and diffusion wear model by Kramer............. 109
III.1.2.1 Dissolution wear ................................................................................ 111

vi

III.1.2.2 Diffusion wear ................................................................................... 116
III.1.3 Upper bound of diffusion wear model ...................................................... 119
III.2 DRIVEN PROCESS OF GENERALIZED THERMOCHEMICAL WEAR ........ 123
Chapter 4: Tool wear improvement in machining of Steel AISI 1045 with micro and
nano-platelets enhanced MQL ................................................................................. 136
IV.1
INTRODUCTION ........................................................................................ 136
IV.1.1
Improved performance of MQL with alternative lubricants ................... 137
IV.1.2
Improved performance of MQL by adding solid lubricants ................... 139
IV.1.3
Improved performance of MQL by varying spray configuration ............ 144
IV.2
BACKGROUND .......................................................................................... 145
IV.2.1
Micro and nano-platelet characterizations ............................................ 147
IV.2.2
Vegetable oil Unist Coolube 2210 ........................................................ 153
IV.3
EXPERIMENTAL SETUP AND PROCEDURES......................................... 154
IV.3.1
Suspension Stability of Mixtures........................................................... 154
IV.3.2
Wetting angle measurement ................................................................ 155
IV.3.3
Surface characterization of two coated inserts using in tribotest and endball milling experiment .......................................................................................... 156
IV.3.4
Tribometer Tests .................................................................................. 157
IV.3.5
Ball Milling Experiments with steel AISI 1045 ....................................... 159
IV.4
EXPERIMENTAL RESULTS AND DISCUSSION ....................................... 163
IV.4.1
Stability of Mixtures .............................................................................. 163
IV.4.2
Wetting angle measurement ................................................................ 164
IV.4.3
Surface characterization of two coated inserts ..................................... 164
IV.4.4
Tribometer Tests .................................................................................. 166
IV.4.5
Tool Wear with Ball Milling Experiments .............................................. 171
IV.4.5.1 Optimal MQL spray angles for ball milling ......................................... 171
IV.4.5.2 Effectiveness of thickness and diameter of platelets to tool wear ..... 173
Chapter 5: ............ Tool wear improvement in machining of Ti64 with nano-platelets
enhanced MQL .......................................................................................................... 178
V.1 EXPERIMENTAL SETUP AND PROCEDURES ............................................ 178
V.1.1 Ball Milling Experiments with Ti64 ........................................................... 178
V.1.2 Tool wear measurements ........................................................................ 180
V.2 EXPERIMENTAL RESULTS AND DISCUSSION .......................................... 182
V.2.1 Flank wear and chipping on cutting edges............................................... 182
V.2.1.1 Flank wear at low cutting speed (2500rpm) ...................................... 182
V.2.1.2 Flank wear at high cutting speed (3500rpm) ..................................... 186
V.2.1.3 Micro-chipping and tool fracture ........................................................ 188
V.2.2 Nose wear of insert .................................................................................. 190
V.2.3 Crater wear .............................................................................................. 192
Chapter 6: Conclusions ............................................................................................ 194

vii

BIBLIOGRAPHY ......................................................................................................... 197

viii

LIST OF TABLES
Table 1: Slip systems in alpha and beta phase ............................................................. 15
Table 2: Tool grades and DOC used ............................................................................. 21
Table 3: Cutting time for the second set of experiment at high DOC (dc= 1.2 mm) ....... 22
Table 4: Cutting time for the first set of experiment at low DOC (dc= 0.635 mm) .......... 23
Table 5: The maximum crater wear depth (µm) of the first turning tests with dc=0.635
[Schrock, 2012; 2014] .................................................................................. 29
Table 6: Johnson-Cooks coefficients for Ti-6Al-4V ....................................................... 70
Table 7: The parameters of air at 20°C and 1atm ......................................................... 72
Table 8: Reynolds, Nusselt number, and heat transfer coefficients for three cutting
speeds ......................................................................................................... 73
Table 9: The area fraction of α-phase and the grain sizes of each phase in Bulk-Ti ..... 84
Table 10: The dimensions of the ‘hard’ α-cluster with Bulk-Ti ....................................... 94
Table 11: Typical damage on the flank face .................................................................. 97
Table 12: The hardness data of tool materials ............................................................ 108
Table 13: Estimated solubility of tool materials in α-Ti (at 800°C) and β-Ti (at 1000°C)
................................................................................................................... 115
Table 14: Diffusivity (m2/sec) of tool components in α-Ti and into β-Ti........................ 117
Table 15: The distance from cutting edge to crater’s center and traveling time with
carbide inserts ............................................................................................ 120

ix

Table 16: Predicted upper bound of diffusion wear at cutting speed of 91 m/min ....... 121
Table 17: Predicted diffusion wear of carbide tool in which carbon is control elements.
................................................................................................................... 122
Table 18: Publication on optimal spray conditions in MQL machining with definition of
Yaw and Pitch angle as shown in Figure 114 ............................................ 145
Table 19: Properties of graphite and hBN ................................................................... 149
Table 20: The diameters and thicknesses of various platelets .................................... 150
Table 21: Parameter for Tribometer tests.................................................................... 159
Table 22: Machining conditions for steel AISI 1045 .................................................... 160
Table 23: Ingot chemical analysis of Ti64 work material ............................................. 179
Table 24: Machining conditions for Ti64 ...................................................................... 180

x

LIST OF FIGURES
Figure 1: Phase diagram of the titanium alloys [E. Gautier and B. Appolaire, Ecole de
Mines de Nancy, France] ................................................................................ 3
Figure 2: Heat distribution of thermal load on tool and chip in turning [Konig, 1979] ....... 5
Figure 3: Phase diagram of titanium [Velsavjevic, 2012] ............................................... 11
Figure 4: Typical phase diagram of Ti alloys: a) α-stabilized system, b) β-stabilized
isomorphous system, c) β-stabilized eutectoid system [Frees, 2011] ........... 12
Figure 5: Alpha phase and its slip systems ................................................................... 14
Figure 6: Beta phase and its slip systems ..................................................................... 15
Figure 7: Critical resolved shear stresses (CRSS) as function of temperature for slip
systems in α phase [Lütjering, 2003] ............................................................ 16
Figure 8: a) Elasticity (E) and b) hardness as a function of the angle ߛ between the caxis of the unit cell and load direction [Lütjering, 2003; Britton, 2009] ........... 16
Figure 9: Elasticity (E) and shear (G) modulus of alpha phase as function of
temperature [Lütjering, 2003]........................................................................ 17
Figure 10: Schematically processing for bi-modal structure of two phase α + β Ti
alloys[Lütjering, 2003].................................................................................. 19
Figure 11: Various types of microstructure of Ti64 alloys [Attanasio, 2013; Maciej
Motyka, 2012; Meyer, 2008]........................................................................ 19
Figure 12: Ti64 alloy turning configuration and chip flow direction ................................ 22
Figure 13: Operating principal of a confocal microscopy ............................................... 24
Figure 14: Flow chart of data collection with confocal microscopy ................................ 25
xi

Figure 15: Measurements of tool wear at the rake face, nose and flank face ............... 27
Figure 16: Types of tool wear according to standard ISO 3685:1993 ........................... 28
Figure 17: The maximum depth and wear rate of crater wear on carbide inserts (YD101)
at low DOC ................................................................................................... 30
Figure 18: The geometric transformation on the nose of tool inserts in 3D view. ......... 32
Figure 19: Nose wear evolution of carbide inserts (YD101) at the cutting speed of
61m/min. ...................................................................................................... 33
Figure 20: Nose wear evolution of carbide inserts (YD101) at 91m/min. ...................... 34
Figure 21: Nose wear evolution of carbide inserts (YD101) at cutting speed of
122m/min. .................................................................................................... 35
Figure 22: Nose wear evolution of PCD1200 inserts at cutting speed of 61m/min. ....... 36
Figure 23: Nose wear evolution of PCD1200 inserts at cutting speed of 122m/min. ..... 37
Figure 24: The average nose wear land of the carbide inserts at the low DOC. ........... 38
Figure 25: The average nose wear land of the PCD inserts at the low DOC................. 38
Figure 26: Comparison of nose wear land on carbide and PCD inserts at low cutting
speed (61 m/min or 200sfm) ........................................................................ 39
Figure 27: Comparison of nose wear land on carbide and PCD inserts at medium
cutting speed (91 m/min).............................................................................. 39
Figure 28: Comparison of nose wear land on carbide and PCD inserts at high cutting
speed (122 m/min). ...................................................................................... 40
Figure 29: The scoring marks on the noses of YD101 inserts ....................................... 40

xii

Figure 30: The evolution of crater wear on carbide inserts (YD101) at 61m/min
(200sfm). ...................................................................................................... 43
Figure 31: The evolution of crater wear on carbide inserts (YD101) at 91m/min
(300sfm). ...................................................................................................... 43
Figure 32: The evolution of crater wear on carbide inserts (YD101) at 122m/min
(400sfm). ...................................................................................................... 44
Figure 33: The similarity of 2D crater wear profiles at different locations (88th and 108th)
on carbide inserts (YD101) at 61m/min (200sfm). ........................................ 44
Figure 34: The evolution of 2-D crater wear profiles at 108th locations on carbide inserts
(YD101) at 91m/min and 122m/min. ............................................................ 45
Figure 35: SEM image of smooth crater wear with YD101 inserts ................................ 45
Figure 36: The evolution of crater wear on PCD1510 inserts at 61m/min (200sfm). ..... 47
Figure 37: The evolution of crater wear on PCD1510 inserts at 91m/min (300sfm). ..... 47
Figure 38: The evolution of crater wear on PCD1510 inserts at 122m/min (400sfm). ... 48
Figure 39: SEM image of rough crater wear with PCD1510 inserts .............................. 48
Figure 40: The dissimilarity of 2-D crater wear profiles at different locations (88th and
108th) on PCD1510 inserts at 61m/min. ....................................................... 49
Figure 41: The evolution of crater wear on PCD1510 inserts at 91m/min and 122m/min.
..................................................................................................................... 49
Figure 42: The maximum depth and wear rate of crater wear on carbide inserts (YD101)
at high DOC ................................................................................................. 51
Figure 43: Comparison in wear rate of crater wear on carbide inserts (YD101) at low
DOC and high DOC ..................................................................................... 51

xiii

Figure 44: Nose wear on carbide YD101 inserts at the longest cutting time of three
cutting speeds. ............................................................................................. 53
Figure 45: Nose wear on the nose of PCD1510 inserts at the longest cutting time of
three cutting speeds. .................................................................................... 54
Figure 46: The flank wear on carbide YD101 inserts at the longest cutting time of three
cutting speeds. ............................................................................................. 54
Figure 47: The flank wear on PCD1510 inserts at the longest cutting time of three
cutting speeds. ............................................................................................. 55
Figure 48: Nose damage of the carbide inserts (YD101) .............................................. 55
Figure 49: The 2D profiles of nose and flank wear at 128th section of carbide YD101 and
PCD1510 inserts at 61m/min. ...................................................................... 56
Figure 50: The 2D profiles of nose and flank at 148th section of carbide YD101 and
PCD1510 inserts at 91m/min. ...................................................................... 56
Figure 51: The 2D profiles of nose and flank wear at 108th section of carbide YD101 and
PCD1510 inserts at 122m/min. .................................................................... 57
Figure 52: The evolution of flank wear land on the flank face of YD101 and PCD1510
inserts at various cutting speeds. ................................................................. 57
Figure 53: The comparison of flank wear land of carbide and PCD1510 inserts. .......... 58
Figure 54: The comparison of flank wear land at low and high DOC. ........................... 58
Figure 55: The SEM images of adhesion layer on the PCD1510 and carbide inserts in
the second set (high DOC). .......................................................................... 60
Figure 56: The elemental contents of adherent layers at white and dark area on
PCD1510 inserts .......................................................................................... 60

xiv

Figure 57: The elemental contents of adherent layers at white and dark area on carbide
YD101 inserts............................................................................................... 61
Figure 58: The SEM images of chip-tool interface of chips generated with carbide and
PCD1510 inserts. ......................................................................................... 62
Figure 59: Elemental content of chips with carbide and PCD1510 inserts for three
cutting speeds .............................................................................................. 62
Figure 60: The SEM images of adhered particles on the chips. .................................... 63
Figure 61: Elemental content of adherent particles on the chips. .................................. 63
Figure 62: The chip morphology (top view). .................................................................. 64
Figure 63: Five parameters represented for chip morphology (side view) ..................... 65
Figure 64: Chip morphology with carbide and PCD1510 inserts at all cutting speeds .. 65
Figure 65: Statistical data for height of peaks. .............................................................. 66
Figure 66: Statistical data for height of valleys. ............................................................. 67
Figure 67: The comparison on chip morphology between carbide and PCD1510 inserts
..................................................................................................................... 68
Figure 68: 2D orthogonal cutting FEM with heat transfer .............................................. 71
Figure 69: Temperature profiles on the chip along tool-chip interface with PCD ........... 74
Figure 70: The temperature in Ti- turning with PCD rake angle: (A) Without heat
transfer, (B) with heat transfer (friction µ=0.35) ............................................ 75
Figure 71: Effect of heat transfer to temperature profiles along the top surface of the
chip with PCD............................................................................................... 75

xv

Figure 72: Temperature profiles on the chip along tool-chip interface with PCD with and
without heat transfer..................................................................................... 76
Figure 73: Effect of JC parameter to temperature profiles on the chip along tool-chip
interface with PCD. ...................................................................................... 76
Figure 74: Temperature profiles on the chip along tool-chip interface with PCD at
various friction coefficients ........................................................................... 78
Figure 75: The chip contact length from experiment and simulation of PCD at various
friction values ............................................................................................... 78
Figure 76: Temperature profiles on the chip with PCD inserts and µ=0.6 at various
cutting speeds. ............................................................................................. 79
Figure 77: Samples for EBSD of material before (Bulk-Ti) and after (Chip) machining . 81
Figure 78: EBSD configuration ...................................................................................... 82
Figure 79: Microstructure of Ti64 alloy used in turning tests ......................................... 84
Figure 80: Grain size distribution of α-Ti in un-deformed work material (Bulk-Ti) .......... 85
Figure 81: The comparison of average diameter of grains in the bulk-Ti and in the
deformed material (chips). ........................................................................... 85
Figure 82: Burgers’ orientation relationship in β → α transformation............................. 87
Figure 83: Microstructures achieved at various intermediate temperatures by slowly
cooling from above the β transus ................................................................. 87
Figure 84: The comparison of misorientation in α-Ti crystals in un-deformed (bulk-Ti)
and deformed (chips) work material at low and high DOC ........................... 89
Figure 85: The comparison area fraction of β-Ti crystals in un-deformed (bulk-Ti) and
deformed (chips) materials. .......................................................................... 89

xvi

Figure 86: The distribution of the β-Ti (dark) and the α-Ti (colored) in the chips at high
DOC. ............................................................................................................ 90
Figure 87: Scoring marks on flank face of inserts in machining of Ti alloys .................. 91
Figure 88: Hardness of α-Ti as a function of the declination angles between c-axis to
vertical line [after Britton, 2009] and the ‘hard’ α-grains respect to flank face.
..................................................................................................................... 92
Figure 89: The distribution of α-crystal in hard orientation (red color) in Bulk-Ti sample
respected to flank face of tool along feed direction ...................................... 93
Figure 90: The size of ‘hard’ α cluster in Bulk-Ti. ......................................................... 94
Figure 91: Interaction of the ‘hard’ α-cluster and the inserts.......................................... 95
Figure 92: The scoring marks on the carbides and PCD inserts at low DOC ................ 95
Figure 93. Adhesion layer on the rake face of carbide and PCD inserts ....................... 98
Figure 94: Width (µm) of ten scoring marks on YD101 ................................................. 99
Figure 95: Range and distribution of width of scoring marks of nose at low DOC respect
to ‘hard’ α-cluster size ................................................................................ 101
Figure 96: Range and distribution of width of scoring marks on flank face at high DOC
respect to ‘hard’ α-cluster size ................................................................... 102
Figure 97: Classifying of scoring marks on YD101 (Left: confocal image, Right: SEM
image) ........................................................................................................ 103
Figure 98: Classifying of scoring marks on PCD1200 (Left: confocal image, Right: SEM
image ......................................................................................................... 103
Figure 99: The hot hardness of Ti64 and typical tool materials ................................... 109

xvii

Figure 100: Distribution of velocity components of the chip in turning ......................... 110
Figure 101: Flow chart for calculation of solubility of tool material .............................. 114
Figure 102: Temperature dependence of the hardness and solubility of HfC tool ....... 115
Figure 103: Dissolution of carbide tool into chip .......................................................... 116
Figure 104: Diffusion of elements into α-Ti and into β-Ti. ............................................ 117
Figure 105: The diffusivity of carbon in α-Ti, β-Ti, Ti64 and TiCx ................................ 118
Figure 106: Generalized thermochemical wear [after Olortegui-Yume, 2007] ............. 123
Figure 107: Tool constituents diffused into chip .......................................................... 125
Figure 108 Dissolution and diffusion wear rate of carbide tool (WC) into β-Ti at 1000°C
................................................................................................................... 130
Figure 109: Dissolution and diffusion wear rate of TiN tool into α-Ti at 800°C ............ 131
Figure 110: Dissolution and diffusion wear rate of TiN tool into β-Ti at 1000°C ......... 132
Figure 111: Dissolution and diffusion wear rate of cBN tool into α-Ti at 800°C ........... 133
Figure 112: Process (dissolution and diffusion) controls thermochemical wear for
machining of ferrous and Titanium with carbide tool. ................................. 134
Figure 113: Flow chart for calculation of relative thermochemical wear rate. .............. 134
Figure 114: Pitch and Yaw angle of the nozzle in End-ball milling .............................. 144
Figure 115: An illustration of hexagonal crystalline structures of ghraphite and hBN
[Encyclopedia Britannica, Inc., 1995] ......................................................... 148

xviii

Figure 116: Temperature range for lubrication of different solid lubricants [Chen N.,
2004] .......................................................................................................... 149
Figure 117: SEM analysis ........................................................................................... 152
Figure 118: SEM images micro- and nano-platelets ................................................... 152
Figure 119: Diameter, thickness and aspect ratio of graphite platelets. ...................... 153
Figure 120: High-speed mixer DAC 150FVZ-K ........................................................... 154
Figure 121: Wetting Measurement system [Park, 2011].............................................. 156
Figure 122: Veeco Dektak 6M Surface Profiler ........................................................... 157
Figure 123: Linear ball-on-disc type tribometer ........................................................... 158
Figure 124: Profile of a cross-section of wear track. ................................................... 159
Figure 125: Experimental Set up for MQL ball milling ................................................. 162
Figure 126: The stability of the mixtures with 0.1wt% micro- and nano platelets after 3
days ........................................................................................................... 163
Figure 127: Wetting angle of MQL lubricants on the surface of TiAlN coated carbide
inserts (left angle, right angle) .................................................................... 164
Figure 128: SEM surface images of tool surfaces ....................................................... 165
Figure 129: Roughness (Rz) and spacing parameters (Sm) ......................................... 165
Figure 130: Friction coefficients of various mixtures on tool A .................................... 167
Figure 131: Friction coefficients of mixtures with xGnP (M5) and hBN300 as function of
sliding speeds. ........................................................................................... 168

xix

Figure 132: The wear track appearance with 0.1wt% of nano-platelets at a speed of
2.5cms and load of 10N after 35000cycles (Left: xGnP M5, Right: hBN300)
................................................................................................................... 169
Figure 133: Depth and Width of wear track under various lubricant conditions on tool A
(Normal load: 10N, Speed: 2.5cm/s) .......................................................... 170
Figure 134: Depth and Width of wear track under various lubricant conditions on tool B
(Normal load: 10N, Speed: 2.5cm/s) .......................................................... 170
Figure 135: Geometric Relationships of micro and nano-platelets on the tool surfaces
................................................................................................................... 171
Figure 136: Minimum pitch angle of MQL nozzle for oil mist entering the cutting zone172
Figure 137: Flank wear at 15° pitch angle and various yaw angles with tool B (dash-line:
Tool chipping)............................................................................................. 173
Figure 138: Top View of MQL Experiment: The distribution of lubricant at 120° and -30°
yaw angle ................................................................................................... 173
Figure 139: Central wear with MQL nano-platelet enhanced mixtures after milling 3
layers ......................................................................................................... 174
Figure 140: Flank Wear after milling 6 layers. ............................................................. 175
Figure 141: Nose wear at 3500 rpm after ball milling six layers with tool A ................. 175
Figure 142: Flank Wear at 3500rpm after ball milling six layers with tool A (dash-line:
Tool chipping)............................................................................................. 176
Figure 143: Flank Wear at 3500 rpm after ball milling six layers with tool B................ 177
Figure 144: Types of tool wear were measured on end-ball nose inserts ................... 181
Figure 145: The flank wear measurement in a) VB2 and b) VB1 on the cutting edge
before and after etching the adhesion layer of titanium.............................. 181
xx

Figure 146: The flanks wear in VB1 on insert with Dry at 2500rpm. ............................ 183
Figure 147: The flank wear in VB1 on the insert with pure Unist oil at 2500rpm. ......... 183
Figure 148: The flank wear in VB1 on insert with xGnP C750 0.1wt% at 2500rpm...... 184
Figure 149: The flank wear in VB1 on insert with xGnP C750 1wt% at 2500rpm. ....... 184
Figure 150: Maximum flank wear with (VBmax) at 2500rpm (after etching) .................. 185
Figure 151: The average flank wear width (VBavg) at 2500rpm (after etching) ............ 185
Figure 152: The flank wear in VB1 after 1st layer with different lubrication conditions at
3500rpm. .................................................................................................... 186
Figure 153: The flank wear in VB1 after 2st layer with different lubrication conditions at
3500rpm. .................................................................................................... 186
Figure 154: Maximum flank wear width (VBmax) at 3500rpm (after etching) ................ 187
Figure 155: The average flank wear width (VBavg) at 3500rpm (after etching) ............ 187
Figure 156: Types of tool damage on ball nose end mill insert. .................................. 189
Figure 157: The first damage at the cutting edge with different lubrication conditions at
2500rpm. .................................................................................................... 189
Figure 158: The first damage at the cutting edge under various lubrication conditions at
3500rpm. .................................................................................................... 190
Figure 159: The largest damage at the cutting edge under various lubrication conditions
at 2500rpm. ................................................................................................ 190
Figure 160: The largest damage at the cutting edge with different lubrication conditions
at 3500rpm. ................................................................................................ 190

xxi

Figure 161: The effective diameter of the cut, De, for ball-nose end-mill insert........... 191
Figure 162: The nose wear at the first and the last cutting layer under different
lubrication conditions at 2500rpm............................................................... 191
Figure 163: The nose wear at the first and the last cutting layer under different
lubrication conditions at 3500rpm............................................................... 192
Figure 164: The crater wear and damage on the tool at the last cutting layer with
different lubrication conditions at 2500rpm. ................................................ 193
Figure 165: The crater wear and damage on the tool at the last cutting layer with
different lubrication conditions at 3500rpm. ................................................ 193

xxii

Chapter 1: Tool wear of carbide and PCD inserts in turning of Ti64
I.1

INTRODUCTION

I.1.1

Machining Titanium overview

Titanium (Ti) industry was established in 1950s mainly to make aerospace parts
such as engine components, rockets, and spacecraft. Nowadays, Ti and its alloys
have become the essential materials for aerospace and medical device industries
due to its outstanding physical properties. The outstanding properties also make Ti
alloys extremely difficult-to-machine materials. In particular, high speed cutting of Ti
alloys is difficulty because of the extremely short tool life. For example, at the cutting
speed of 122 m/min, polycrystalline diamond (PCD) cutting tools can last for 2-4
minutes and only about 1-2 minutes for uncoated carbide tools [Schrock, 2013]. The
current recommended cutting speeds for Ti alloys are less than 30m/min with high
speed steel (HSS) and 60m/min with carbide tools in dry machining [Rahman, 2003].
Overall the machinability of Ti alloys is considered to be poor in terms of many
variables such as cutting forces, tool life, metal removal rate, and surface finish
[Jaffery, 2008].
Ti alloys have the low density of around 4.5g/cm3, high hot strengths [Oosthuizen,
2010] and extremely low thermal conductivities, somewhere between 6.6 and
20W/m.K depending on the alloys. The phase transformation temperature (β-transus
temperature or Ts) and the melting temperatures for pure titanium are around 882°C
[Yang, 1999] and 1650°C [Oosthuizen, 2010], respectively. These temperatures vary
depending on the content of alloy ingredients and pressure. For example, the beta

1

transus temperatures for Ti-5Al-2.5Sn, Ti-6Al-4V and Ti-5Al-2Sn-2Zr-4Cr-4Mo are
1040°C, 995°C, and 885°C, respectively [Semiatin, 1996]. The transus temperatures
of pure titanium are 1055, 1013, and 873 °K for pressure of 0, 5 and 10 GPa,
respectively

[Velsavjevic,

2012].

Titanium

is

classified

in

two

categories,

commercially pure titanium (at least 99.67 wt% of Ti) and alloyed titanium (Ti alloys).
Commercially pure titanium has low strength but outstanding corrosion resistance,
which has very limited applications except in the chemical process industries. Ti
alloys have higher strength and, thus, wider applications in aerospace and medical
device industries. The allotropic nature makes the classification of Ti alloys into three
categories, Alpha, Beta, and Alpha-Beta.
Alpha titanium alloys (Ti5Al2.5Sn, Ti8Al1Mo1V, etc.), hexagonal close packed
microstructures, have the α-phase stabilizers such as Al, O, B, N, Sn that raise the βtransus temperature. Alpha and near-alpha alloys generally are not heat-treatable,
brittle and have low to medium tensile strengths, high corrosion resistance, and good
weldability. They are mainly used in corrosion resistance applications.
Beta

alloys

(Ti11.5Mo6Zr4.5Sn,

Ti5553,

etc.),

body

centered

cubic

microstructures, contain the β phase stabilizers such as V, Nb, Ta, and Mo that
reduce the β-transus temperature. Alloys in this group are readily heat-treatable,
ductile and exhibit high strengths and have slightly higher density [Machado, 1990].
The alloys in this group show higher hardenability.
Alpha-beta alloys (Ti6Al4V, Ti5Al4V, etc.) contain the combination of both α and β
stabilizers. The mechanical properties of these alloys vary substantially depending on
the heat treatment schedule, which can lead to high strength between room

2

temperature and moderately high temperatures. In general, the alloys in this group
have higher strength than the alpha and near alpha alloys, and are widely used in the
general purpose (strength applications) and aerospace industries as well. In the
alpha-beta alloys group, Ti-6Al-4V is the most commonly used alloy (accounting for
up to 60% of the titanium production [Boyer, 1996]).
The beta titanium alloys has much worse machinability than alpha and alpha-beta
(α+β) titanium alloys [Arrazola, 2009; Machai, 2013].
Beside these alloys, there are two transition types of alloys, near alpha and near
beta alloys. The near alpha alloys have up to 5% of beta stabilizers with the
microstructure containing alpha phase and about 3-5% beta phase. The near beta
alloys has the high fraction of metastable beta phase (βM) and the similar properties
of α+β alloys [Maciej Motyka, 2012].

Figure 1: Phase diagram of the titanium alloys [E. Gautier and B. Appolaire, Ecole de
Mines de Nancy, France]

3

In machining metal, approximately 90% of the heat generation come from plastic
deformation [Boothroyd, 1975; Shahan, 1993]. However, the main distinction of Ti
compared to other metallic alloys is its low thermal conductivities. They have the
thermal conductivity around 6.6 W/m·K [Calamaz, 2008] compared to ferrous
materials around 50.7W/m·K [Andriya, 2012; Nandy, 2008] and aluminum alloys
around 200W/m·K [Toh, 2004] at room temperature. The thermal conductivity of
titanium could reach 20W/mK at high temperature [Calamaz, 2008]. This makes the
dissipation of the heat generated during machining extremely slow. Konig et al.
[1979] stated that about 80% of heat generation is absorbed by the cutting tool in
machining Ti alloys while only 50% for machining ferrous materials as shown in
Figure 2. Therefore, most of the heat is concentrated on the cutting edge of the tool
when machining Ti alloys. The effect of the low conductivity of Ti alloys can be
evident by comparing the reported cutting temperatures; ~1000°C [Ezugwu, 1997;
Hartung, 1982] when turning Ti alloys, 600°C when turning ferrous materials [Dhar,
2002; 2007] and 200°C when turning aluminum alloys [List, 2005]. The relatively
small chip-tool contact area (typically three times less than that of steels [Ramesh,
2008]) leads to high stress and stress gradient on the tool. The high cutting
temperature could be slightly relieved by reducing the cutting speed and using a
large amount of coolant. High thermal stress and cutting force with the confined
contact area result in extremely high stresses near the cutting edge (within 0.5mm
[Ezugwu, 1997]) as well as excessive tool deformation and failure. In addition, the
high temperature increases the wear rate of dissolution, diffusion, and chemical
reaction of a tool material with Ti alloys, often beyond the transus temperature

4

(882°C for pure titanium [Yang, 1999]) where the drastic change in physical
properties occurs .

Figure 2: Heat distribution of thermal load on tool and chip in turning [Konig, 1979]
The surface integrity is another reason for reduction of tool life. It not only causes
work hardening beneath the machined surface but also increases surface roughness.
In particular, because of the high chemical affinity with a tool material, Ti alloys have
a tendency to gall and weld to the tool, thus causing the fracture and tool geometry
failure and resulting in premature tool failure. Applying coolant at high pressure helps
overcome these problems. However, the high chemical reactivity of Ti with lubricants
and additives creates other problems such as additional reaction possibilities
[Andriya, 2012; Ezugwu, 1997; Rahman, 2003].
The low elastic modulus [Andriya, 2012; Machado, 1990] and high work
hardenability of Ti alloys are additional obstacles in machining. The low elastic
modulus of Ti alloys causes a higher deflection (twice as much as steels [Ezugwu,
1997; Machado, 1990]), which leads to chatter during machining.

5

Although most metals generates continuous chips in typical machining conditions,
Ti alloys are notorious for segmented chips, experiencing plastic instability and
discrete bursts of catastrophic thermoplastic shear in the primary shear zone. The
frequency of the chip segmentation in α−β alloys is higher than those in α-alloys and
β-alloys [Joshi, 2014]. Two shear bands was observed on each segmentation in alloys [Motonishi, 1987]. The chip morphology and serrated frequency of the chip
also depends on cutting speed and microstructure [Gente, 2001; Joshi, 2014;
Molinari, 2002; Rahim, 2008]. The lower chip segmentation frequency (larger spacing
between segments) was found in a higher cutting speed [Bayoumi, 1995].
The challenges mentioned above are the principal problems associated with
machining Ti alloys. When machining Ti alloys, the tool life dramatically decreases as
the cutting speed increases. Thus, the tool materials should provide (1) high hot
hardness, (2) good thermal resistance, (3) good thermal conductivity to reduce
thermal gradient (4) good chemical inertness to minimize galling and welding, (5)
high toughness to withstand vibration force, and localized stress and chip
segmentation. The development of advanced coated tool materials has improved the
machining productivity significantly in many materials such as ferrous and aluminum
alloys but not for Ti alloys.
I.1.2

Motivation

In machining Ti alloys, beside PCD, the most successfully commercially available
tool material for cutting titanium is straight grade (uncoated) carbide tools (WC-Co or
WC). Even though many carbide tools with coatings are available in the market, the
coated carbide tools are not effective in machining Ti alloys. The reason behind this

6

is not clearly discovered at the present time. Therefore, the selection of tool materials
in machining Ti alloys is very difficult.
This thesis attempts to identify the wear mechanisms behind two main modes of
tool wear, crater and flank wear, by understanding metallurgical structure of a
selected Ti work material. This study can provide the fundamental knowledge in
designing effective tool materials for machining Ti alloys.
I.1.3

Tool materials and tool wear mechanisms reported in literature

Over the past six decades, the production of Ti alloys has increased to meet the
increased usage in aerospace applications. Many researchers have studied heavily
the subject of machining Ti alloys to understand how to design effective tool materials
and develop techniques to minimize machining cost. Numerous studies on the
machinability of Ti alloys have been carried out. In fact, many hypotheses on tool
wear mechanisms in machining Ti alloys have been introduced. Many researchers
believe that both mechanical wear (abrasion, attrition) and chemical wear
(dissolution, diffusion) are the main wear mechanisms in machining Ti alloys.
Tool materials for machining Ti alloys can be mainly classified in three groups:
•

High speed steel tools (HSS)

•

Coated and uncoated carbide tools

•

Super hard tool materials: PCD and CBN

Narutaki et al. [1983] conducted turning experiment on both alpha (Ti-5Al-2.5Sn)
and alpha-beta (Ti64) titanium alloys with straight carbide tools, CBN tools, cemented
TiN tools, sintered Al2O3 tools, sintered and natural diamond tools, and TiC coated
tools. He concluded that natural diamond tools offered excellent cutting performance

7

at the cutting speed of 1.67 m/s due to high thermal conductivity and low chemical
reactivity in both dry and water based coolant conditions. The cutting speed for these
diamond tools could go up to 3.33 m/s when applying sufficient coolant. Cemented
TiN, TiC-coated, and CBN tools were not recommended. In addition, CBN tools
caused very large and unusual groove marks on both the rake and flank faces. The
sinter diamond tools exhibited similar performances as the carbide tools at high
cutting speeds. Rahman et al. [2003] found that the binderless CBN tools without any
cobalt binder yielded significant improvement in tool life compared to regular CBN
tools, and even comparative to PCD tools at the cutting speed of 400m/min with high
pressure coolant.
With the turning experiment of Ti64 with multilayer coated (TiN-Al2O3-TiCN-TiN)
carbide tools, Ibrahim et al. [2009] stated that adhesive wear and welding were
predominant wear mechanisms both in flank and the rake face. The adhesion wear
took place after the coating had gone (worn out or flaked off). Corduan et al. [2003]
conducted the machining experiment of Ti64 with PCD, CBN and TiB2 coated
carbide. He recommended TiB2 coated carbide tools only for cutting speeds of less
than 100 m/min. While PCD tools showed the lowest wear rate, CBN tools were only
recommended for finishing cutting. He mentioned the delamination of TiB2 coating is
the main reason why the coated carbide tools did not work. Due to the mismatch in
thermal expansions between coating and substrate, internal stress is generated as
the temperature rises. The internal stress can intensify enough to break and flake off
the coating.

8

I.1.3.1 Dominant tool wear mechanisms
In milling of Ti6246, Jawaid et al. [1999] claimed that the main wear mechanisms
were dissolution/diffusion and attrition wear with PVD-TiN and CVD-TiCN+Al2O3
coated carbide tools which caused the carbide grain to pull out. Nabhani et al. [2001]
conducted turning experiments of alpha-beta Ti48 with TiC/TiC-N/TiN coated WC,
CBN (Cubic Boron Nitride) and PCD tools with “quick-stop” device to capture the insitu tool condition during cutting process. By investigating tool wear appearance and
coherent metallic layer, they concluded that diffusion/dissolution and attrition were
the dominant tool wear mechanisms. The coating was not beneficial in resisting tool
wear since these layers were rapidly worn off, leading to immediate exposure of the
carbide substrate. The thickness of the protective adherent metal layer was strongly
influenced by the balance between the diffusion rate of tool material through layer
and the dissolution rate of the layer into work material.
I.1.3.2 TiC protection layer
Hartung and Kramer [1982] carried out a comprehensive study on turning Ti64
with various tool materials (WC, TiC, CBN, Al2O3, TiCN, PCD, etc.) and various
coatings (HfO2, HfC, TiC, HfN, TiN, etc.) on carbide inserts. They reported that Al2O3
had the highest tool wear and PCD was the best in terms of wear resistance.
Uncoated carbide tools showed better performance than coated ones. A higher wear
rate was recorded with all coated carbide tools. Furthermore, they claimed that the
least soluble tool component controls the solubility of tool material. For example, the
solubility of WC was not greater than that of C (0.6 at%) and less than that of W (100
at%). Thus, they concluded that dissolution and diffusion wear models of tool

9

constituents in titanium were not sufficient to describe the tool wear in machining of Ti
alloys. They believed that, because of the high reactivity of Ti, the reaction layer,
titanium carbide (TiC), is formed, which becomes the main factor to control tool wear.
In the comparative research of turning Ti64 and Ti555.3 by Arrazola et al. [2009], the
results supported the conclusion by Hartung and Kramer [1982]. He showed that the
presence of TiC layer was less stable at the higher cutting speed (90m/min) which
accelerated tool wear.
I.1.3.3 Cobalt diffusion
A cobalt-based diffusion tool wear model was introduced by Hua et al. [2005].
They carried out the turning experiment with Ti64 with uncoated carbide tools with
two distinct cobalt contents (6wt% and 10wt%). In the model, the wear rate was
calculated as the ratio of flux rate of diffused cobalt at tool-chip interface over density
of cobalt. He found that the results of the model agreed well with the experimental
data. Furthermore, the simulation indicated that the temperature increases with the
increase in cutting speeds while the chip contact length decreases, which leads to
rapid tool wear. The maximum depth of crater profile coincided with the peak
temperature which moved closer to the tool nose as the cutting speed increased.
I.1.4

Phases and microstructure of Ti alloys

I.1.4.1 Phases in Titanium alloys
It is known that the cementite phase (Fe3C) is present as the main abrasive
contributing to the flank wear in machining ferrous materials. Titanium can exist in
alpha, beta and rarely omega phases. Figure 3 shows the phase diagram as a

10

function of temperature and pressure.

However, the phase diagram is strongly

influenced by the alloying elements and their content as shown in Figure 4. The
mechanical properties and hardness of Ti alloys influence the size, composition, and
volumetric fraction of α, β and ω-precipitated phases. No significant hard phase
exists in Ti alloys, which make hard to pinpoint the root cause of tool wear. In
addition, the microstructural features, which are affected by heat treatment and
alloying elements, also plays very important role in tool wear in machining Ti alloys.

Figure 3: Phase diagram of titanium [Velsavjevic, 2012]

11

Figure 4: Typical phase diagram of Ti alloys: a) α-stabilized system, b) β-stabilized
isomorphous system, c) β-stabilized eutectoid system [Frees, 2011]
Among the three phases of Ti, the α-phase with HCP structure is stable at room
temperature and pressure without any alloy stabilizer. At room temperature, the
hexagonal unit cell of the α-phase has the lattice parameters a (0.295 nm) and c
(0.468 nm) as shown in Figure 5. The c/a ratio for pure α-phase (1.587) is smaller
than the ideal ratio for the archetypal hexagonal crystal structure (1.633). Crystalline
structures accommodate plastic deformation along certain planes and directions
within the crystal lattice. As a general rule, the slip occurs in the densest packing
crystal planes (number of atoms/area) along the directions of the highest linear
density (atom/length). The α-phase has four slip planes, ({0001}, {1100}, {1101}, {1
1 22}, {1101}) and two slip directions (<1120>, <11 2 3>) making to 24 possible slip
systems as shown in Table 1. To determine which slip system is likely more active,

12

the critically resolved shear stress (CRSS) and the geometric relation between the
slip plane and the applied stress are used. The CRSS shown in Figure 7 is the inplane stress component required for dislocation movement as a function of
temperature. The predominant slip mode in the α phase is in {1100}, {0001}, {1101}
along <1120> direction. The highest CRSS is required for slip <11 2 3> direction.
Because of the intrinsically anisotropic character, the α phase has pronounced
variation of the elastic modulus (Ε) and hardness (H) as a function of the angle γ
between the c-axis of the unit cell and load direction. The elastic modulus and
hardness reach the highest values along the c-axis but the lowest in perpendicular
direction to the c-axis as presented in Figure 8. Furthermore, the elastic and shear
modulus and hardness decrease with the temperature as shown in Figure 9. The αphase has two variational phases, martensite structure (α′) and orthorhombic
martensite (α″).
The beta (β) phase is a metastable phase with body centered cubic (BCC)
structure. It has more slip systems than the α-phase making it more ductile and
easier to deform. The elastic modulus, shear modulus and hardness of the β phase
are well below that of the α phase [Meier, 1992].
The omega (ω) phase only presents in metastable β-alloys. The isothermal ω
particles have either an ellipsoidal or a cuboidal shape. It is well accepted that,
the ω phase has the highest elastic modulus and hardness followed by the α and α′,
phases [Zhou, 2004, 2008]. Many studies [Gabriel, 2013; Hickman B.S., 1969; Hsu,
2013; Jon, 1972; Jones, 2009] revealed that the ω (ellipsoidal) precipitates forms at
the beginning of the aging treatments and quenching β alloys from the beta field.

13

Although the ω-phase cannot be identified with optical analysis or the X-ray
diffraction (XRD) pattern, it can be investigated by the selected area diffraction
(SAED) patterns and bright-field image with transmission electron microscopy (TEM)
analysis.

Figure 5: Alpha phase and its slip systems

14

Figure 6: Beta phase and its slip systems
Table 1: Slip systems in alpha and beta phase

β phase

α phase

Slip plane family
No. of variants

Slip direction
No. of variants

Total no of slip systems
Designation

{0001}: 1

<a> or <1120>: 3

3 “Basal”

{1100}: 3

<a> or <1120>: 1

3 “Prismatic”

{1 101}: 6

<a+c> or <1123>: 1

6 “1 order Pyramidal”

{1 101}: 6

<a> or <1120>: 1

6 “1 order Pyramidal”

{1 1 22}: 6

<a+c> or <1123>: 1

{110}: 12

<111>: 1

12

{112} : 12

<111>: 1

12

{321}: 24

<111>: 1

24

15

st

st

nd

6 “2

order Pyramidal”

Figure 7: Critical resolved shear stresses (CRSS) as function of temperature for slip
systems in α phase [Lütjering, 2003]

Figure 8: a) Elasticity (E) and b) hardness as a function of the angle ߛ between the caxis of the unit cell and load direction [Lütjering, 2003; Britton, 2009]

16

Figure 9: Elasticity (E) and shear (G) modulus of alpha phase as function of
temperature [Lütjering, 2003]
I.1.4.2 Phase transformation in Titanium alloys.
The phase transformation happens with the transition metals in the group IVB
such as Ti, Zr, Hf under applied pressure (p), temperature (T) or both (p, T). The α ⇔
β is more favorable with temperature while pressure is more critical for β ⇔ ω
transformation.
The α → β transformation happens when the temperature reaches transus
temperature (100% β phase). The transus temperature is strongly influenced by
interstitial and substitutional alloy ingredients.
The β phase transforms to α phase during continuous cooling of the α-β Titanium
alloys from above transus temperature. Depending on the cooling rate, the phase
transformation mechanism could be β → α for slow cooling rate and diffusionless
transformation β → α’ for fast cooling rate at temperatures below 700-750 °C.

17

Quenching Ti64 alloy from the 750–900 °C temperature range produces an
orthorhombic martensite (α″).
The β ⇔ ω transformation is reversible and diffussionless. The phase
decomposition mechanism of the metastable beta phase in this material follows the
classical behavior for this type of alloy: route 1 (β → β + ω + α → β + α) and route 2
(β → β + ω → β + ω +α → β + α) [Grabriel, 2013].
I.1.4.3 Microstructure of α + β Ti alloys
The properties, hardness and machinability of each Ti alloy are directly related to
its microstructure. The dual phase (α + β) Ti alloys have different types of
microstructure depending on the heat treatment (solution treated and aged
temperatures; cooling rate and cooling rate. Four common microstructures are milled
annealed, equiaxed structure, fully lamellar structure and bi-modal (or duplex)
structure which is a combination of equiaxed and lamellar structure as shown in
Figure 11. Equiaxed structure has good creep and crack growth resistance, but
suffers from low tensile ductility and moderate fatigue properties. The bi-modal
structure is characterized by high ductility and fatigue strength, high yield and tensile
stress. High resistances to crack propagation and fracture toughness are the notable
properties of a fully lamellar structure [Zhang, 2014]. The bi-modal microstructure is
typical in the metallurgical processing of α + β alloys when heat-treated in the α + β
field as shown in Figure 10 [Lütjering, 2003]. The fully lamellar structure can be
achieved by heat treatment typically above beta-transus temperature (beta annealed)
then slowly cooled in furnace or in the air. In this microstructure, the thickness of
Widmanstätten α-laths, colony size, and prior grain size are important parameters

18

that can be controlled by the cooling rate. The fully lamellar structure is believed to be
harder and more difficulty to machine than other microstructures.

Heat treatment in α + β

Figure 10: Schematically processing for bi-modal structure of two phase α + β Ti
alloys[Lütjering, 2003].

Mill annealed
389±18HV

955°°C-ST/1hr + FC
Bimodal
411±15HV

925°°C-ST/4hr + FC
Fully Equiaxed
397±24HV

1035°°C-ST/0.5hr + FC
Fully lamellar
365±45HV

1065°°C-ST + QA
Martensitic

Necklace

Figure 11: Various types of microstructure of Ti64 alloys [Attanasio, 2013; Maciej
Motyka, 2012; Meyer, 2008]

19

I.2

EXPERIMENTAL SETUP AND PROCEDURES

I.2.1

Turning experiments of Ti64 alloy

The turning experiments were conducted with Ti-6Al-4V with the average
hardness of 326HV on Yama Seiki GA-30 lathe using tungsten carbide (WC6wt%Co) and PCD inserts. The tests were in dry condition to understand the wear
mechanisms between Ti alloy and cutting tool. The tool insert geometry was ANSI
designation CNMA-432.
To investigate wear evolutions in both the rake and flank faces, all the inserts were
flat rake face without chip breaker. Two grades (YD101, YD201) of Zhuzhou
Cemented Carbide Cutting Tools, Co, LTD, (ZCCCT, Zhuzhou, Hunan, China) brand
uncoated carbide inserts were chosen for the test for their flat rake face. Grade
YD201 was a straight grade carbide containing approximately 94% WC and 6% Co.
The average grain size was 2µm. Carbide grade YD101 had a composition
consisting of 93.6% WC, 0.15%NbC, 0.25%TaC, and 6%Co. The average grain size
was 1µm.
Three PCD grades and two tungsten carbide grades were used as tool inserts in
this study. The PCD tips were Compax® 1200 and 1500 grades, 92% diamond by
volume, with average diamond grain size of 1.5 µm and 25 µm, respectively,
manufactured by Diamond Innovation. Shape-Master Tool Company, Kirkland, IL,
brazed each PCD tip onto an ISO CNMA120408 carbide base. The 1200 grade PCD
inserts with 0° rake angle had an average grain size of 1.5µm. Two grades, 1210 and
1510, of PCD inserts with a 10° positive rake angle. The feed rate remained constant
for all samples at 0.127 mm/rev (0.005 in/rev). Two sets of experiments with the

20

depth of cut (DOC) of 0.635 mm (low DOC) and 1.2 mm (high DOC), respectively,
were conducted to see the effects of DOC to cutting process and tool wear. The
information for tool grades and DOC was summarized in Table 2. The low DOC
showed wear at the nose while the high DOC showed traditional flank wear. Lead
angle refers to the angle between the imaginary line perpendicular to the direction of
feed and the line parallel to the cutting edge of the insert. Rake angle in this context
refers to the inclination angle in the tool holder that gives the insert its clearance with
respect to the work piece. For consistency, both of the inserts were run with a -5°
°
lead angle and a -5°
° rake angle with respect to the work material. The turning setup
and the chip flow direction respect to tool were presented in Figure 12. Tables 1 and
2 below lists the cutting time used in the turning tests for carbide and PCD inserts at
each cutting speeds.
Table 2: Tool grades and DOC used
Tool

Grade
(Tool ID)

Rake
angle

Grain
size (µm)

Thermal
conductivity
(W/m·K)

DOC
(mm)

Carbide

YD101

0°

1

65

0.635 & 1.2

Carbide

YD201

0°

2

75

0.635

PCD

PCD1200

0°

1.5

450

0.635

PCD

PCD1210

+10°

1.5

450

1.2

PCD

PCD1510

+10°

25

450

1.2

21

Tool

Side view

Front view

CNMA-432

8

S(rpm)

0
-5° Rake
angle

Flat

d

c

Positive
rake angle

Depth of
cut line

feed
direction

32°

5° Lead angle

53°

chip
flow
direction

5°

PCD1200 _61m/min_2.5min
Figure 12: Ti64 alloy turning configuration and chip flow direction
Table 3: Cutting time for the second set of experiment at high DOC (dc= 1.2 mm)
Cutting speed
Inserts

61m/min

91m/min

122m/min

(200sfm)

(300sfm)

(400sfm)

2.4 min

1.6 min

0.6 min

5.2 min

3.5 min

1.1 min

6.8 min

4.5 min

1.8 min
1.9 min

YD101

2.2min
2.6 min (chipped)
2.7 min (chipped)
3.5 min (chipped)
PCD1510

2.5 min

1.6 min

1.2 min

5.4 min

3.5 min

2.7 min

6.8 min

5.4 min

3.5 min

22

Table 4: Cutting time for the first set of experiment at low DOC (dc= 0.635 mm)
Cutting speed
Inserts

YD101

61m/min

91m/min

122m/min

(200sfm)

(300sfm)

(400sfm)

3 min (chipped)

1 min

6 min (chipped)

2 min

9 min

3 min (chipped)

12 min

4 min

3 min (chipped)
YD201

6 min

All Inserts

9 min (chipped)

Chipped

12 min
6 min
12 min
PCD1200

18 min
24 min

I.2.2

1 min (chipped)
2 min (chipped)
30 sec
1 min (chipped)
2 min
1 min (chipped)

No Inserts

2 min

Tested

3 min
4 min

30 min (chipped)
PCD1210

30 sec

6 min

4 min

2 min

12 min

6 min

3 min

24 min

8 min

4 min

Confocal Microscopy

In this work, Ziess LSM 210 Confocal Laser Scanning Microscope (CSLM) was
used to capture both 2-D and 3-D images of crater and flank wear. This particular
confocal system can work in confocal, non-confocal and conventional optical modes.

23

Laser beam

Collecting
Lens

Beam splitter

Photo
detector

Computer

PMT

Objective Lens

Focal point

Focal distance
Depth of field

Focal plane

Z-movement

Specimen
Movement stage

2D and HEI
images

Monitor

X-movement
(manually)

Figure 13: Operating principal of a confocal microscopy
Each 2-D picture was the overlapping image of 50 regular microscope pictures
taken through the wear depth using Z-series function of CSLM. That revealed more
information of the wear region at high to low points.
The CLSM is used to obtain the surface information from a very thin ‘‘optical slice’’
at a focal plane by eliminating the reflected light from above and below that plane.
For this research, the number of optical slices (50-100), step size (500-3000nm) and
objective (100X, 200X) were used to construct the comprehensive 3D wear surfaces.
By manipulating a stack of optical slices, a grayscale encoded z-matrix [z(x,y)] that
contained the height information of the worn surface was generated as an output
image of the surface. The particular image is called a height encoded image (HEI).
The 3D wear topographies as a three-dimensional mesh of (X, Y, Z) data of both the

24

rake and flank faces of the cutting tools were created from the HEI images. With the
similar process, the 2D images are generated in a much higher quality, which have
the detailed information to analyze the wear patterns on the inserts.
Objective, step size,
number of optical slides,
current section

Yes

Current section is
greater than number
of optical slides?

No
Collect points at
current section (optical
slide)

Move to next optical
slide (current section
+step size)

Finish
collection of
data

2D & HEI
images

Figure 14: Flow chart of data collection with confocal microscopy
To reduce the noise in the capturing process, a wavelet filtering procedure
described by Olortegui-Yume and Kwon [Olortegui-Yume, 2010; Park, 2011] was
used to process the HEI images in Matlab. The main advantage of the wavelet
transform for image processing is to extract the surface topography clearly from the
raw image data without losing any surface details. In this work, the two-dimensional
discrete wavelet transform (2D-DWT) was used in a multi-resolution scheme with a
25

two-channel filter bank, which consisted of a pair of filters, low-pass and high-pass,
based on the chosen mother wavelets, Daubechies7 (db7) [Rioul, 1991].
Finally, the combination of multiple 2D wear profiles extracted from the 3D
topographies for each cutting time provided the quantitative data on wear surfaces,
which can be accumulated to construct the wear evolution as a function of time. For
the crater wear on the rake face, the 2D crater wear profiles were taken in the
direction of chip flow to insure maximum wear profiles. For nose and flank wear, 2D
wear profiles were taken along the y direction at various x-positions (1st to 256th
section) as shown in Figure 15. The PCD inserts with +10° rake angle inserts were
placed on the 10° incline stage to measure them based on the horizontally flat plane.
The tool wear measurement was conducted both before and after an adhered
layer (most likely Ti and TiC) was removed with a weak hydrofluoric acid solution (HF
at 10 vol%).

26

Wavelet
Transform
(Matlab)

HEI

50

2D	wear	
profiles	
th

3D	wear	
topographies	

128 section

The evolution of Nose wear of YD101 tool at 200sfm (-num80-num16-num17-num18-128sec)
10
5

100

0
-5
Height(µm)

y (pixels)

Crater Wear

HEI	image

150
200

-10
-15

Rake face

-20
YD101_0min0sec_0m

250

-25

YD101_2min24sec_146m
YD101_5min12sec_316m
YD101_6min48sec_414m

50

100 150 200 250
x (pixels)

-30
100

150

200

250
300
Distance(µm)

350

400

450

th

128 section

Wavelet
Transform
(Matlab)

5
0
-5
Height(µm)

Nose Wear

The evolution of Nose wear of YD101 tool at 200sfm (-num80-num16-num17-num18-128sec)
10

-10
-15

Geometric
transformation

-20
YD101_0min0sec_0m
YD101_2min24sec_146m
YD101_5min12sec_316m

-25

YD101_6min48sec_414m
-30
100

150

200

Nose
st

th
1
128
section section

350

400

450

th

256
section

Wavelet
Transform
(Matlab)

Flank Wear

250
300
Distance(µm)

th

128 section
The evolution of Nose wear of YD101 tool at 200sfm (-num80-num16-num17-num18-128sec)
10
5
0

Height(µm)

-5
-10
-15
-20

15

YD101_0min0sec_0m

Flank face

-25

YD101_2min24sec_146m
YD101_5min12sec_316m
YD101_6min48sec_414m

st

1 section

th

-30
100

th

128 section 256 section

150

200

250
300
Distance(µm)

350

400

450

Figure 15: Measurements of tool wear at the rake face, nose and flank face
I.2.3

SEM picture and element mapping

The optical view and the element mapping of worn tools were captured and
characterized before and after cleaning the adhesion layer using JEOL 6610LV
Scanning Electron Microscope with Energy Dispersive X-ray Spectroscopy. The
accelerating voltage and working distance were 20kV and 15mm, respectively.

27

I.3

EXPERIMENTAL RESULTS AND DISCUSSION

I.3.1

Wear Characteristics of tools inserts at low DOC

The machining experiments were divided into two sets based on the DOC. The
crater wear characteristic of the first set with the low DOC was reported in detail
elsewhere [Schrock, 2012; 2014]. In this thesis, the flank wear of the first set was
studied and presented. Furthermore, the second set of the machining experiments
with high DOC was conducted to study the wear characteristics of both flank ad
crater wear. Figure 16 presents the definitions of crater and flank wear studied in this
research.

Figure 16: Types of tool wear according to standard ISO 3685:1993
I.3.1.1 Crater wear at low DOC
The maximum crater depths of carbide and PCD inserts from the first set of
experiments are summarized in Table 5. The depths were calculated from the 2Dprofiles of craters reported in [Schrock, 2012; 2014] after excluding the fractured and

28

chipped inserts. These data were used to compare those of high DOC presented in
the next chapter.
Table 5: The maximum crater wear depth (µm) of the first turning tests with dc=0.635
[Schrock, 2012; 2014]
Tool

YD101

YD201

PCD1200

PCD1210

Cutting time
61 m/min (200sfm)
3min

--

--

--

---

6min

17

24

--

--

9min

26

33

--

--

12min

45

42

--

--

91 m/min (300sfm)
1min

11

--

--

--

2min

35

--

--

--

3min

50

--

--

--

4min

64

--

--

--

5min

--

--

--

--

7min

--

--

--

--

122 m/min (400sfm)
1min

15

14

--

--

2min

62

--

--

--

3min

--

--

--

--

29

70

35

60

Depth (µm)

50

YD101
dc=0.635mm

40

30

20

400sfm
300sfm
200sfm

YD101
dc=0.635mm

30

Wear rate (µm/min)

400sfm
300sfm
200sfm

25

20

15

10

10

5

0

0
0

100

200

300

400

500

600

700

800

0

100

200

Cutting Length (m)

300

400

500

600

700

800

Cutting Length (m)

Figure 17: The maximum depth and wear rate of crater wear on carbide inserts
(YD101) at low DOC
I.3.1.2 Nose wear at low DOC
Nose wear and flank wear land were also analyzed at various cutting conditions
(cutting speed and time). However, the first set generates notably the nose wear,
which is slight different from traditional flank wear. It should be noted that the depths
of nose wear along the tool curvature were much smaller than the tool radius.
Therefore, the evolution of nose wear was not clearly represented in 3D topographies
in its natural view. For a better representation, the 3D nose wear topographies are
presented after the curved surface has been flattened by the geometric
transformation as shown in Figure 18. In this transformation, the highest point for
each profile along X-axis direction in the unworn area was determined based on the
tool geometry. For each point along the X-profile, an addition in height (h) was
calculated based on the tool radius (r) and the difference in x direction between this
point and the highest point as shown in Figure 18. This transformation only affected
the 3D view of the nose wear. The 2D wear profiles nose wear were plotted from
initial data without any effect of geometric transformation.
30

The nose wear on carbide YD101 and PCD1200 inserts at various cutting speeds
as a function of time were presented in Figures 19-23. Each figure consists of three
sets of images from CLSM, (2D images, 3D images and 2D wear profiles at 128th
location in X direction of the HEI images which consist 256 X-locations (or the
crosstion at middle of measured wear region), see Figure 15). Each set contains
from the fresh insert to the worn inserts after cutting for the longest time at each
machining condition. The 3D images from CLSM show that the flank wear are
characterised by two major damages, the scoring marks and the chipping in both
macro- and micro-scales. For low and medium cutting speeds, the carbide inserts
show smoother appearance on the wear land than PCD inserts. At the high cutting
speed, as expected, both carbide and PCD inserts experienced more extensive
chipping on the flank face. Figures 24 - 28 show the average nose wear land (VBavg)
for both grades of carbide and PCD inserts. Despite of some missing data due to the
macro chipping on most PCD1200 at 91m/min, the obtained data showed the general
trend in flank wear. It should be pointed out that many smooth and wide scoring
marks (width > 40µm) were obtained on the flank wear of the carbide inserts at
medium and high cutting speeds. The scoring marks on the PCD inserts were much
more narrow and much less in number.

31

Y
Profiles along X-axis

YD101_91m/min_3.5min

݄ ൌ ‫ ݎ‬െ ඥ‫ ݎ‬ଶ െ ‫ ݔ‬ଶ
Z
highest point
h

transformed
point x

r : tool radius

Y

unworn
area

X

Profiles along X-axis

X

Regular 3D view (90° rotation)

Geometric
transformation

Geometric transformation 3D view

Figure 18: The geometric transformation on the nose of tool inserts in 3D view.

32

YD101 _61m/min_0min
(fresh tool)

YD101 _61m/min_~6min

st

Cutting edge

th
1
128
2D-profile

YD101 _61m/min_9min

256

th

YD101 _61m/min_12min

Scoring marks
100µ
µm

100µ
µm

a) 2D image

b) 3D image

10

Nose face

Cutting edge

0

Depth(µm)

-10

Nose wear at 61m/min
128thlocation

-20

Depth of
scoring mark

-30

Rake face
-40

-50

-60
-300

YD101_0min0sec_0m
YD101_5min57sec_363m
YD101_9min0sec_549m
YD101_12min0sec_732m
-250

-200

-150

-100

-50

0

Distance to cutting edge(µm)

c) 2D profile of wear at 128th location
Figure 19: Nose wear evolution of carbide inserts (YD101) at the cutting speed of
61m/min.

33

YD101	_91m/min_0min	
(fresh	tool)	
Cutting	edge	
	

YD101	_91m/min_2min	

128th		
2D-profile		

Dust	
contamination	

100µ
µm
100µ
µm µ
100µ
m
YD101	_91m/min_7min	
YD101	_91m/min_4min	
Scoring	marks	

100µ
µm

100µ
µm

a) 2D image

b) 3D image

10
Nose	wear	land	

0

D epth(µ m )

-10

Nose wear at 91m/min
128thlocation
Chipping	 at	 the	
cutting	 edge	 &	
the	rake	face
	

-20

-30

-40

-50

-60
-300

YD101_0min0sec_0m
YD101_2min0sec_122m
YD101_4min10sec_254m
YD101_7min0sec_427m
-250

-200

-150

-100

-50

0

Distance to cutting edge(µm)
th

c) 2D profile of wear at 128 location
Figure 20: Nose wear evolution of carbide inserts (YD101) at 91m/min.

34

YD101	_122m/min_0min	
(fresh	tool)	
	

YD101	_122m/min_2min	

100µ
µm
YD101	_122m/min_3min	

100µ
µm
YD101	_122m/min_~4min	

100µ
µm

100µ
µm

b) 3D image

a) 2D image
10

0

Depth(µm)

-10

Nose wear at 122m/min
128thlocation

-20

Chipping	at	the	
cutting	edge	&	
the	rake	face

-30

-40

-50

-60
-300

YD101_0min0sec_0m
YD101_2min0sec_122m
YD101_3min0sec_183m
YD101_3min57sec_241m
-250

-200

-150

-100

-50

0

Distance to cutting edge(µm)
th

c) 2D profile of wear at 128 location
Figure 21: Nose wear evolution of carbide inserts (YD101) at cutting speed of
122m/min.

35

PCD1200	_61m/min_0min	
(fresh	tool)	
	

PCD1200	_61m/min_6min	

Scoring	marks
100µ
µm
PCD1200	_61m/min_12min	

100µ
µm
PCD1200	_61m/min_18min	

100µ
µm

100µ
µm

a) 2D image

b) 3D image

10

0

Depth(µm)

-10

Nose wear at 61m/min
128thlocation

-20

-30

-40

-50

-60
-300

PCD1200_0min0sec_0m
PCD1200_6min0sec_363m
PCD1200_12min0sec_732m
PCD1200_18min0sec_1097m
-250

-200

-150

-100

-50

0

Distance to cutting edge(µm)
th

c) 2D profile of wear at 128 location
Figure 22: Nose wear evolution of PCD1200 inserts at cutting speed of 61m/min.

36

PCD1200	_122m/min_0min	
(fresh	tool)	

PCD1200	_122m/min_2.5min	

100µ
µm
PCD1200	_122m/min_3min	

100µ
µm
PCD1200	_122m/min_4min	

100µ
µm

Scoring	marks

100µ
µm

a) 2D image

b) 3D image

10

0

Depth(µm)

-10

Nose wear at 122m/min
108thlocation

-20

-30

-40

-50

-60
-300

PCD1200_0min0sec_0m
PCD1200_2min30sec_305m
PCD1200_3min0sec_366m
PCD1200_4min0sec_488m
-250

-200

-150

-100

-50

0

Distance to cutting edge(µm)
th

c) Wear profile at 128 location
Figure 23: Nose wear evolution of PCD1200 inserts at cutting speed of 122m/min.

37

180

180

N o s e w e a r la n d (µ m )

140

N o se w e a r la n d (µ m )

YD101
dc=0.635mm

160

122m/min
91m/min
61m/min

120
100
80
60

YD201
dc=0.635mm

160

122m/min
61m/min

140

120

100

40
20

80
0

100

200

300

400

500

600

0

100

200

300

400

500

600

700

800

Cutting Length (m)
Cutting Length (m)
Figure 24: The average nose wear land of the carbide inserts at the low DOC.

70

PCD1200
dc=0.635mm

120

122m/min
61m/min

N o se w e a r la n d (µ m )

N ose w ea r land (µ m )

140

100

80

60

40

20

60

PCD1210
dc=0.635mm

122m/min
91m/min

50

40

30

20

10
0

200

400

600

800

1000

1200

1400

1600

200

300

400

500

600

700

Cutting Length (m)
Cutting Length (m)
Figure 25: The average nose wear land of the PCD inserts at the low DOC.

38

800

200

dc=0.635mm
V=61m/min

Nose wear land (µm)

180
160

YD101
YD201
PCD 1200

140
120
100
80
60
40
20
0

200

400

600

800

1000

1200

1400

1600

Cutting Length (m)
Figure 26: Comparison of nose wear land on carbide and PCD inserts at low cutting
speed (61 m/min or 200sfm)

Nose wear land (µm)

200

dc=0.635mm
V=91m/min

YD101
PCD1210

150

100

50

0

200

400

600

800

Cutting Length (m)
Figure 27: Comparison of nose wear land on carbide and PCD inserts at medium
cutting speed (91 m/min)

39

200

Nose wear land (µm)

PCD1200
PCD1210
YD101
YD201

dc=0.635mm
V=12m/min

180
160
140
120
100
80
60
40
20
0

100

200

300

400

500

600

Cutting Length (m)
Figure 28: Comparison of nose wear land on carbide and PCD inserts at high cutting
speed (122 m/min).
It should be pointed out that YD101 inserts at the medium cutting speed had larger
scoring marks (>60µm) on the nose in comparison to low and high cutting speeds. In
addition, all YD201 inserts at 91 m/min were fractured. The possible reason could be
chatter and vibration at this cutting condition. There were less and smaller scoring
marks observed on the nose of PCD inserts.

YD101_91m/min
_ 4min10sec

YD101_91m/min
_5min0sec

YD101_91m/min
_7min0sec
YD101_400fsm_n
um3_3min0sec_20x

Figure 29: The scoring marks on the noses of YD101 inserts

40

I.3.2

Wear characteristics of tools inserts at high DOC

I.3.2.1 Crater wear at high DOC
It is interesting that the catastrophic failure from tool fracture of the cutting edge
only happened after a cutting distance of more than 400m at high cutting speeds. At
a low cutting speed, the chipping happened randomly on both carbide and PCD
inserts of all grades at each cutting speed. Meanwhile, catastrophic fracture
happened on most of the YD201 and PCD1210 rake inserts at 91 m/min and 61
m/min with low DOC, respectively. The possible cause for chipping and fracture with
a low DOC was high temperature gradients and stress concentration along the short
chip-tool contact lengths as evident in the study of temperature and stress distribution
at the rake face of the tool using the FEM simulation to be discussed in the next
section.
The 2D and 3D images of fresh and worn carbide inserts are presented in Figures
30-32 for all three cutting speeds. Interestingly, the results from the second set show
smoother crater surface, a representative of thermochemical wear (dissolution and/or
diffusion) on the carbide inserts. The smooth crater surfaces were seen on carbide
inserts at even low cutting speed (200sfm) after cutting 5 min. The carbide inserts
with the cutting time of less than 5 minutes at 200sfm showed a few abrasive marks
but mainly dissolution/diffusion wear (insert after cutting 2.4min in Figure 30). The
crater surfaces became more even and shinier as the cutting speed increases. The
dissolution/diffusion may be more dominant at the medium and high cutting speed.
As reported in the literature [Ezugwu, 1997; Hartung, 1982; Narutaki, 1983], the
temperatures of medium and high cutting speeds with carbide tools were above the
41

transus temperature leading to acceleration of dissolution and diffusion rates. The
crater wear developed closer to the cutting edge on the inserts at all cutting speeds.
Therefore, the cutting edge of carbide inserts was weaken and chipped quickly. This
phenomenon resulted from the high temperature near the cutting edge. The stable
and sharp cutting edge was an important factor for long tool life in cutting Ti alloys. In
machining Ti, not only were the cutting temperatures higher but also the crater wear
occurred near the cutting edge alloys compared to machining ferrous materials
[Ezugwu, 1997].
To represent the crater depth information, the 2D crater wear profiles along chip
flow direction, where the maximum of crater was expected happen, were plotted
various locations (88th, 108th, 128th and 148th sections in Figure 30). These maximum
crater depths were obtained by taking the maximum depth value from the 2D wear
profiles. The 2D crater profiles of the carbide inserts were presented in Figures 33
and Figure 34. Figures 33 showed the 2D crater wear profiles at 88th and 108th
locations with the consistent depth of crater wear which presented the evidence of
smooth crater wear on carbide inserts.

42

times higher than that of 61 m/min and two times higher than that of 91 m/min,
respectively. Figure 43 compared the crater wear rate of carbide insets at low at high
DOC. As expected, the wear rate at high DOC was much higher than that at low
DOC at given cutting speed.

0

50

YD101
dc=1.2mm

40

W ear rate (µm/min)

Maxim um Depth (µm )

-20

-40

-60

122m/min
91m/min
61m/min

-80

122m/min
91m/min
61m/min

YD101
dc=1.2mm

30

20

10

-100

0
50

100

150

200

250

300

350

400

450

50

100

Cutting Length (m)

150

200

250

300

350

400

450

Cutting Length (m)

Figure 42: The maximum depth and wear rate of crater wear on carbide inserts
(YD101) at high DOC

50

122m/min_low DOC
91m/min_low DOC
61m/min_low DOC
122m/min_high DOC
91m/min_high DOC
61m/min_high DOC

Wear rate (µm/min)

40

30

YD101

20

10

0
0

100

200

300

400

500

600

700

800

Cutting Length (m)

Figure 43: Comparison in wear rate of crater wear on carbide inserts (YD101) at low
DOC and high DOC

51

I.3.2.2 Nose wear and flank wear at high DOC
The nose and flank wear on the carbide and PCD1510 inserts at the longest
cutting time of three cutting speeds were presented in Figures 44-47. In appareance,
the carbide inserts showed even (in depth) nose and flank wear in appearance while
uneven nose and flank were observed on PCD1510 inserts at the low and medium
cutting speeds. In the literature, many researchers reported that titanium galls and
welds to tool material lead to the chipping of the cutting edge. The welded or
fractured chipping abraded the nose and flank face of tools causing scoring marks
when the tool travels along the feed direction. The scoring marks were observed with
both carbide and PCD1510 inserts. However, these scoring marks appeared more
frequently with smooth and clearly appearance than those of PCD1510. Obviously,
the temperatures at the flank face were much lower than those at the rake face.
Therefore, the dissolution and diffusion mechanisms which cause smooth wear
cannot be dominant in the flank face. Then, a question here is what causes these
scoring marks on the flank faces of the carbide and PCD1510 inserts, respectively.
The answer will be discussed in next section.
At the high cutting speed, as expected, with high temperature and less edge
stability on the rake face, both carbide and PCD1510 inserts experienced
catastrophic fracture.

The inserts with a large fractured area suppressed the

appearances of the steady-state wear patterns observed in nose wear, flank wear
and crater wear. Figure 48 shows a very large damaged area on the carbide inserts
resulted from a large fracture. For the cutting speed of 122 m/min, the carbide inserts
last up to 2 minutes, and the PCD1510 tools withstand for up to 3.5 minutes.

52

Furthermore, the main failure of the carbide inserts was at the nose while the flank
face of PCD1510 was more damaged. The failure of the PCD1510 inserts at all
cutting speeds was affected by the notch wear as shown in Figure 47. The notch
wear occurred at most cutting speeds and cutting times. It happened in the earlier
stage of cutting and become more severe as the cutting speed increases.

YD101_0m/min_0min
(fresh tool)

YD101_61m/min_
5.2min

YD101_0 m/min_0 min

YD101_61 m/min_5.2 min

th
2D	profile	
256 		
st
at	1 	
section 128th		

100µ
µm
YD101_91m/min_
4.5min

YD101_122m/min_
2.2min

YD101_91 m/min_4.5 min

2D profile at
st
1 section
th

128
th
256

even	(in	depth)	
wear

uneven	(in	depth)	
wear
YD101_122 m/min_2.2 min

a) 2D images

b) 3D images

Figure 44: Nose wear on carbide YD101 inserts at the longest cutting time of three
cutting speeds.

53

PCD1510 _0m/min_
0min (fresh tool)

PCD1510_0 m/min_0 min

PCD1510_61m/min_
6.8min

PCD1510_61 m/min_6.8 min

256th		
2D	profile	
at	1st	
128th		
section
section

100µ
µm

PCD1510_122m/min_
3.5min

PCD1510_91m/min_
5.4min

2D	profile	at	
1st	section

PCD1510_91 m/min_5.4 min

uneven	(in	depth)	
wear

128th
256th

uneven	(in	depth)	
wear

PCD1510_122 m/min_3.5 min

a) 2D images

b) 3D images

Figure 45: Nose wear on the nose of PCD1510 inserts at the longest cutting time of
three cutting speeds.

YD101_0m/min_0min
(fresh tool)

YD101_61m/min_
6.8min

YD101_0 m/min_0 min

YD101_61 m/min_6.8 min

YD101_91 m/min_4.5 min

YD101_122 m/min_2.2 min

100µm
YD101_91m/min_
4.5mim

YD101_122m/min_
2.2min

uneven	(in	depth)	
wear
a) 2D images
b) 3D images
Figure 46: The flank wear on carbide YD101 inserts at the longest cutting time of
three cutting speeds.

54

PCD1510_61m/min_
6.8min

PCD1510_0m/min_
0min (fresh tool)

PCD1510_0 m/min_0 min

PCD1510_61m/min_6.8 min

Notch wear
100µ
µm
PCD1510_91m/min_
5.5min

PCD1510_122m/min_
3.5min

PCD1510_91m/min_5.5 min

PCD1510_122m/min_3.5 min

Notch
wear

uneven	(in	depth)	
wear
a) 2D images

b) 3D images

Figure 47: The flank wear on PCD1510 inserts at the longest cutting time of three
cutting speeds.

YD101_122m/min_2.6min
250µm

YD101_122m/min_2.7min YD101_122m/min_3.5min

Figure 48: Nose damage of the carbide inserts (YD101)
The 2D nose and flank wear profiles were also plotted for each tool type as a
function of cutting distances for three cutting speeds. The 2D wear profiles provide
wear land as well as wear depth which was not determined with typical microscope
and SEM images reported in literatures. The nose and flank wear profiles of the
carbide and PCD1510 inserts can be compared in Figures 49-51. The 2D wear
profiles provided consistent data of nose and flank wear land at four different
locations (88th, 108th, 128th, 148th sections shown Figure 44). Because of the multiple

55

scoring marks, however, the 2D wear profiles was consistent in depth, especially for
the carbide inserts at the medium and high cutting speed and the PCD1510 inserts in
all cutting speeds. This happened in all cutting speeds with PCD1510 insert because
of uneven wear which grain fractured out was also contributed to. Similar to crater
wear, flank wear land at high DOC was much higher than low DOC as shown in
Figure 54.
10

10

0

0

-20

Nose
Nose wear
wearat
at200sfm
61 m/min
th
thlocation
128
128 location

-10

Depth(µ m)

Depth(µ m)

-10

-30

-40

Flank wear
Flank
wearatat200sfm
61 m/min
th
th
128
location
128 location

-20

-30

-40

YD101_2min24sec_146m

YD101_2min24sec_146m
PCD1510
PCD1500_2min30sec_150m
YD101_6min48sec_414m
PCD1510
PCD1500_6min48sec_414m

PCD1510
PCD1500_2min30sec_150m
-50

-60
-300

-50

YD101_6min48sec_414m
PCD1500_6min48sec_414m
PCD1510
-250

-200

-150

-100

-50

-60
-300

0

-250

Distance to cutting edge(µm)

-200

-150

-100

-50

0

Distance to cutting edge(µm)

Figure 49: The 2D profiles of nose and flank wear at 128th section of carbide YD101
and PCD1510 inserts at 61m/min.
10

10

0

0

-20

Nosewear
wear
Nose
at at
91300sfm
m/min
th th
148
location
148 location

-10

Depth(µm)

Depth(µm)

-10

-30

-40

-50

-60
-300

-200

-150

-100

148 location

-50

-50

-60
-300

0

Distance to cutting edge(µm)

Flank
wear at 91 m/min
th
Flank
wear at 300sfm
148 th
location

-30

-40

YD101_1min36sec_147m
PCD1510
PCD1500_1min36sec_144m
YD101_4min30sec_413m
PCD1500_5min24sec_496m
PCD1510
-250

-20

YD101_1min36sec_147m
PCD1500_1min36sec_144m
PCD1510
YD101_4min30sec_413m
PCD1500_5min24sec_496m
PCD1510
-250

-200

-150

-100

-50

0

Distance to cutting edge(µm)

Figure 50: The 2D profiles of nose and flank at 148th section of carbide YD101 and
PCD1510 inserts at 91m/min.

56

10

0

0

-10

-10

Nose
wear at 122 m/min
th wear at 400sfm
Nose
108 th location
108 location

-20

Depth(µ m)

Depth(µm)

10

-30

-40

-60
-300

-250

-200

-150

-30

Flank
wear
at 400sfm
Nose
wear
at 122m/min
th th
108
location
108 location

-40

YD101_1min6sec_135m
PCD1510
PCD1500_1min12sec_149m
YD101_2min12sec_262m
PCD1510
PCD1500_2min42sec_327m

-50

-20

YD101_1min6sec_135m
PCD1510
PCD1500_1min12sec_149m
YD101_2min12sec_262m
PCD1510
PCD1500_2min42sec_327m

-50

-100

-50

-60
-300

0

Distance to cutting edge(µm)

-250

-200

-150

-100

-50

0

Distance to cutting edge(µm)

Figure 51: The 2D profiles of nose and flank wear at 108th section of carbide YD101
and PCD1510 inserts at 122m/min.

350

350

Flank wear (µm)

300

122m/min
91m/min
61m/min

250

200

150

100

dc=1.2mm
PCD1510

122m/min
91m/min
61m/min

300

Flank wear (µm)

dc=1.2mm
YD101

250

200

150

100

50

50

0
50

100

150

200

250

300

350

400

450

100

Cutting Length (m)

200

300

400

500

600

Cutting Length (m)

Figure 52: The evolution of flank wear land on the flank face of YD101 and PCD1510
inserts at various cutting speeds.

57

350

dc=1.2mm
V=61m/min

F la n k w e a r l a n d V B ( µ m )

300

F la n k w e a r ( µ m )

300

YD101
PCD1510

250

350

dc=1.2mm
V=91m/min

YD101
PCD1510

250

200

200

150

150

100

100

100

50

50

0

0
100

200

300

400

50

0
100

500

YD101
PCD1510

dc=1.2mm
V=122m/min

300

250

200

150

F la n k w e a r la n d V B ( µ m )

350

200

300

400

500

600

50

100

Cutting Length (m)

Cutting Length (m)

150

200

250

300

350

400

450

Cutting Length (m)

Figure 53: The comparison of flank wear land of carbide and PCD1510 inserts.

400

500

YD101_high DOC
YD101_low DOC
PCD1510_high DOC
PCD1210_low DOC

V=91m/min

400

F lank wear (µm )

F lank wear (µm )

300

200

100

YD101_high DOC
YD101_low DOC
PCD1510_high DOC
PCD1210_low DOC

V=122m/min
300

200

100

0

0
0

100

200

300

400

500

600

700

800

0

Cutting Length (m)

100

200

300

400

Cutting Length (m)

Figure 54: The comparison of flank wear land at low and high DOC.

58

500

600

I.3.3

SEM images and element mapping results

Figure 55 shows the adhesion layers on the PCD1510 and carbide inserts for all
cutting speeds in the second set. Based on the images, the adhesion layers on
PCD1510 inserts were more stable and uniform. These layers are known to protect
the cutting edge during cutting process. Those layers strongly adhered to the rake
face and cutting edge of the inserts. Because of this, the fragments of a tool material
were pulled out when the layers were detached by the flowing chip. To determine the
possible reaction between work and tool materials, the elemental mapping was
conducted in the adhesion layers. Figures 56 and 57 represent the elemental
composition of the adhesion layer represented by the white and dark areas (shown in
Figure 55) for both PCD1510 and carbide inserts. Despite the deviation caused by
the carbon contamination, the weight content of carbon increased and that of titanium
decreased in the dark area as the cutting speed increased. The dark areas on the
inserts from the low and medium cutting speeds could be a compound of Ti and C
(TiCx). The standard atomic weight ratio of C:Ti is 12u: 47.68u (about 1:4). Based on
the weight content in the dark area, only adherent layer at low cutting speed with
PCD1510 had the weight ratio of C:Ti at 4:1 which converted into 1:1 in atom ratio (or
TiC). This result does not provide the definitive proof of the formation of TiC in the
adhesion layer at the low cutting speed. However, the results provided the low
possibility of the TiC formation at the medium and high cutting speeds because the
weight ratio C:Ti was less than 4:1 (or 1:1 in atom ratio) at the medium and high
cutting speed as shown in Figures 53 for PCD inserts and 54 for carbide inserts.

59

Hartung and Kramer [1982] and Arrazola et al. [2009] claimed that the TiC layer did
not form at the high cutting speed.

PCD_91 m/min_5.4min

PCD_61 m/min_5.4min

PCD_unworn

PCD_122 m/min_3.5min

white area

dark area

dark area

dark area
white area
YD101_unworn

YD101_61	m/min_6.8min

YD101_91 m/min_4.5min YD101_122 m/min_2.2min

Figure 55: The SEM images of adhesion layer on the PCD1510 and carbide inserts in
the second set (high DOC).

100
Unworn tool
Adherent layer
Adherent layer
Adherent layer
Adherent layer
Adherent layer
Adherent layer

Weight content (%)

80

60

at
at
at
at
at
at

200sfm
300sfm
400sfm
200sfm
300sfm
400sfm

(white)
(white)
(white)
(dark)
(dark)
(dark)

(dc=1.2mm)
40

20

0

C

Al

Ti

V

Co

W

Figure 56: The elemental contents of adherent layers at white and dark area on
PCD1510 inserts

60

100

W eight content (%)

80

60

(dc=1.2mm)

Unworn tool
Adherent layer at 200sfm (white)
Adherent layer at 300sfm (white)
Adherent layer at 400sfm (white)
Adherent layer at 200sfm (dark)
Adherent layer at 300sfm (dark)
Adherent layer at 400sfm (dark)

40

20

0

C

O

Al

Ti

V

Co

W

Figure 57: The elemental contents of adherent layers at white and dark area on
carbide YD101 inserts

The SEM experiments were also conducted on the chips to determine if the tool
material was dissolved into or attached to the chips. Figure 58 showed the SEM
images of the tool-chip interface with the attached particles. The elemental mapping
of the interface shown in Figure 59 indicates no traces of tool constituents (W, Co).
To be more certain, the elemental mapping was performed on those particles (see
Figure 60) to find high carbon and oxygen contents as shown in the element mapping
of Figure 61. This indicates that those particles were the carbon contaminations, not
the tool material.

61

Chip_PCD1510_61 m/min

Chip_PCD1510_91 m/min Chip_PCD1510_122m/min

attached particles
attached particles

Chip_YD101_61 m/min

Chip_YD101_122m/min

Chip_YD101_91 m/min

attached particles

attached particles

Figure 58: The SEM images of chip-tool interface of chips generated with carbide and
PCD1510 inserts.

100

Weight content (%)

80

Bulk-Ti
Chip at 200sfm (PCD1500)
(PCD1510)
(PCD1510)
Chip at 300sfm (PCD1500)
(PCD1510)
Chip at 400sfm (PCD1500)
Chip at 200sfm (YD101)
Chip at 300sfm (YD101)
Chip at 400sfm (YD101)

Chips
(dc=1.2mm)

60

40

20

0

C

Al

Ti

V

W

Figure 59: Elemental content of chips with carbide and PCD1510 inserts for three
cutting speeds

62

Chip with PCD1510_122m/min

Chip with YD101_122m/min

Figure 60: The SEM images of adhered particles on the chips.

100
Particle on chip at 400sfm (PCD1510)
(PCD 1500)
Particle on chip at 400sfm (YD101)

Chips
(dc=1.2mm)

Weight content (%)

80

60

High content of Ti from
the background (chip)

40

20

0

C

O

Al

Ti

V

W

Figure 61: Elemental content of adherent particles on the chips.
I.3.4

Chip morphology

Figure 62 presents SEM image of chip morphology with the carbide and PCD1510
inserts at three cutting speeds. The segmented chips were observed in all cutting
conditions with both carbide and PCD1510 inserts. Figure 63 presents five
parameters (height of peak, height of valleys, spacing of peaks, spacing of valleys,
and spacing of peak-to-valley) represented for chip morphology. The morphology of

63

chip generated with carbide and PCD1510 inserts was shown Figure 64. Statistical
analysis for height of peaks and valleys were presented in Figure 65 and Figure 66.
The higher height of peaks and valleys were found at the high cutting speed for both
carbide and PCD1510 inserts. The comparative data for chip morphology between
carbide and PCD1510 inserts was summarized in Figure 67. The carbide tool insert
showed segmented chips with higher peaks and shorter spacing of peaks as well as
spacing valleys (lower serrated frequency).

Chip_PCD1510_
61 m/min

Chip_YD101
_61 m/min

Chip_PCD1510_
91 m/min

Chip_PCD1510_
122 m/min

Chip_YD101
_91 m/min

Chip_YD101_
122 m/min

100µ
µm
Figure 62: The chip morphology (top view).

64

Height of peak

Spacing of peaks

Spacing of
peak-to-valley
Height of valley

Spacing of valleys

Figure 63: Five parameters represented for chip morphology (side view)

a) YD100 at 61 m/min

d) PCD1510 at 61 m/min

b) YD100 at 91 m/min

e) PCD1510 at 91 m/min

c) YD100 at 122 m/min

f) PCD1510 at 122 m/min

Figure 64: Chip morphology with carbide and PCD1510 inserts at all cutting speeds

65

YD101_61m/min
YD101_200sfm

4

0.005

2
0

Frequency (counts)

8

100
200
300
400
Height of peaks-Bins(µm)
YD101_300sfm
YD101_91m/min

0
500

0.01
0.008

6

0.006
4
0.004
2
0

0.002
0

100
200
300
400
Height of peaks-Bins(µm)
YD101_400sfm
YD101_122
m/min

0
500

6

0.015

4

0.01

2

0

0.005

0

100
200
300
400
Height of peaks-Bins(µm)

0.02

15

0.015
10
0.01
5
0

Frequency (counts)

6

Frequency (counts)

0.01

0.025

20

0.005
0

100
200
300
400
Height of peaks-Bins(µm)
PCD1510_91m/min
PCD1500_300sfm

0.015

10

0.01

5

0.005

0

100
200
300
400
Height of peaks-Bins(µm)
PCD1500_400sfm
PCD1510_122
m/min

20

0
500

0
500

0.025
0.02

15

0.015
10
0.01
5
0

0.005
0

100
200
300
400
Height of peaks-Bins(µm)

Figure 65: Statistical data for height of peaks.

66

0
500

15

0

Frequency (counts)

Frequency (counts)

8

0

Frequency (counts)

PCD1500_200sfm
PCD1510_61m/min
0.015

10

0
500

PCD1510_61m/min
PCD1500_200sfm

8

0.008

6

0.006

4

0.004

2

0.002

0

0

100
200
300
400
Spacing of peaks-Bins(µm)
YD101_300sfm
YD101_91m/min

12
Frequency (counts)

10

0
500

6
0.004
4

0

0.002

0

100
200
300
400
Spacing of peaks-Bins(µm)

0.015

10

0.01

5

0.005

0

20

0.01

0
500

15
0.008
10

0.006
0.004

5
0.002
0

0

10

0.008

8

0.006
4
0.004
2

0

0.002

0

100
200
300
400
Spacing of peaks-Bins(µm)

100
200
300
400
Spacing of peaks-Bins(µm)

0
500

PCD1510_122
PCD1500_400sfm m/min
0.01

0
500

Frequency (counts)

Frequency (counts)

6

0
500

0.012

YD101_122
m/min
YD101_400sfm
8

100
200
300
400
Spacing of peaks-Bins(µm)

PCD1510_91m/min
PCD1500_300sfm

0.006

2

15

0

0.01
0.008

8

Frequency (counts)

0.01

Frequency (counts)

Frequency (counts)

YD101_61m/min
YD101_200sfm
10

0.015

0.01

6
4

0.005
2
0

0

100
200
300
400
Spacing of peaks-Bins(µm)

Figure 66: Statistical data for height of valleys.

67

0
500

Height
of peaks
Height of Peaks
350

Height
valleys
Height ofof
Valleys
350

YD101
PCD1500
PCD1510

300
Mean height of valleys (µm)

Mean height of peaks (µm)

300

250

200

150

100

250

200

150

100

50
40

YD101
PCD1500

50
60

80
100
120
Cutting speed (m/min)

140

40

60

Spacing
of peaks
Spacing of Peaks
260
YD101

240

YD101

240

PCD1500
PCD1510

PCD1500
PCD1510

220
Mean spacing of valleys (µm)

220
Mean spacing of peaks (µm)

140

Spacing
valleys
Spacing of of
Vaylleys

260

200
180
160
140
120
100

200
180
160
140
120
100

80

80

60

60

40
40

80
100
120
Cutting speed (m/min)

60

80
100
120
Cutting speed (m/min)

40
40

140

60

80
100
120
Cutting speed (m/min)

140

Figure 67: The comparison on chip morphology between carbide and PCD1510
inserts

68

I.4

CUTTING TEMPERATURE PROFILES WITH FEM SIMULATION

I.4.3

2D-FEM with the Johnson-Cook (JC) model

The cutting temperature played a very important role in identifying the main wear
mechanisms involved in the cutting process. Both mechanical and thermochemical
wear (dissolution and diffusion) are functions of temperature which were discussed in
detail in Chapter 4. Thus, the temperature profiles along the chip thickness and rake
and flank faces on the workpiece needed to be determined. In spite of the
development

of

experimental

techniques

(thermocouple,

infrared

camera,

temperature indicating liquid, etc.), these techniques only measure in-situ local
temperatures. However, this study requires the estimation of temperature profiles on
both the tool and the workpiece. Therefore, Finite Element Analysis (FEA) software
was utilized. Many researchers have been using FEA simulation of cutting process to
predict the temperature distribution. This study also used 2D-FEM with the JohnsonCook (JC) model and Arbitrary Lagrangian-Eulerian (ALE) formulation to obtain the
temperature profile. The simulations were run with few sets of JC coefficients of the
Ti64 for comparison as shown in Table 6. However, the difference in the predicted
temperatures was minute. The developed model only simulated the case of
continuous chip as only rough estimation of the temperature profiles on tool and work
materials was required. The flow stress equation for the Johnson-Cook model is
given below.

σ ( ε , ε& , T ) =  A + B ( ε )


n

m

 ε&     T − T0  
 1 + C ln    1 − 
 
&

ε
−
T
T

 0     m 0  


69

(0.1)

where:

ε : the plastic strain
ε& : the strain rate ( s −1 )
ε& : the reference plastic strain rate ( s −1 ) ,
T : the temperature of the work material ( °C ) ,
Tm : the melting temperature of the work material (1399 °C )
T0 : the room temperature ( 20 °C )
and the Johnson-Cook coefficients:

A : the initial yield strength (MPa)
B : the hardening modulus (MPa)
C : the strain rate sensitivity coefficient
n : the hardening coefficient
m : the thermal softening coefficient.
Note

σ ( ε , ε& , Tm ) = 0 → there is no temperature effect to flow stress at melting

temperature.
Table 6: Johnson-Cooks coefficients for Ti-6Al-4V
Tm (°°C) A (Mpa) B (Mpa)

C

n

m

References

1650

880

695

0.04

0.36

0.8

This study [Meyer, 1998]

1650

870

990

0.028

0.25

0.8

Shivpuri et. al. [2002]

1630

783

498.4

0.028

0.28

1.03

Ozel et. al. [2004]

1600

870

990

0.008

1

1.4

Umbrello et. al. [2007]

1656

863

656

0.013

0.5

0.8

Meijer et. al. [2001]

70

I.4.4

2D-FEM JC model with heat transfer

The heat transfer simulation requires all three modes of heat transfer, conduction,
convection and radiation, which were investigated in order to improve the accuracy in
obtaining the temperature profile. The convection and radiation heat transfer was set
up at the exposed surfaces of chip (top, chip flow and chip-tool contact surfaces).
The surrounding was the air at room temperature with the appropriate properties
shown in Table 7. Based on this, the convection had an important effect on the
temperature profiles. Figure 68 shows the 2D simulation models with flat PCD and
10° positive rake angle PCD inserts. The model with the flat carbide insert was also
developed.
Convective
Chip flow (3)
surfaces

Tool

(3)

(1,2,4,5)
(4)

Top surface
(2)

Chip-tool
contact surface (4)

(2)

(1)
(5)
Workpiece
a) 0° rake angle

b) +10° rake angle

Figure 68: 2D orthogonal cutting FEM with heat transfer

71

Table 7: The parameters of air at 20°C and 1atm
Temperature

Density

Kinematic
Thermal
viscosity conductivity

Specific heat
capacity
7

µ.10

ρ (kg/m ) Cp (kJ/kg.K)
3

T (K)
293.5 K

1.1614

3

3

(Ns/m )

1.0007

6

α.10

ν.10 (m /s)

k.10
(W/m.K)

(m /s)

15.89

26.3

22.5

6

184.6

2

2

The heat transfer coefficient for free convection to the air was 5–25 (W/m2K) while
the value for forced convection was 10-200 (W/m2K). In this simulation, the forced
convection was appropriate because of the chip travels up to 2.03 m/sec
approximated by the cutting speed. Table 8 reports the value for heat transfer
coefficients for each cutting speed calculated based on Reynolds, Nusselt, Pardtl
number, and chip flow using the equations (1.2) to (1.7).
Local heat flux :

q′′ = h(Ts − T∞ )

W / m 2 

(0.2)

Total heat rate :

q = ∫ q′′dAs = (Ts − T∞ ) ∫ hdAs
As

As

or
q = hAs (Ts − T∞ ) ;

[W ]
(0.3)

The average heat transfer coefficient :

h=

1
As

∫

As

72

hdAs

(0.4)

The average heat transfer coefficient:

h=

NuL ⋅ k
L

(0.5)

Reynolds number:

Re L =

u∞ L

ν

=

ρ u∞ L
µ

(0.6)

Nusselt number:

Nu L = 0.664 ⋅ ( Re L )

1/ 2

( Pr )

1/3

(0.7)

Table 8: Reynolds, Nusselt number, and heat transfer coefficients for three cutting
speeds
Chip flow
(m/min)

Chip flow
(m/sec)

Chip's Length
(m)

Reynolds
number

Nusselt
number

h_avg

61

1.02 m/sec

0.0010

64.0

4.7

124.4

91.00

1.52 m/sec

0.0010

95.4

5.8

152.0

122.00

2.03 m/sec

0.0010

128.0

6.7

176.0

Although the majority of the heat generated during the process was from plastic
deformation, the simulation results showed that the friction has some influence on the
temperature profile. The comparison of chip-tool contact length between experiment
data and those of the simulation model was made to determine a reasonable friction
coefficient to be used in this model.
I.4.5

Cutting temperature profiles with 2D-FEM simulation

The temperature profiles on the chip along the rake face and through the thickness
of the chip were the main interest. Figure 69 plots the temperature profile along the

73

rake face starting at the end-point of the round corner (see Figure 69) with the friction
coefficient set at 0.8. The simulation results were adequate with the high temperature
profile for the high cutting speed. The maximum temperatures did not happen at the
cutting edge. The effects of heat transfer are reported in Figures 70-72 which show a
significant reduction of temperature. The maximum temperature from the simulations
without the heat transfer was along rake face while the simulations with the heat
transfer showed the maximum temperature close to the shear bands presented in
other publications [Sima, 2010; Y. C. Zhang, 2011; X. Zhang, 2011] with segmented
chips.

Temperature profile in the chip along rake face (µ=0.8)
1200
122m/min

1100

Temperature (°C)

91m/min
1000

61min/min

900
800

Temperature along this curve
700

End- point of the round corner

600

Cutting edge
553.9

526.4

498.9

471.4

443.9

416.4

388.9

361.4

333.9

306.4

278.9

251.4

223.9

196.4

168.9

141.4

113.9

86.4

58.9

31.4

3.9

0.0

500

Distance from the end point (mm)
Figure 69: Temperature profiles on the chip along tool-chip interface with PCD

74

the maximum
temperatures along
rake face

the maximum
temperatures close
to shear band

A

B
cutting edge

Figure 70: The temperature in Ti- turning with PCD rake angle: (A) Without heat
transfer, (B) with heat transfer (friction µ=0.35)

Temperature profile in the chip along top surface (V=61m/min, µ=0.35)
700

Temperature (°C)

600
500
400
w/o Convection and Radiation

300

With Convection and Radiation (h=20W/mK)
200

With Convection and Radiation (h=70W/mK)
With Convection and Radiation (h=150W/mK)

100

With Convection and Radiation (h=150W/mK) at Rake
0
1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20

Node along top surface (mm)
Figure 71: Effect of heat transfer to temperature profiles along the top surface of the
chip with PCD

75

900

V=61m/min, µ=0.35

Temperature (°C)

800

700

600

w/o Convection and Radiation
With Convection and Radiation (h=20W/mK)

500

With Convection and Radiation (h=70W/mK)
400

With Convection and Radiation (h=150W/mK)

0.57

0.54

0.51

0.48

0.45

0.42

0.39

0.35

0.32

0.29

0.26

0.23

0.20

0.17

0.15

0.12

0.10

0.08

0.06

0.04

0.03

0.02

0.00

0.01

With Convection and Radiation (h=150W/mK) at Rake

300

Distance from end- point (mm)
Figure 72: Temperature profiles on the chip along tool-chip interface with PCD with
and without heat transfer
1200

V=122m/min, µ=0.6, h=150W/mK
1100

Temperature (°C)

1000
900
800
700
600
500

This study (="A=880_B=695_n=0.36_m=0.8_C=0.0.04)
400

Shivpuri (A=870_B=990_n=0.25_m=0.8_C=0.28)
0.00
0.01
0.02
0.03
0.04
0.06
0.08
0.10
0.12
0.15
0.17
0.20
0.23
0.26
0.29
0.32
0.35
0.39
0.42
0.45
0.48
0.51
0.54
0.57

300

Distance from end-rounded point(µm)
Figure 73: Effect of JC parameter to temperature profiles on the chip along tool-chip
interface with PCD.

76

Figure 74 represents the significant effects of the friction coefficient in FE
simulation on the temperature profile along the rake face. To estimate the reasonable
friction coefficients for FEM simulation, the chip-tool contact lengths between
simulation results and experiments were compared. These values of experiment
were approximated by the maximum width of crater wear on the rake face of inserts.
The simulated results shown in Figure 75 also indicate that the chip-tool contact
length depended strongly on the friction coefficient. There was a small difference in
the length among three cutting speeds. As shown in Figure 75, the chip-tool contact
length of PCD and PCD+10° in the turning test was most likely close to those of the
simulation model with the friction coefficients of 0.6. For the simplicity, 0.6 was used
as friction coefficients for both PCD. The simulated temperature for all cutting speed
with friction coefficient of 0.6 was shown in Figure 76. The temperatures at tool-chip
interface were above transus temperature at medium and high cutting speed. The
maximum cutting temperatures of 2D FEM simulation with carbide insert were
reported at least 100 °C higher than those PCD at all cutting speeds [Schrock, 2012].
This thesis accepted that results although the FEM model was simulated without heat
transfer.

77

1100

µ=0.6
0.6
friction

0.8
friction
µ=0.8

µ=0.3
0.35
friction

1000

Temperature (°C)

V=61 m/min
900
800
700
600

553.9

526.4

498.9

471.4

443.9

416.4

388.9

361.4

333.9

306.4

278.9

251.4

223.9

196.4

168.9

141.4

113.9

86.4

58.9

31.4

3.9

0.0

500

Distance from end-point (mm)
Figure 74: Temperature profiles on the chip along tool-chip interface with PCD at
various friction coefficients

Chip- Tool contact length vs. friction

Contact length (mm)

µ=0.35
0.35
friction

µ=0.6
0.6
friction

0.8
friction
µ=0.8

PCD

PCD10

0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
400sfm
122m/min

91m/min
300sfm
Cutting Speed (sfm)

61m/min
200sfm

Figure 75: The chip contact length from experiment and simulation of PCD at various
friction values

78

1100

Temperature (°C)

1000
900
800
700

122 min/min

600

91 m/min

500

61 m/min

400

Transus temperature
0

0.1

0.2

0.3

0.4

0.5

0.6

Distance from the end point (mm)
a) Temperature profile in the chip along tool-chip interface with PCD
(µ=0.6)

1100

Temperature (°C)

1000
900
800
122 min/min

700

91 m/min

600

61 m/min
500
Transus temperature
400
0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Distance from the end point (mm)
b) Temperature profile in the chip along shear band with PCD (µ=0.6)
Figure 76: Temperature profiles on the chip with PCD inserts and µ=0.6 at various
cutting speeds.

79

Chapter 2: Evidence of phase change and root causes flank wear
and scoring marks with orientation imaging microscopy (OIM)
II.1

OIM SETUP FOR CRYSTAL ORIENTATION IN TI64 ALLOY AND CHIPS

To explain the wear mechanism on both rake and flank face, the microstructure of
Ti workpiece (before machining, to be called Bulk-Ti) and after machining (chip) was
investigated with Orientation Imaging Microscopy (OIM) based upon electronbackscattered diffraction (EBSD). Two samples (Sample A and Sample B) were cut
out from the Bulk-Ti to study the microstructure before machining in DOC and feed
rate directions as shown in Figure 77a. The microstructure in the YZ plane is noted to
be Sample A (in Feed direction) while that in the XY plane is Sample B (in DOC
direction). To accurately obtain the original structure of the work material and limit
effects deformation and heat-affected zones and heat of the cutting process, Wire
Electron Discharge Machining (EDM) was applied to cut off these samples from the
round bar of Ti workpiece. With Wire-EDM cutting, the heat-affected zones from the
process can be easily removed [Hasçalık, 2007]. After polishing, EBSD scans were
conducted at various locations on both samples. With the Bulk-Ti samples, the EBSD
was scanned on the region of 800µmx800 µm, step size 2 µm, exposed time of
0.13sec. To have the reliable data, EBDS was conducted at various regions (named
A1, A2, A3, B1, B2, B3) on both samples as showed in Figure 77a.

80

Y

Y

Sample

Z

A

Chips (PCD, dc=0.635 mm)

Sample B

Bulk-Ti

X

Z

Feed direction
X

Top of chip

Sample A
(YZ plane)

Lateral side
of chips

Chip
tool-chip
interface

A1

B1

A2

B2

A3

B3 Sample B
(XY plane)

YD101
PCD
Sample A & B in
Bulk-Ti
a) Bulk -Ti

Chips (dc=1.2 mm)
b) Chip

Figure 77: Samples for EBSD of material before (Bulk-Ti) and after (Chip) machining
The microstructures of chips generated with carbide and PCD inserts at all speeds
were investigated to compare with those of bulk-Ti as shown in Figure 77b. The
difference in the microstructures between bulk-Ti and chip contains the information
about cutting temperature, stress, plastic deformation, phase change, etc. The
scanned regions of 200µm x 200µm were taken on the chips with step size 1µm and
exposed time of 0.13sec.
The EBSD studies shown in Figure 78 were carried out using a CamScan 44 FE
scanning electron microscope, with a field emission gun operated at 20 kV using

81

TSL/Link OIM TM system. The working distance of 33 mm was used, and the
exposure time for each Kikuchi pattern was 0.08 s. The TSL data collection software
was connected with the EBSD camera to instantaneously convert the acquired
Kikuchi patterns into crystal orientations. Each crystal orientation thus obtained was
recorded as three Euler angles. The step size used for each scan varied and was
indicated in the result session. Two clean-up steps were performed on all the EBSD
datasets. The first step was to conduct a neighboring confidence index (CI)
correlation procedure, where a data point with a CI smaller than 0.07 (user defined) is
reassigned the orientation and CI of its neighbor with the highest CI.

-

e beam

Normal

EBSD camera

Direction
(ND)

Transverse
Direction
(RD)

Rolling Direction (RD)

Figure 78: EBSD configuration

82

II.2

OIM RESULTS AND DISCUSSION

II.2.1 Evidence of phase change (α →β) in machining of Ti64
The OIM scan of the microstructure is very important to indicate the indirect
evidence of phase change of titanium in the cutting process. Based on the average
results of six scan areas (A1, A2, A3, B1, B1, B2, B3) on the Bulk-Ti sample as
shown in Figure 79, the microstructure before machining was mostly the alpha phase
accounting for up to 99% of area fraction with the rest being the β-phase as
summarized in Table 9. The distribution of α-grain size in samples A and B was
plotted in Figure 80. It should be emphasized that that the average grain sizes of αphase and β-phase in the Bulk-Ti samples were 6.78 µm and 3.89µm, respectively.
However, the grain size in chips was much smaller, as shown in Figure 81. This was
likely the result of high deformation and recrystallization near the cutting zone on the
chips. The lateral side (see Figure 77b) of the chip recrystallized more than the chiptool interface as evident by the much smaller grain size on the side of chips. Only
small differences in the average grain size in relation to the cutting speeds and tool
materials were evident.

83

a-phase

b-phase

Figure 79: Microstructure of Ti64 alloy used in turning tests
Table 9: The area fraction of α-phase and the grain sizes of each phase in Bulk-Ti
Area

Area fraction
of α-Ti

Grain size of
α-Ti (µm)

A1

99.6%

6.93

3.99

A2

99.6%

6.86

3.54

A3

97.3%

6.52

3.61

B1

99.5%

6.74

3.45

B2

99.4%

6.68

4.74

B3

98.9%

6.95

4.46

Average

99.1%

6.78

3.97

84

Grain size
of β-Ti (µm)

14

35

α−Ti grains

Region A1
Region A2
Region A3

A re a fra c tio n (% )

12

Region B1
Region B2

α−Ti grains

30
25

Bulk-Ti (sample A)

10

A re a fra c tio n (% )

16

8
6
4

Bulk-Ti (sample B)

20
15
10
5

2

0

0

0

10

20

30

40

50

60

70

0

Grain Size (µm)

10

20

30

40

50

60

Grain Size (µm)

Figure 80: Grain size distribution of α-Ti in un-deformed work material (Bulk-Ti)

10

10

8

6

Microstructure on the
lateral side of chip

4

2

(dc=1.2mm)
A v e r a g e D ia m e te r ( µ m )

A v e r a g e D ia m e te r ( µ m )

(dc=.635mm)

Bulk- Ti
Chip at 200sfm (PCD)
Chip at 300sfm (PCD)
Chip at 400sfm (PCD)

0

8

6

Microstructure
on tool-chip
interface of
chip

Bulk- Ti
Chip at 200sfm (PCD)
Chip at 300sfm (PCD)
Chip at 400sfm (PCD)
Chip at 200sfm (YD101)
Chip at 300sfm (YD101)
Chip at 400sfm (YD101)

4

2

0

alpha-crystals

beta-crystals

alpha-crystals

beta-crystals

Figure 81: The comparison of average diameter of grains in the bulk-Ti and in the
deformed material (chips).
In Ti machining, the evidence of phase change is not easy to prove. Due to the
low thermal conductivity, the temperatures of chip are expected to be higher than the
transus temperature (Ts) at which the phase transformation (α to β) occurs. At

85

atmospheric pressure, the α-phase in Ti64 transformed to the β-phase when the
temperature increased above Ts = 995°C. The β-phase transformed back to the αphase when the temperature was cooled to below Ts. However, due to the low
thermal conductivities, the chips cooled slowly. This process helped the most of βphase to have enough time to transform back to the α-phase. Therefore, the amount
of β-phase formed during the cutting process cannot be determined directly. Because
the β→α transformation has a well-defined interphase misorientation relationship, it is
possible to see evidence of α →β phase by determining the misorientations in the α
phase.
In a polycrystalline material, the misorientation is defined as the difference in
crystallographic orientation between two adjacent crystals. When Ti64 is slowly
cooled below the beta transus temperature, α-plates form with their base planes
parallel to the transformation planes (110) of the β-phase as shown in Figure 82. The
α-plates grow relatively slowly in perpendicular direction to the planes and much
faster along the planes making acicular (plate-like) structures (see Figure 83).
Because of the symmetry of the β-phase, the transformation planes are oriented 60°
& 90° to each other resulting in alpha grains with precisely 60° & 90° misorientation
as shown in Figure 82. A high fraction of boundary in the α-phase with 60° & 90°
misorientations indicates proof of the phase change roughly.

86

α-phase
(110)

90° misorientation

(110)

(110)

basal plane
β-phase

60° misorientation

(110)

Figure 82: Burgers’ orientation relationship in β → α transformation

Figure 83: Microstructures achieved at various intermediate temperatures by slowly
cooling from above the β transus

87

Figure 84 shows the higher peak of 60° and 90° misorientations in the chips than
those in un-deformed work material. In addition, the portions of the β-phase in the
chips were greater than those in Bulk-Ti as presented in Figure 85. This evidence
provides a strong indication of the phase change during cutting process. It should be
noted that the determination of the phase change is valuable in predicting the
diffusion/dissolution wear. In observing the chips between low and high DOCs, a
higher fraction of the β-phase and the higher peak of 60° and 90° misorientations
with low DOC were observed. However, OIM scans were conducted on the side and
tool-chip interface of chips for low and high DOC, respectively. One possible
explanation is the higher and more uniform temperatures at the chip-tool interface
while the temperatures on the side were closer to the transus temperature and the
temperature gradients were greater due to the faster cooling on this side. The fast
cooling process retains more β-phase in the chip. 3D FEM simulation which was not
presented in this thesis may reveal the difference in terms of temperatures with two
DOCs in order to have a more convincing explanation. Figure 86 shows the
distribution of the β-phase among the α-phase in the underside of the chip at high
DOC for both carbide and PCD inserts for three cutting speeds. It was interesting to
see that, at high DOC, the β-phase crystals were distributed mostly at shear band
where the temperature was reported highest in most FEM simulation with serrated
chips.

88

40

40

(dc=0.635mm)
α−Ti crystals

20

60 & 90° misorientations
mostly caused by
phase change

10

0

Bulk-Ti
Chips PCD
Chips YD101

(dc=1.2mm)
α−Ti crystals

30

A r e a fr a c tio n ( % )

30

A re a fra c tio n (% )

Bulk-Ti
Chip PCD

20

60 & 90° misorientations
mostly caused by
phase change

10

0

0

20

40

60

80

0

Misorenntation (degree)

20

40

60

80

Misorenntation (degree)

Figure 84: The comparison of misorientation in α-Ti crystals in un-deformed (bulk-Ti)
and deformed (chips) work material at low and high DOC

30

25

30

Bulk- Ti
Chip at 200sfm (PCD)
Chip at 300sfm (PCD)
Chip at 400sfm (PCD)

(dc=.635mm)
25

15

20

F ra tio n (% )

F ra tio n (% )

20

15

10

10

5

5

0

0

Bulk- Ti
Chip at 200sfm (PCD)
Chip at 300sfm (PCD)
Chip at 400sfm (PCD)
Chip at 200sfm (YD101)
Chip at 300sfm (YD101)
Chip at 400sfm (YD101)

(dc=1.2mm)

Figure 85: The comparison area fraction of β-Ti crystals in un-deformed (bulk-Ti) and
deformed (chips) materials.

89

PCD1510 at 61 m/min

PCD1510 at 91 m/min

PCD1510 at 122m/min

β-Ti

YD101 at 61 m/min

YD101 at 91 m/min

YD101 at 122m/min

Figure 86: The distribution of the β-Ti (dark) and the α-Ti (colored) in the chips at
high DOC.
II.2.2 Root causes flank wear and scoring marks in Ti alloys machining
Scoring marks were typically found on the flank wear in machining of Ti alloys
[Bermingham, 2011; Dearnley, 1986; Hughes, 2004; Narutaki, 1983; Venugopal,
2007; Wright, 1981] as shown in Figure 87. This section mainly focuses on the
microstructure of a work material in an attempt to determine the root cause for flank
wear and scoring marks on flank face.

90

Dearnley, 1986

Wright, 1981

Hughes, 2004

Narutaki, 1983

Bermingham, 2011

Venugopal, 2007

Figure 87: Scoring marks on flank face of inserts in machining of Ti alloys
The α-Ti has the characteristic anisotropy in mechanical properties. The hardness
of α-Ti in the c-direction (0001) was reported to be much harder (~1.5 time) than any
other directions [Britton, 2009; Kwon, 2013]. Figure 6 shows the anisotropy in
hardness. Based on these data, the α-Ti with the c-axis within ±20° interacting with
the wear land of the inserts was chosen as the grains with ‘hard’ orientations or the
‘hard’ α-grains respect to the inserts. During machining, the tool will be worn or
damaged when the ‘hard’ α-grains interacts with the tool surfaces.

91

Average
‘hard’ α-grains
c-axis

Tool
Flank face

Figure 88: Hardness of α-Ti as a function of the declination angles between c-axis to
vertical line [after Britton, 2009] and the ‘hard’ α-grains respect to flank face.
The results of EBSD scan showed that Bulk-Ti consists mostly of the α-Ti.
Therefore, the anisotropy in hardness of α-grain had significant effect to wear. As
shown in Figure 89, Sample A shows the distribution of the hard-orientation grains
considering the hard orientation within 90±20° to the rake face. It shows that the
clusters of hard α-grains in the Y-Z plane (the declination angle within ±20°) are
distributed in the direction of Z. Sample B has the hard-orientation grains
agglomerate more less at the 30° from the negative Y axis as shown in Figure 89.

92

Y
Bulk-Ti
workpiece

90°
rotation

Z
X

A1

c-axis

B1

α-Ti

A2

B2

Figure 89: The distribution of α-crystal in hard orientation (red color) in Bulk-Ti
sample respected to flank face of tool along feed direction
The interaction with the tool in the hard α-grains can cause the abrasion, chipping
and the fracture. The size of the fracture may depend on the size of α-clusters in the
Bulk-Ti as well as the detailed interaction conditions (e.g. impact angle and
orientation). Table 10 summarizes the average dimensions of the ‘hard’ α-clusters
based on the microstructures in both XY and YZ faces. Based on these OIM
measurements, the size of the ‘hard’ α-clusters is roughly estimated and represented
in Figure 90. In this particular microstructure, the ‘hard’ α-cluster can be easily
approximated into an ellipsoid. Figure 91 shows how an ellipsoidal hard α-cluster
interacts with the flank surface, generating the flank wear. In addition, the ‘hard’
clusters happen to be in a certain plane, horizontally on the YZ face and on the 30°orientation on the XY face. Machining the same ‘hard’ α-cluster range in the Bulk-Ti,
however, the much harder PCD insert has much better wear resistance, resulting in

93

smaller and shallower scoring marks on flank wear compared to the carbide insert
as shown in Figure 92.
Table 10: The dimensions of the ‘hard’ α-cluster with Bulk-Ti
Minimum width
(µm)

Sample
Sample A
(YZ plane)
Sample B
(XY plane)

Maximum
width (µm)

Average
width (µm)

25.78

301.56

87.49

16.41

89.06

54.49

Y
Depth of Cut Direction

54µm
Bulk-Ti

30°

YZ plane
16µm

Feed Direction

89µm

301µm

87µm 26µm

Z

X

Figure 90: The size of ‘hard’ α cluster in Bulk-Ti.

94

Figure 91: Interaction of the ‘hard’ α-cluster and the inserts.

YD101_91m/min
4min10sec

YD101_91m/min
5min0sec

YD101_91m/min
7min0sec

YD101_122m/min
7min0sec

PCD1200_61m/min
6min0sec

PCD1200_122m/min_
4min0sec

100µm

Figure 92: The scoring marks on the carbides and PCD inserts at low DOC
Obviously, because of the flank temperature being much lower than the crater
temperature and the ‘rough’ wear surface, the dissolution and diffusion rate is not
significant for the flank wear. Discounting the macro-fractures which may be more

95

related to the microstructural variations, two distinct types of steady-state wear
patterns were identified, (a) micro-fracture at the cutting edge and (b) scoring marks.
Both types of damage can be caused by the “hard” α-phase grains and clusters.
However, micro-fracture can be affected by additional mechanisms, the adhesion and
detachment of the work material. These two patterns on the flank face manifest
themselves into four main types of observed damage presented in Table 11. Type I
is the micro-fracture which occurs near the cutting edge most likely as a result of the
impact with the ‘hard’ grain or cluster, which are observed only on the PCD inserts.
Type II occurred near the cutting edge in both carbide and PCD. The distinction
between Type I and II is the inverted shape of the fracture.
Both Types I and II are expected to be caused by the larger ‘hard’ α-clusters in a
work material. Because the incidences of Type I damage are much rare and only
occurred to the PCD inserts in our experiment, more detailed interaction conditions
must be explained with the adhesion layer. Many researchers [Dhar, 2002; Ezugwu,
1997; Hartung, 1982; Rahman, 2003] reported that Ti adheres to a tool material,
which leads to the chipping when detached. Our observations indicated that not only
was the adhesion on carbide inserts substantially more extensive than PCD inserts
but the adhesion layers were also mostly located on the crater surface as shown in
Figure 93. Consequently, the cutting edge of the carbide inserts was more
extensively damaged while the edge of the PCD inserts remained sharper. The
difference in the adhesion layer between carbide and PCD inserts enabled Type I to
cause damage on PCD inserts. With Type II, additional damage may occur as the
loose fragments can further damage the surface evident by the damage at the trailing

96

end of the fracture. Type III was the scoring marks abraded by the ‘hard’ α-grains or
cluster. Type IV was the combination of the first and third types of damage. Besides
of Type III, the more common damage on PCD inserts were Type I and II while those
of carbides were Type II and IV.
Table 11: Typical damage on the flank face
Typical damage

Explanation

Type I

Micro-fracture by the Impact of
the hard α-cluster only on PCD

Type II

Micro-fractures by the Impact of
the hard α-cluster

Type III

Scoring marks caused by the
‘hard’ α-crystal or cluster.

Type IV

The Combination of Type I and
Type III

97

PCD_unworn

PCD_200sfm_
5.4min

PCD_91
m/min_5.4min

PCD_400sfm_
3.5min

YD101_91
m/min_4.5min

YD101_400sfm_
2.2min

Adhesion
layer

Adhesion
layer

YD101_unworn

YD101_200sfm
_6.8min

Figure 93. Adhesion layer on the rake face of carbide and PCD inserts
The scoring marks, Type II damage, were much more prevalent than any other
types on both carbide and PCD inserts. However, the scoring marks on the carbides
were not as distinct as those of PCD. Our hypothesis is that all four types of damage
were caused by the ‘hard’ α-crystals and the adhesion may enhance Type I and II
damage. To substantiate this claim, the widths were measured for at least ten scoring
marks on both carbide and PCD inserts at various cutting lengths as shown in Figure
94. The distribution of width was presented using a ‘box and whisker’ plot with the
minimum, maximum and average width as shown in Figure 95 for low DOC and
Figure 96 for high DOC. These data were compared to the size range of the ‘hard’ αclusters obtained in Sample B (mostly close to tool flank face) shown in the light gray
boxes in Figures 95 and 96. Notice that the widths of the damage were substantially
smaller as only a smaller part of each ellipsoidal cluster interacts with the tool during
the abrasion process. The width of the scoring marks was larger with longer cutting
distance for both carbide and PCD tool as demonstrated in Figures 95 and 96 as the
probability of encountering larger clusters increases. Typically wider scoring marks

98

were observed at high cutting speed except the cutting speed of 122 m/min for
carbide inserts. This may happen because the tool material becomes much softer at
the high cutting temperature. The scoring marks on the carbide YD101 inserts were
smooth and consistent in the width of each scoring mark as shown in Figure 97. For
low DOC, with comparative grain size among carbide and PCD1200 and PCD1210
grades, grooves were smaller and less on PCD inserts than those on carbide inserts
for all cutting speeds. For high DOC, the sizes of grooves on PCD1510 inserts were
comparative to those on carbide inserts. As discussion early, diamond grains pulled
out grains was main wear mechanisms on both rake and flank faces of PCD inserts.
The scoring marks were produced from impact of hard α-clusters and flank face. The
size of scoring marks was equivalent to the size of one or two diamond grains
fractured out (much larger grain size of PCD1510 inserts).

Figure 94: Width (µm) of ten scoring marks on YD101
With similar phenomena, the comparable grain size among PCD1200, PCD1210
and carbide YD101, YD201 inserts, the scoring marks were narrower and less in

99

number for PCD inserts than those for carbide insert for all cutting speeds due to the
superior hardness of the PCD. For PCD inserts, the width of scoring marks was more
independent of cutting speed. PCD inserts have much higher hardness than the
carbide inserts and the ‘hard’ α-grain (6000kgf/mm2 for PCD, 1800kgf/mm2 for
carbides [Groover, 2007] and 550kgf/mm2 for α-grain in hard orientations [Britton,
2009] at room temperature) while their transverse rupture strength and fracture
toughness of PCD (1550MPa and 6.86MPa—m1/2 [Lammer, 1988]) were lower than
those of carbide insert (3133MPa and 10.18MPa—m1/2 [Fang, 2005]). Most of scoring
marks on PCD were rough with uneven width (and depth) as the pulled-out diamond
grains can abrade the tool surface, as shown in Figure 98. In addition, PCD inserts
were more prone to micro-fracture than the carbide inserts, an indication of lower
fracture toughness of PCD. The similar phenomenon was observed on the rake face
reported in [Schrock, 2014]. At high cutting speed, the micro-fractures are more
frequent and larger to suppress the appearances of the scoring marks on flank wear.
There is no difference between flat and positive tool in term of scoring marks due to
the end relief angle is the same for both inserts.

100

Width of scoring marks (µm)

140
120

hard α-Ti range

2Q Box

3Q Box

Mean

100
80
60
40
20
0

a) Vc = 61m/min, DOC = 0.635 mm
Wiidth of scoring marks (µm)

140
120

hard α-Ti range

2Q Box

3Q Box

Mean

100
80
60
40
20
0

b) Vc = 91 m/min, DOC = 0.635 mm
Wiidth of scoring marks (µm)

140
120

hard α-Ti range

2Q Box

3Q Box

Mean

100
80
60
40
20
0

c) Vc = 122m/min, DOC = 0.635 mm
Figure 95: Range and distribution of width of scoring marks of nose at low DOC
respect to ‘hard’ α-cluster size

101

Wiidth of scoring marks (µm)

140

hard α-Ti range

2Q Box

3Q Box

Mean

120
100
80
60
40
20
0

a) Vc =61m/min, DOC = 1.2 mm
Wiidth of scoring marks (µm)

140

hard α-Ti range

2Q Box

3Q Box

Mean

120
100
80
60
40
20
0

b) Vc =91m/min, DOC = 1.2 mm
Wiidth of scoring marks (µm)

140

hard α-Ti range

2Q Box

3Q Box

Mean

120
100
80
60
40
20
0

c) Vc =122m/min, DOC = 1.2 mm
Figure 96: Range and distribution of width of scoring marks on flank face at high
DOC respect to ‘hard’ α-cluster size

102

Cutting edge fractured out on
rake face

Smooth scoring marks
YD101_91 m/min 4min10sec

Tool fracture
100µm

50µm

Crack
Figure 97: Classifying of scoring marks on YD101 (Left: confocal image, Right: SEM
image)

Rough
scoring marks

PCD1200_200min 12min0sec
100µm

20µm

Figure 98: Classifying of scoring marks on PCD1200 (Left: confocal image, Right:
SEM image
For both carbide and PCD inserts, the flank wear is directly related to the cutting
speed. The thermal conductivity and the hardness of PCD are much higher
compared to those of the carbide inserts, resulting in lower cutting temperatures and
superior resistance to abrasion wear. Therefore, the flank wear rates of PCD inserts
were much smaller than those of carbide inserts. The tool fracture started at 91m/min
for both PCD and carbide inserts. These fractures are caused by the large ‘hard’ α-

103

cluster and the detachment of the adhesion layer. The fractured fragments of the
inserts can also abrade the tool evident by the groove marking emanating from the
fracture area. On PCD inserts, the tool wear including crater wear did not progress as
rapidly as those of the carbide inserts and more importantly the cutting edge
remained sharp and stable throughout the cutting process. In other words, the edges
of PCD inserts were preserved while those of carbide inserts were destroyed.
Consequently, the fracture on the PCD inserts was less frequent than that of the
carbide inserts. This is an important factor to reduce cutting force and temperature in
machining of Ti alloys [Hosseini, 2014].

104

Chapter 3: Driven process of thermochemical wear in Ti alloys
machining
In thermochemical wear, beside the temperature, the wear rate of a tool material is
controlled by both the solubility and the diffusivity of a tool material into a particular
phase of a work material. The dominant wear mechanism may exist under a given
machining condition to determine if a process is limited by dissolution or diffusion.
Once determined, this information will help to design/select effective tool materials or
coating materials for machining Ti alloys. According to Olortegui-Yume and Kwon
[2007], the dissolution wear must take place prior to the diffusion wear. In ferrous
machining, the dissolution of a tool material is much lower than the diffusion and the
wear process is dominated by the dissolution. This has not been corroborated in
machining Ti alloys. This section focused the thermochemical wear process involved
in machining Ti alloys in order to determine which wear mechanism dominates the
wear process.
I.1

BACKGROUND ON WEAR MECHANISMS
In general, the comprehensive wear model was considered as a combination of

mechanical wear and thermochemical wear as:

W = η ⋅Wm + ζ ⋅Wc

(3.1)

where Wm and Wc are mechanical and thermochemical wear rates and η and ζ are
the weight factors for mechanical and thermochemical wear rates, respectively. The
weight factors, η and ζ must be determined from experimental data.

105

I.1.1

Mechanical wear

The abrasion wear is one of the primary mode of mechanical wear model for flank
wear in machining Ti alloys if harder inclusions are present. However, Ti alloys do not
have hard inclusions like cementite (Fe3C) in ferrous alloys. As shown in previous
chapters, the hard inclusions in Ti alloys are the α-grains or clusters in the hard
orientation. Depending on the general morphology of the hard α-phases, they can
exhibit both 2-body and 3-body abrasion. The 2- body abrasion model presented in
Rabinowicz et al. [1961] can be expressed as:

V2mB =

F tan θ
x
π Pt

(3.2)

where

V2mB : wear volume ;

F: load between interacting surface

θ : roughness angle of abrasive
x : silding distance ;
Pt : the hardness of abraded surface (work material)
Although F and x is depended on the cutting condition, θ is influenced by the
roughness characteristics of the abrasives. They can be assumed to be the same for
given cutting condition and work material. Consequently, the relative abrasive wear
volume, RWm, can be expressed as a function of ratio of the hardness of a tool
material (2) and the reference tool material (1).

V2mB1 Pt 2
RWm = m 2 = 1
V2 B
Pt

106

(3.3)

If the hard inclusions are not constrained within the Ti matrix, an empirical
quantitative 3-body wear model abrasion introduced by Rabinowicz [1977] can be
used. In this model, hard particles can roll as well as slide between two surfaces. The
model indicates the strong dependency on the hardness ratio of abrasive and tool
material.

 L tan θ

 3Pt

 L tan θ
m
V3 B = 
 5.3Pt

 L tan θ
 2.43Pt

m

x
P
x t 
 Pa 
P
x t 
 Pa 

where V3 B is the wear volume and

if

Pt
< 0.8
Pa

if

0.8 <

if

Pt
> 1.25
Pa

−5/2

−6

Pt
< 1.25
Pa

(3.4)

Pa is hardness of the abrasives.

While the 2-body model is more relevant for large hard clusters of grains or the
hard particles with complex morphologies, the 3-body is expected to be more
relevant with small isolated grains in the regions experiencing high deformation and
recrystallizations. The α-phase has the hard orientation when the c-axis happens to
be perpendicular to the rake face. The hardness in the c-direction was reported to be
much harder (~1.5 time) than any other directions [Britton, 2009; Kwon, 2013]. To
compare tool materials at the same cutting condition, the parameters, x and θ, for the
hard clusters can be assumed to be the same. Therefore, the hardness ratio
determines the relative abrasive wear rate of a tool material. In addition, due to the
effects of thermal softening, the hardness of the tool and work materials were given

107

as a function of temperatures, so called ‘hot hardness’, which are needed to predict
the relative abrasion wear rate. The hot hardness was introduced by the exponential
function of temperature.

P = Po e −α T

(3.5)

where Po is hardness at 0°C, α is thermal softening factor, T is temperature. The
cutting tool temperature is much higher compared to the temperature of the hard
clusters as the workpiece undergoes a thermal transient condition.
Table 12: The hardness data of tool materials

Material

WC

Po
(Kg/mm2)

α x104
(/°°C)

Temperature
range(°°C)

Softening
temperature
(°°C)

References

1100

Kwon et al.
(1985)

2350

2.87

0 – 740

7260

18.11

740 - 900

pCBN

3500

7.56

0 - 1000

1500

Almond et
al. (1982)

PCD

5500

8.12

0 -1000

1500

Almond et
al. (1982)

Al2O3

2300

3.89

0 - 1000

1500

Kwon et al.
(1985)

Ti64

255

29.3

0 - 800

NA

Bilous et al.
(2005)

108

6000
PCD
WC
Ti64

Hardness(Kg/mm2)

5000

pCBN
Al2O3

4000
3000
2000
1000
0
0 °C

200 °C

400 °C

600 °C

800 °C

1000 °C

Temperature (°C)
Figure 99: The hot hardness of Ti64 and typical tool materials
I.1.2

Independent of dissolution and diffusion wear model by Kramer

The thermochemical wear includes chemical reaction, dissolution and diffusion.
The thermochemical wear occurs most likely on the rake face in the form of crater
wear. A quantitative dissolution wear model was first proposed by Kramer [1979].
The model was proved to be effective to predict the crater wear in machining ferrous
materials [Kramer, 1986; Wong, 2004].

109

Y
Vy

\Chi
Vx

Workpiece
V

p

Contact
length

X

X Tool

Figure 100: Distribution of velocity components of the chip in turning
Suh [1986] stated that thermochemical wear is a combination of dissolution and
diffusion wear:



∂c
Wc = W dif + W dis = K ⋅ S  − D + Vy 
∂y

 y =0
where :

Wc [ cm / sec] : thermochemical wear rate of the tool material
W dif
W
K

dis

: diffusion wear rate of the tool material
: dissolution wear rate of the tool material
: ratio of molar volumes of the tool material and the chip material

110

(3.6)

D cm 2 / sec  : diffusion coefficient of the slowest diffusing tool constituent in the chip
c [ at %]

: concentration of the tool material in the chip

S [ at %]

: equilibrium concentration (solubility) of the tool material in the chip

Vy [ cm / sec] : bulk velocity of chip material at the tool-chip interface
perpendiclar to the interface
The first term represents diffusion wear, which is strongly influenced by the
concentration gradient and the diffusivity of the slowest diffusing tool constituents.
The second term is dissolution wear which depends on the solubility of a tool
material. At the present time, neither dissolution nor diffusion wear model has been
applied successfully in predicting tool wear in machining Ti alloys. In particular,
Hartung and Kramer [1982] showed the dissolution/diffusion wear model did not
correctly predict the crater wear in machining Ti alloys.
I.1.2.1 Dissolution wear
In the dissolution wear model, the tool material dissolves into the chips and its
solubility influenced by the cutting temperature affects the wear rate. The dissolution
wear becomes more dominant with higher cutting temperatures. The dissolution wear
is influenced by the bulk velocity term, Vy which cannot be easily determined. For the
simplicity, Vy is considered as constant for each combination of cutting condition and
tool material. Thus, the relative wear rate of two tool materials (1, 2) can be
expressed as:

111

RW

dis

M=

W1dis M 1S1
= dis =
W2
M 2 S2

GMW

(3.7)

ρ

where M, GMW and ρ are the molar volume, the gram molecular weight and the
density of a tool material, respectively.
Assuming the tertiary compounds as a tool material, AxByCz has three
constituents, A, B and C while x, y and z are their stoichiometric ratios, respectively.
The free energy of formation per mole of a tool material can be calculated as:

∆GAx By Cz = x ∆G AM + y ∆GBM + z ∆GCM

(3.8)

where ∆GiM : relative partial molar energy of component i
The relative partial molar energy can be determined from excess Gibbs free
energy, ∆Gixs , and the solubility, Si, of component i in work material.

∆GiM = ∆Gixs + RT ln Si

(3.9)

At the equilibrium state the solubility of tool constituents is given by

S A = x ⋅ S Ax By Cz , S B = y ⋅ S Ax By Cz , SC = z ⋅ S Ax By Cz
Substitute in Eq. (3.8)

112

(3.10)

∆GAx ByCz = x ∆GAxs + y∆GBxs + z ∆GCxs + RT ( x ln x + y ln y + z ln z )
+ ( x + y + z ) RT ln S Ax By Cz

(3.11)

Then the solubility tool material AxByCz in work material is calculated:

S Ax By Cz

 ∆G Ax By Cz − x∆GAxs − y ∆GBxs − z ∆GCxs − RT ( x ln x + y ln y + z ln z ) 
= exp 
 (3.12)
( x + y + z ) RT



To determine the solubility of a tool material, AxByCz, into a particular phase (α or
β) of a work material, the values of

∆G Ax By Cz , ∆G Axs , ∆GBxs , and ∆GCxs for the

phase are needed. The free energy of formation is only dependent on the tool
material, not the phase of work material while the excess Gibbs free energy is
influenced by the phase of the work material as the solubility of a tool material was
strongly affected by the phases (α, γ) of a ferrous material [Wong, 2004 ] as well.
The free energy of formation of tool material can be expressed as a function of
temperature as below:

∆G Ax By Cz = K1 + K 2 ⋅ T log T + K 3 ⋅ T

(3.13)

where T is the temperature and K1, K2 and K3 are the curve-fit coefficients.
It is worth noting that that the solubility of tool material cannot be greater than the
solubility of tool constituents over their stoichiometric ratios.

113

SA B C ≤ SA x
 x y z
 S Ax By Cz ≤ S B y

 S Ax By Cz ≤ SC z

(3.14)

Hence, the solubility of tool material was determined by following procedure:

Given properties of tools constituents

Calculation of

by eq. (*)

Yes

No

Figure 101: Flow chart for calculation of solubility of tool material
The solubility of a tool material into a work material strongly depends on the
cutting temperature and the phase of a work material. As the cutting temperature
increases, the hardness decreases and the solubility increases, accelerating tool
wear. For example, Figure 102 presents the temperature dependence of hardness

114

and solubility of hafnium carbide (HfC) into α-grain. However, these properties of only
few tool materials in α- and β- phases of titanium were available, which were used for
the calculation in this thesis as shown in Table 13..

Figure 102: Temperature dependence of the hardness and solubility of HfC tool
Table 13: Estimated solubility of tool materials in α-Ti (at 800°C) and β-Ti (at 1000°C)
Tool
(AxByCz)

SA in α- SA in β- SB in α- SB in β- SC in α- SC in β- SA xBy Cy in SA xBy Cy in Sto ol in α- Sto ol in β Ti (at
Ti (at
Ti (at
Ti (at
Ti (at
Ti (at
α-Ti (at
Ti (at
Ti (at
β -Ti (at
800°C) 1000°C) 800°C) 1000°C) 800°C) 1000°C)
800°C)
1000°C)
800°C)
1000°C)

TiN

100.0

100.0

11.2 [1]

1.1 [1]

TiAlN

100.0

100.0

15.3 [2]

8.4 [2]

11.2 [1]

1.1 [1]

1.12E+01 1.10E+00

TiCN

100.0

100.0

0.9 [3]

0.6 [3]

11.2 [1]

1.1 [1]

9.00E-01

6.00E-01

cBN

0.1 [4]

0.3 [4]

11.2 [1]

1.1 [1]

7.69E+07 * 9.86E+07 * 1.00E-01

3.00E-01

Al2 O 3

15.3 [2]

8.4 [2]

15.3 [2]

8.4 [2]

1.04E+03 * 2.17E+03 * 5.10E+00 2.80E+00

WC

0.2 [5]

16.5 [5]

0.9 [3]

0.6 [3]

2.51E+04 * 1.07E+04 * 2.00E-01

1.02E-04

2.49E-03

Ref : [1] Murray (1987) ; [2] CA LPHA D ; [3] Bandyopadhyay (2000) ;[4] Rahaei ; [5] Murray (1981)
* solubility >100 (mol%) w as replaced by a modif ied solubility

115

1.02E-04

2.49E-03

6.00E-01

I.1.2.2 Diffusion wear
The diffusion takes place as the atom moves from the regions of high
concentration to low concentration as shown in Figure 103. The diffusion is possible
only if the contact interface reaches high enough temperature for an adequate
amount of time. In high speed turning, the condition is favorable for the diffusion of
tool material into the chip [Trent, 1991]. Besides the temperature, the diffusivity of a
tool element is affected by the particular phase present in a work material during
machining. The diffusivities into the β-phase are reported two or three orders of
magnitude higher than those into the α-phase as shown in Figure 104 for some
selected tool materials. The diffusivity data of tool constituents in the α-phase at
800°C and the β-phase at 1000°C was represented in Table 14. The diffusion of tool
constituents in the β-phase is much faster than that into the α-phase leading to a
higher diffusion wear rate with the β-phase.

Y

Vx
W

Chip at high
temperature

C

Contact length
WCXTool
Figure 103: Dissolution of carbide tool into chip

116

X

log_Difffusion ofn Ti (m2/sec)

1.E+03

Ti in α-Ti_Perlmutter (1962)

Ti in b-Ti_Perlmutter (1962)

V in α-Ti_Perlmutter (1962)

V in b-Ti_Perlmutter (1962)

Mo in α-Ti_Perlmutter (1962)

Mo in b-Ti_Perlmutter (1962)

O in α-Ti_Perlmutter (1962)

O in b-Ti_Perlmutter (1962)

Zr in α-Ti_Perlmutter (1962)

Zr in b-Ti_Perlmutter (1962)

Al in α-Ti_Perlmutter (1962) [12]

Al in b-Ti_Perlmutter (1962) [12]

1.E+00
1.E-03
1.E-06
1.E-09
1.E-12
1.E-15
1.E-18
600

800

1000

1200

1400

1600

1800

Temperature (K)
Figure 104: Diffusion of elements into α-Ti and into β-Ti.
Table 14: Diffusivity (m2/sec) of tool components in α-Ti and into β-Ti
Tool
(AxByCz)

DA in α-Ti DA in β-Ti DB in α-Ti DB in β-Ti DC in α-Ti DC in β-Ti (at
(at 800 ° C) (at 1000 ° C) (at 800 ° C) (at 1000 ° C) (at 800 ° C)
1000 ° C)

TiN

3.69E-14 [1]

1.51E-09 [1]

2.06E-15 [2]

1.18E-12 [2]

TiAlN

3.69E-14 [1]

1.51E-09 [1]

1.13E-10 [1]

1.89E-10 [1]

2.06E-15 [2]

1.18E-12 [2]

TiCN

3.69E-14 [1]

1.51E-09 [1]

6.57E-13 [3]

2.11E-12 [3]

2.06E-15 [2]

1.18E-12 [2]

cBN

1.33E-11 [4]

NA

2.06E-15 [2]

1.18E-12 [2]

Al2O3

1.13E-10 [5]

1.89E-10 [5]

5.29E-12 [1]

7.84E-04 [1]

WC

NA

9.80E-15 [6]

6.57E-13 [3]

2.11E-12 [3]

Refs. : [1] Perlmutter (1962); [2] Bellot (1996) ; [3] Bucur (1954); [4] Divinski (2008);
[5] Perlmutter (1962) [12] ; [6] Hartung (1981) [71]

More importantly, the reaction layer formed at the interface also affects the
diffusion process. As mentioned in the previous chapter, the TiC layer was most likely
formed in the adherent layer at the low cutting speed. The presence of TiC layer

117

substantially reduces the diffusion of tool constituents into the work material. Figure
105 shows the significant reduction of carbon diffusing into the TiC layer. The
adhesion layers (consisting of Ti and TiC) covered most of the contact zone on the
rake face of PCD inserts in all cutting speeds while the carbide inserts had fewer
localized areas covered by the adhesion layer. However, as discussed in pervious
section, the dominant wear mechanism with the carbide insert was dissolution and
diffusion while that for the PCD inserts is grain-pulled out. Hence, the diffusion wear
is mostly relevant to the carbide inserts.

1.0E-08

Difffusion (m^2/sec)

1.0E-10

1.0E-12

1.0E-14
C in TiCx_Barros (1998)
C in α-Ti_Barros (1998)

1.0E-16

C in Ti-6Al-4V_Barros (1998)
C in a-Ti_Bucur (1954)
C in b-Ti_Bucur (1954)

1.0E-18
600

800

1000

1200

1400

1600

1800

Temperature (°K)
Figure 105: The diffusivity of carbon in α-Ti, β-Ti, Ti64 and TiCx

The diffusion wear was influenced by the concentration gradient in y direction,

∂c
∂y

in Equation (3.6), of tool constituents in the chip, which cannot be accurately
determined. Therefore, the upper bound diffusion wear model was adopted by

118

Hartung and Kramer [1982] to simply predict the relative diffusion wear rate of tool
materials into titanium.
I.1.3

Upper bound of diffusion wear model

The upper bound calculation of diffusion wear model was introduced by Cook and
Nayak [1986] to predict the maximum diffusion wear rate with the assumption of no
chip deformation as it traverses across tool chip interface

Vwear = − K ⋅ S ( D / π t )
where:
K

0.5

(3.15)

: ratio of molar volumes of the tool material and the chip material

D cm / sec  : diffusion coefficient of the slowest diffusing tool constituent in the chip
2

S [ at %]

: equilibrium concentration of the tool material in the chip

t [sec]

: travelling time, taken for the chip to move from the edge of the tool to the
center of the crater

The accuracy of this model depends on the diffusivity and solubility of tool
constituents used in the model. As the cutting time increases, the size of crater wear
grows. Hence, the distance from the cutting edge to the center of the crater wear
increases with cutting time for a given cutting speed. Then, the respective traveling
times can be calculated. Table 15 lists the traveling times after cutting for certain
amounts of times in all cutting speeds of carbide inserts. In fact, the difference in the
travelling times among various cutting times is relatively small. In this thesis, the
average value at 91m/min was used as the travel time for each cutting speed.

119

Table 15: The distance from cutting edge to crater’s center and traveling time with
carbide inserts
Cutting speed
61 m/min

91 m/min

122 m/min

Cutting time
(min)
2.4
5.2
6.8
1.6
3.5
4.5
0.6
1.1
2.2

Distance to center
of crater (µm)
130
150
190
150
180
190
150
150
150

Time to
travel (sec)
6.40E-05
7.38E-05
9.35E-05
9.84E-05
1.18E-04
1.25E-04
1.48E-04
1.48E-04
1.48E-04

Average
time (sec)
7.71E-05

1.14E-04

1.48E-04

Table 16 shows the upper bounds of diffusion wear for a tool material used in
Hartung and Kramer (1982). The results were calculated with the diffusivity of the
slowest diffusing tool constituent, the solubility of the slowest dissolving tool
constituent and the same travelling time, t=1.14x10-4sec (see Table 15), at the cutting
speed of 91 m/min. For instance, the diffusivity of W and solubility of C were used to
calculate the diffusion wear rate of carbide. The results of predicted diffusion wear did
not agree with experiment data. For example, for turning Ti alloy with a high DOC,
carbide inserts showed that the crater wear rate was 20.6 µm/min at the cutting
speed of 91 m/min (see Figure 42), much higher than the predicted value.
Furthermore, the predicted wear rate of the carbide tool was much lower than those
of diamond tool. The experiment showed the opposite fact. It is worth stressing that
the wear mechanism (diffusion or dissolution) has not been well defined for titanium
machining. In general, the diffusion is known to be slower than the dissolution in Ti
alloys.

120

Table 16: Predicted upper bound of diffusion wear at cutting speed of 91 m/min

Tool at
300sfm

Diffusion coeff.
D(cm2/sec) of
Ratio molar
least
D
volumes (K)
REF
consti
2
(cm /sec)
tuent

Solubility C(at%) of
constituents
least
consti C (at%)
REF
tuent

wear rate
(um/min)
with t=1e-4
(sec)

Diamond

0.321

C

2.28E-06

Ha rtung
[11,12]

C

0.6

98.44

NbC

1.266

Nb

3.66E-09

Ha rtung
[11,12]

C

0.6

15.56

TiC

1.147

Ti

4.82E-09

Ha rtung
[11,12]

C

0.6

16.17

VC

1.025

V

2.92E-09

Ha rtung
[11,12]

C

0.6

11.25

ZrC

1.472

Zr

1.40E-08

Ha rtung
[11,12]

C

0.6

35.37

TiN

1.082

Ti

4.82E-09

Ha rtung
[11,12]

Ni

23.5

597.54

ZrO2

2.068

Zr

1.40E-08

Ha rtung
[11,12]

O

17

1408.03

TiB2

1.492

Ti

4.82E-09

Ha rtung
[11,12]

B

0.5

17.53

WC

1.165

W

1.05E-10

Ha rtung
[11,12]

C

0.6

2.42

Diffusivity of the slowest
diffusing tool constituent

Solubility of the slowest
dissolving tool constituent

The effects of reaction layer TiC to diffusion wear rate
In the case of the carbide insert, the diffusivity of tungsten is two or three orders of
magnitude lower than that of carbon into titanium [need reference]. Although the
diffusivity of Co is comparable to that of carbon, the amount of Co is very small in
comparison with W and C.

In addition, tungsten, which was used as the wear

121

controlling constituent in Hartung and Kramer [1982], diffuses slower than C in
titanium. However, the diffusivity data of tungsten and cobalt in α-Ti, β-Ti, Ti64 and
TiC is scarce. Thus, carbon was assumed to be the wear controlling constituent for
calculation of the upper bound diffusion wear of carbide tool to study the effect of TiC
reaction layer. Table 17 represents the upper bound of diffusion wear rate of carbide
tools in α-Ti at 1000°K, β-Ti at 1400°K, Ti64 and TiC at 1200°K for three cutting
speeds based on the diffusivity and solubility data of carbon in α-Ti, β-Ti, Ti64 and
TiC collected from few publications. The effect of the TiC reaction layer is significant
in controlling the wear rate of the carbide tool.
Table 17: Predicted diffusion wear of carbide tool in which carbon is control elements.
2

Diffusion coeff. D(cm /sec) of
constituents
least
D
constit
2
(cm /sec)
uent

REF

Solubility C(at%) of
constituents
least
consti C (at%)
tuent

REF

wear rate (µm/min)

200sfm

300sfm

400sfm

C

8.02E-09

C in α-Ti_Barros (1998)

0.6

17.44

19.87

24.13

C

1.10E-07

C in Ti64_Barros (1998)

0.6

64.63

73.63

89.43

C

1.72E-10

C in TiCx _Barros (1998)

0.6

2.55

2.91

3.53

C

1.42E-09

C in α-Ti_Buc ur (1954)

0.6

7.34

8.36

10.15

C

2.85E-06

C in β-Ti_Buc ur (1954)

0.6

328.74

374.55

454.91

The experimental results from the high DOC showed the crater wear rates of
carbide inserts at 7.2, 20.6 and 43.7 µm/min (see Figure 42) for cutting speeds of 61
m/min, 91 m/min, and 122 m/min, respectively. In comparison with the predicted
diffusion wear of the carbide insert shown in Table 17, the diffusion wear observed in
the experiments behave similar to the diffusion prediction of the carbide into the TiC

122

reaction layers at the low cutting speed, into both Ti64 and the TiC reaction layers at
the medium cutting speed, and into Ti64 at the high cutting speed as the formation of
TiC reaction layer is deterred at higher cutting speeds.
I.2

DRIVEN PROCESS OF GENERALIZED THERMOCHEMICAL WEAR
Olortegui-Yume and Kwon [2007] have introduced a generalized thermochemical

wear where the tool constituent undergoes as a sequence of dissociation, chemical
reaction if possible, dissolution, and diffusion.

Dissociation of tool material
eg AxBy tools:

AxBy = xA + yB

Chemical reaction
Chemical reaction ofdissociated species with
work material
Solubility controls
the dissolution rate

Dissolution
Atomic transport of the dissociated species that
have not been consumed in the chemical across
the tool-chip interface (work material and
reaction layer)
Diffusivity controls
the diffusion rate

Diffusion
Diffusion of the dissociated species in work
material
Figure 106: Generalized thermochemical wear [after Olortegui-Yume, 2007]

123

Assuming that the tool material, so called AxBy, dissociates into species, xA and
yB at the early stage of thermochemical wear process.

Ax B y = xA + yB

(3.16)

Then, the dissolution of tool constituents into a work material or the reaction layer
happened. The dissolution is controlled by the solubility (mol% or at%) of tool
material in work material. Therefore, at one instance of time, the specific mole of A
i

and B atoms in tool-chip interface, so called “dissolution mole limit”, ndis , i is A, B ,
are present.
Considering a volume of chip, Vdis, in the dissolution zone as shown in Figure 107,
the mole of Ti can be calculated as:

n Ti =

Vdis
ρ ⋅ M Ti
Ti

[ mol ]

Ti
3
Ti
where ρ  g / cm  : density; M [ g / mol ]: molar mass

124

(3.17)

ht=f (chip thickness)

x

hdidf
0

dc
Chemical
AxBy
reaction

lc

Figure 107: Tool constituents diffused into chip
The solubility (mol%) of tool component i (A or B) in tool-chip interface is defined
as:
i
ndis
S [ mol % ] = i
ndis + n Ti
i

(3.18)

The mole of component i dissolved in dissolution zone of chip is calculated as:

V
Si
⋅ Ti dis Ti
n =
i
1− S ρ ⋅ M
i
dis

[ mol ] → called as "solubility mole limit"

(3.19)

Concentration of tool component i (A or B) in dissolution zone is obtained as:

ni
Si
1
⋅ Ti Ti
csi mol / cm3  = dis =
i
Vdis 1− S ρ ⋅ M

→ called as "solubility concentration limit" (3.20)

The solubility concentration limit is also the fixed surface concentration at the
boundary of dissolution and diffusion zone.

125

Subsequently, the diffusion process takes place. Depending on how fast A
and B atoms diffuse in the work material, the numbers of A and B atoms at the tool஺
஻
chip interface are determined. Let, ݊ௗ௜௙௙
and ݊ௗ௜௙௙
are the moles of A and B’s

diffused into the work material in time t (sec). These numbers are functions of
diffusivity and temperature.
The net flux of atom passing through the unit area per unit time along x direction is
(Fick’s first law of diffusion):

dc( x)
J = −D
dx

 mol 
 cm2 ⋅ sec  ;

 cm2 
where D 
 is diffusivity (3.21)
sec



The Fick’s second law of diffusion:

∂c( x, t )
∂ 2 c ( x, t )
=
∂t
∂2 x

(3.22)

with a boundary condition of c (0, t ) = csi = const and initial condition of c( x, 0) = 0

Si
1
 mol / cm 3  = const ; I.C : c ( x, 0) = 0
B.C: c (0, t ) = c =
⋅ Ti
i
Ti 
1− S ρ ⋅ M
i
s

Hence, the diffusion concentration of component i (A or B) in diffusion zone at time
t and location x can be determined as:

c ( x, t ) = c 1 − erf

i

i
s

2
 

 x 
dci ( x, t )
csi
x   (3.23)
i dz d

= cs
exp − 
erf ( z ) =

 ⇒

 2 D t 
  2 Di t  
dx 3
dx dz
π Di t
1
424
i 



conc . gradient

126

where a = 2 Di t is characteristic distance for diffusion.
2

erf ( z ) =

π

z

∫e

−ζ 2

d ζ ; erfc( z ) = 1 − erfc( z ) is Gauss Error function

0

erf (0) = 1; erf (∞ ) = 1;

2 − z2
d
erf ( z ) =
e
dz
π

The diffused mole of component i passed through cross-section, Ac, at location x
of diffusion zone at time t (sec) is:

ni ( x, t ) [ mol ] = c( x, t ) ⋅ Ac ⋅ dx

; Ac = d c ⋅ lc ;

(3.24)

where lc is chip-tool contact length , dc is depth of cut , dx is thickness of a cross-section
Hence, the total mole of the tool component i diffused in chip can be calculated as:
ht
i
diff

n

(t ) = lim

hdiff → 0

∫

ht

n ( x, t )dx = Ac ⋅ c lim
i

i
s

hdiff
ht / a

= Ac ⋅ c ⋅ a lim
i
s

z = ht / a

where

∫
0

hdiff →0

∫

hdiff / a

hdiff →0

∫

hdiff


 x 
−
1
erf
   dx

 a 



 x  x
i
−
1
erf
   d = Ac ⋅ cs ⋅ a

 a  a


∫

h 

− t 
a
h
 h  1− e  
i
(t)[ mol ] = Ac ⋅ csi ⋅ a  t ⋅ erfc t  +
ndiff
π
a
a


2

erfc( z )dz

0


h
1 − e− z 
h
= t ⋅ erf  t
erfc( z )dz =  z ⋅ erfc( z ) +

π  z = h / a a
a

t
2

(3.25)

z = ht / a

h 
− t 
a

2

 1− e
+
π




 →called as "diffusion mole limit" (3.26)



The average of diffusion rate limit the tool component i in chip at time t is defined
as:

127

i
WRdiff


i
i
n
(
t
)
Ac ⋅ cs 
 ht
 mol  diff
=
=
 cm3 ⋅ sec  V ⋅ t A ⋅ f ⋅ t  ht − ht ⋅ erf  a

diff
c



h 
− t 
a

 1− e
+a
π


2



 ∝t



(3.27)

whereas the dissolution rate limit the tool component i at time t is calculated as :
i
Si
1
 mol  ndis
i
=
= cdis =
⋅ Ti
∝ t but ∝ S A B
WR  3
i

x y
1 − S ρ ⋅ M Ti
 cm ⋅ sec  Vdis
i
dis

(3.28)

The process that controls the thermochemical wear was determined by comparing
i
i
the diffusion rate limit, WRdiff
, and the dissolution rate limit, WRdis
. The process with a

smaller rate limit controls the thermochemical wear of tool constituents. Figures 108 111 compared the dissolution and diffusion wear rates of the constituents of WC, TiN
and cBN tool material into α-phase and β-phase using the data of solubility of tool
material, S A B C , in Table 13 and diffusivity tool constituents in Table 14. At the very
x y z

beginning stage (less than 1E-04 sec), the thermochemical wear is dominated by the
dissolution mechanism. Then, for most tool components such as W, C, Ti, B (exclude
N into β-phase at 1000°C), diffusion will control the thermochemical wear rate after
that. To summarize, traffic jam analogy was used and demonstrated the difference in
the thermochemical wear between ferrous and titanium alloys with the carbide tool on
Figure 112. The dissolution controls the thermochemical wear in machining ferrous
while thermochemical wear of mostly tool material in machining Ti will be restrained
by the diffusion the fastest tool constituent. Thermochemical wear of carbide into β-Ti
is an example of this. After dissociating from carbide tool, W and C dissolved and
diffused independently into chip in which diffusion is driven process for both W and C

128

as shown in Figure 108. The diffusion of W and C is influenced by solubility of
carbide which controls the dissociation of tool at tool-chip interface as the way how
solubility of tool was calculated. To put it more simply, diffusion of W and C control
both the dissociation of carbide and their dissolution. Therefore, thermochemical
wear of carbide depends on how fast W or C diffused in the chip. This means that C
with higher diffusion rate (see Figure 108) will control thermochemical wear of
carbide tool. The slower diffusion rate, a part of W will diffuse in chip while the other
left will be taken away with the chip flow. The relative thermochemical wear of tool 1
and tool 2 was determined by the chart shown in Figure 113.

129

x 10

-14

The average wear rate of Tungsten (W ) in the chip at 1220K

x 10
8
7

Due to t ≅ 0

6
5

W component into
β-Ti at 1000°C

1.5

4
1

Diffusion driven

3

diffused atoms (mol)
0.5

0

2

diffusion wear rate

0

0.1

0.2

0.3

0.4 Time0.5
t(sec) 0.6

0.7

0.8

0.9

1

1
x 10

2.5

x 10

-13

The average wear rate of Carbon (C) in the chip at 1220K

x 10
8

-4

7
6
5

Diffusion driven

1.5

0

-3

C component into
β-Ti at 1000°C

2

4
1

3
diffused atoms (mol)

0.5

2

diffusion wear rate
1

diss olution wear rate
0

Wear Rate(mol/[cc*sec])

2

dissolution wear rate

Total mole of diffused atoms in the chip(mol)

-5

Wear Rate(mol/[cc*sec])

Total mole of diffused atoms in the chip(mol)

2.5

0

0.1

0.2

0.3

0.4 Time0.5
t(sec) 0.6

0.7

0.8

0.9

1
x 10

0

-3

Figure 108 Dissolution and diffusion wear rate of carbide tool (WC) into β-Ti at
1000°C

130

7

The average wear rate of Titanium (T) in the chip at 800C

x 10
3.5

Ti component into
α–Ti at 800°C

3

6

2.5

Diffusion driven

5

2
4
1.5
3
1

diffused atoms (mol)

2

diffusion wear rate
0.5

1
0

dissolution wear rate

0

0.1

0.2

0.3

0.4 Time0.5
t(sec) 0.6

0.7

0.8

0.9

1
x 10

2.5

Total mole of diffused atoms in the chip(mol)

-8

Wear Rate(mol/[cc*sec])

-18

x 10

-18

-3

The average wear rate of Nitrogen (N) in the chip at 800C

x 10
7

N component into
α–Ti at 800°C

2

0

6

5

Diffusion driven
1.5

4

3

1

2

diffused atoms (m ol)
0.5

Wear Rate(mol/[cc*sec])

Total mole of diffused atoms in the chip(mol)

8

x 10

diffusion wear rate
1
dissolution wear rate

0

0

0.1

0.2

0.3

0.4 Time0.5
t(sec) 0.6

0.7

0.8

0.9

1
x 10

0

-3

Figure 109: Dissolution and diffusion wear rate of TiN tool into α-Ti at 800°C

131

-9

x 10

-15

The average wear rate of Nitrogen (N) in the chip at 1000C

x 10
4
3.5

Ti component into
β-Ti at 1000°C

1

0.8

Diffusion driven

2.5

0.6

2
1.5

0.4
diffused atoms (mol)

1

diffusion wear rate

0.2

0

0

0.1

0.2

0.3

0.4 Time0.5
t(sec) 0.6

0.7

0.8

0.9

0.5

1
x 10

x 10

3.5

-14

0

-3

The average wear rate of Titanium (T) in the chip at 1000C

x 10
1.4

N component into
β-Ti at 1000°C

1

Dissolusion driven

2.5

0.8
2
0.6
1.5
0.4

diffused atoms (mol)

1

diffusion wear rate
0.2

0.5

dissolution wear rate

0

0.1

0.2

0.3

-4

1.2

3

0

Wear Rate(mol/[cc*sec])

3

dissolution wear rate

Total mole of diffused atoms in the chip(mol)

-6

0.4 Time0.5
t(sec) 0.6

0.7

0.8

0.9

1
x 10

0

-3

Figure 110: Dissolution and diffusion wear rate of TiN tool into β-Ti at 1000°C

132

Wear Rate(mol/[cc*sec])

Total mole of diffused atoms in the chip(mol)

1.2

x 10

-13

1.4

The average wear rate of Boron (B) in the chip at 800C

x 10
6

B component into
α–Ti at 800°C

5

Diffusion driven

4

0.8

3

0.6
2
diffused atoms (mol)

0.4

diffusion wear rate
0.2

1

dissolution wear rate

0

0.1

0.2

0.3

0.4 Time0.5
t(sec) 0.6

0.7

0.8

0.9

1
x 10

2.5

Wear Rate(mol/[cc*sec])

1

0

Total mole of diffused atoms in the chip(mol)

-4

1.2

x 10

-15

-3

The average wear rate of Nitrogen (N) in the chip at 800C

x 10
7

N component into
α–Ti at 800°C

2

0

6

5

Diffusion driven

1.5

4

3

1

2

diffused atoms (mol)
0.5

Wear Rate(mol/[cc*sec])

Total mole of diffused atoms in the chip(mol)

1.6

diffusion wear rate
1
dissolution wear rate

0

0

0.1

0.2

0.3

0.4 Time0.5
t(sec) 0.6

0.7

0.8

0.9

1
x 10

0

-3

Figure 111: Dissolution and diffusion wear rate of cBN tool into α-Ti at 800°C

133

-6

Traffic Jam Analogy: Fe vs. Ti
Ti
Diffusion

Fe Diffusion

Reaction
Dissolution

Dissolution
Dissociation
WC-Co

Dissociation
WC-Co

a) Dissolution controls in cutting
steels

b) Diffusion controls in cutting
titanium

Figure 112: Process (dissolution and diffusion) controls thermochemical wear for
machining of ferrous and Titanium with carbide tool.

Tool 1:
AxBy

Tool 2:
CzDw

No

Dissolution
driven

Dissolution
driven

Diffusion
driven ?
Yes

No

Diffusion
driven ?

Diffusion
driven

Diffusion Yes
driven

Find the fastest
diffusing component
(for example: A)

Find the fastest
diffusing component
(for example D)

Figure 113: Flow chart for calculation of relative thermochemical wear rate.

134

According to the pervious chapters, the thermochemical (dissolution/diffusion) is
proven to be the dominant wear mechanism of carbide tools in machining Ti64. The
diffusion mechanism takes place by the diffusion action of W and C or Co from carbides
into Ti alloys. However, the diffusion process is hampered due to the presence of TiC
layer that forms by the reaction of Ti and C from carbides. The diffusivity of W and C in
TiC is fairly lower than in titanium and Ti64. The reaction layer TiC formed at tool-chip
interface helps to reduce wear rate. This can be proved with the nano-lubricants, which
can provide the additional carbon from graphitic platelets. However, providing the nanolubricants during turning is not always successful with MQL as the cutting tools is
always engaged with the workpiece. Instead, the nano-graphitic platelets can provide
additional lubricity at the tool-chip interface.
In next chapters, the study of tool wear improvement was conducted with MQL
machining with micro and nano-platelets of graphite added in MQL oil. However, due to
the difficulty in setting up MQL turning experiment, the ball mill was selected. The AISI
1045 steel was initially chosen for the workpiece to reduce the cost of the experiments.
Based on the MQL ball-mill experiments with the steel, the investigation was carried out
to find the effect of diameter and thickness of platelets and presented in in Chapter 4.
Finally, the performance of MQL experiment with nano graphite platelets was evaluated
with Ti64 in Chapter 5.

135

Chapter 4:

Tool wear improvement in machining of Steel AISI

1045 with micro and nano-platelets enhanced MQL
IV.1 INTRODUCTION
Flood cooling has been the backbone of machining processes for manufacturing
industries. The traditional flood cooling, which uses and discards a large amount of
metal working fluid, is widely used in industrial machining operations. The primary
purposes of metalworking fluid (MWF) are to rapidly dissipate heat, reduce friction
between the tool and the chip and carry the chip away from the cutting zone.
Consequently, the reduction in force and power consumption and the improvement in
tool life and surface finish [1] are apparent. However, flood cooling has many
disadvantages due to its associated maintenance and disposal cost (7-17%) [Kajdas,
2010] as well as occupational hazards and environmental impacts. According to the
U.S. Occupational Safety and Health Administration (OSHA) [Aronson, 1995] and the
U.S. National Institute for Occupational Safety and Health (NIOSH) [U.S. DHHS, 1998],
the permissible exposure level (PEL) for metalworking fluid aerosol concentration is 5
mg/m3 and 0.5 mg/m3, respectively. The coolant mist level in U.S. automotive parts
manufacturing facilities has been estimated to be on the order of 20-90 mg/m3 with the
use of traditional flood cooling and lubrication [Bennett & Bennett, 1985]. As a result, the
need to reduce cutting fluid consumption is strong and the discrepancy in fluid aerosol
concentration needs to be aggressively addressed. Dry machining has offered an
alternative solution by completely eliminating cutting fluid. However, its applications are
limited due to the poor surface quality, the accelerated tool wear and the high

136

concentration of airborne particles on the factory shop floor [R Khettabi, 2008; 2009].
Recently, Minimum Quantity Lubrication (MQL)-based machining has been introduced
as an ideal alternative which provides a compromise between productivity and
environmental concerns. MQL machining refers to a class of machining processes
where the lubricant use is limited to only a extremely small amount – typically a flow rate
of 2.5 to 500 ml/h [Gaitonde, 2008; Shen, 2008]. This amount of lubricant is about three
to four orders of magnitude lower than the amount commonly used in flood cooling
which dispenses up to 60 l/h [Srikant, 2009]. Thus, MQL offers many benefits such as
the diminution of the environmental and occupational hazards and the reduction of
maintenance and disposal cost. A general study on the efficiency of MQL to dry
machining was introduced in Weinert et al. [2004]. In mild machining conditions, MQL
provides a great substitute for flood cooling. However, MQL-based machining
possesses have few major drawbacks such as chip disposal and limitation in machining
conditions.

Chip disposal can be somewhat remedied by adding compressed air.

However, the lack of cooling capacity with MQL-based machining has been a major
drawback preventing this technology to reach its full potential in industrial applications.
To improve MQL performance and expand the range of MQL applications, many spray
configurations and cutting parameters are proposed and investigated while others study
to enhance the lubricity.
IV.1.1 Improved performance of MQL with alternative lubricants
Several works have been published to evaluate the characteristics of various
lubricants for finding optimum lubricants in MQL machining. The most common

137

lubricants in MQL are mineral oil [Braga, 2002; Brandao, 2011; Brinksmeier, 1999;
Shen, 2008; Varadarajan, 2002], emulsions (soluble or synthetic fluid) [Brinksmeier,
1999; Gaitonde, 2008; Wakabayashi, 2006], synthetic oils, esters [Brinksmeier, 1999;
Wakabayashi, 2006], neat oil [Sarhan, 2012; Thepsonthi, 2009] and vegetable oil [Cetin,
2011; Ozcelik, 2011; Shen, 2008; Singh, 2013; Wakabayashi, 2006]. Some researchers
used alcohol [Heinemann, 2006; Shen, 2008] and deionized water [Mao, 2012; Shen,
2008]. The mixture of water and oil-free synthetic lubricant provided longer tool life
compared to both synthetic ester with and without alcohol in deep-hole drilling due to its
high cooling ability and low viscosity of water [Heinemann et al., 2006]. A new approach
to enhance the cooling effect with the MQL process, oil-on-water droplets produced
through a new design nozzle system, reduced cutting forces in end milling [Itoigawa,
2006; Yoshimura, 2005]. Cutting performance in the tapping test as well as secondary
factors (biodegradability, oxidation, and storage stability) for synthetic polyor esters
were superior to vegetable oil [Wakabayashi, 2006; Suda, 2002].
In general, vegetable oil is favorable in MQL machining due to reduction in cutting
force [Sadeghi, 2008], better surface finish [Emami, 2014; Singh, 2013; Tai, 2011] and
improvement in tool life [Nguyen, 2012; Park, 2011] in comparison with MQL with other
lubricants. More importantly, the biodegradability makes it safe for the workers in the
environment. Thus, MQL offers many benefits such as the diminution of the
environmental and occupational hazards and the reduction of maintenance and disposal
costs. However, MQL lacks the ample cooling offered by flood cooling. Due to the
extremely small amount of a liquid lubricant, MQL mainly provides essential lubricant
function, which makes effective to only mild cutting conditions.

138

IV.1.2 Improved performance of MQL by adding solid lubricants
Solid lubricants have been used either to eliminate the use of cutting fluids or to mix
with cutting fluid to enhance its lubricity. Many of them have lamellar or layered crystal
structures where each layer readily slides against adjacent layers to provide lubricity.
Some examples of such solids include soft metals, graphite, hexagonal boron nitride,
boric acid, and the transition-metals dichalcogenides MX2 (M is molybdenum, tungsten
or

niobium

and

X

is

sulfur,

selenium

or

tellurium),

such

as

MoS2 and

monochalcogenides (e.g., GaSe and GaS) [Erdemir, 2001]. However, no single solid
lubricant can provide the lubricity over a wide range of cutting conditions (different tool
and work materials). For example, molybdenum sulfide (MoS2) works well in vacuum or
dry conditions but degrades quickly in moist and oxidizing environments [Wiener, 1967].
Most transition metal dichalcogenides tend to oxidize at elevated temperatures, and
thus lose their lubricity. MoS2 can provide lubrication up to 350°C while WS2 endures up
to 500°C [Sliney, 1982]. In general, those with higher oxidation resistance or
chemical/structural stability perform the best at elevated temperatures. Oxide and
fluoride-based solid lubricants (e.g., CaF2, BaF2, PbO, and B2O3) [Sliney, 1993] as well
as some soft metals (e.g., Ag, Au) function quite well at elevated temperatures [Erdemir,
1990; Erdemir and Erck, 1996; Maillat, 1993), but all fail to provide low friction at room
or lower ambient temperatures. The lubricity of these solids at elevated temperatures is
largely controlled by their ability to soften and resist oxidation. Solid lubricants such as
molybdenum disulfide (MoS2), tungsten disulfide (WS2) graphite, hexagonal boron
nitride (hBN) and polytetrafluoroethylene (PTFE) have been used as dry powders or
coating materials [Donnet, 2004]. The effectiveness of the solid lubricant was

139

demonstrated in some publications. For example, the recent developments and future
trends of well-known solid lubricants were summarized at Donnet et al. [2004]. Yukhno
et al. [2001] reviewed the frictional behavior and wear resistance of solid lubricant
coatings at low temperatures. Prasad et al. [2010, 2005] inspected the effects of some
solid lubricants such as graphite, talc, MoS2, and lead suspended in oil on the wear
performance of cast iron and zinc–aluminum alloy. Hsu et al. [2004] discussed the
important characteristics of nano-lubricants such as being non-volatile, oxidation and
thermal decomposition resistant, and self-repairing as well as having a more effective
film organization of nano-lubricant compared to conventional lubrications. The effect of
the concentration of solid lubricants (MoS2 and graphite) were studied and the “optimal
concentrations” were found to be 3wt.% and 10wt.% for MoS2 and graphite, respectively
[Bartz, 1971]. Below the critical concentrations of MoS2 and graphite, they did not
effectively protect the surfaces against wear.
The main advantage of solid lubricants is the preservation of lubricity even under
extreme pressure and temperature. Thus, by mixing the solid lubricants into a MQL
lubricant, MQL-based machining can be enhanced for more aggressive conditions.
Several works were published to evaluate the effectiveness of MQL applications at
elevated temperatures by adding small amounts of solid lubricant to MQL oils.
However, to be effective in general machining applications, solid lubricants should
provide lubricity in a broader range of temperatures.
Molybdenum sulfide (MoS2)

140

Shen et al. [2008] used MoS2 to enhance grinding processes. By mixing MoS2 nanoparticles into MQL grinding fluid, the MQL-grinding process showed excellent
performance on cutting forces. However, its effectiveness, especially in traditional
machining applications, is questionable due to the low dissociation temperature of MoS2
(350°C in oxidizing environments). Kalita et al. [2012] added 2 and 8 wt.% of MoS2
nano-particles (40-70 µm) and micro-particles (3–5 µm) into paraffin and soybean oils
used in MQL grinding tests of cast iron and EN 24 steel. He concluded that the grinding
process was improved with MQL oil with added MoS2 (both nano- and micro- particles)
to flood cooling and MQL with pure oil. Furthermore, the nanoparticles less than 100nm
exhibited superior performance to micro-particles (3-5 µm), resulting in up to 30% and
50% reduction in the friction coefficient and grinding ratio, respectively. The increase in
the nano-particle content reduced the friction, energy consumption and cutting forces. In
milling tests of AL6061-T6 alloy, Rahmati [2013] used the MQL mixture with MoS2
nanoparticles (20–60 nm) added in ECOCUT HSG 905S neat cutting oil at various
concentrations. The machining experiment with the presence of MoS2 nanoparticles in
the MQL oil showed outstanding performance compared to pure MQL oil in term of
cutting force, temperature and surface finish. He claimed that optimum concentrations
for minimizing cutting force and temperature were 1 wt.% and 0.5 wt.%, respectively.
The surface roughness was minimized by applying the MQL oil mixed with 0.5 wt.% of
MoS2 nanoparticles.

141

Graphite (C)
The main advantage of graphite over MoS2 is that the dissociation of graphite is
much higher than that of MoS2. The graphite was found to start decomposing at
moderately high temperatures (e.g.: 500°C in oxidizing environments [Zhu, 1998]). The
nano-graphene platelet enhanced lubricants were studied under various conditions. Lee
et al. [2009] studied nano lubricants by mixing graphite nanoparticle additives (at 0.1
vol% and 0.5 vol%) with industrial gear oil to improve the lubrication properties in the
disk-on-disk tribo-test between two sliding grey cast iron plates which showed superior
performance in reducing the friction coefficient, sliding temperature, and surface
roughness. Huang et al. [2006] compared the performance of pure paraffin oil to
mixtures of paraffin oil with flake graphite (particle size: 48 µm) and with graphite nanosheets (average diameter: 500nm, thickness: 10-20nm) in pin-on-disk tests using steel
balls. He pointed out that the mixture with graphite nano-sheets formed a film on the
rubbing surface that not only improved the friction coefficient but also reduced the wear.
The pure graphite powder was used as solid lubricant to prolong the tool life and
improve the cutting force and surface quality in milling AISI 1045 [Reddy, 2006]. Amrita
et al. [2013] studied MQL turning of AISI 1040 steel using varying concentrations of
graphite nano-particles (particle size < 100nm) suspended in water soluble oil (20:1).
The cutting temperature, flanks wear and cutting force were significantly reduced
compared to those of dry, flood and pure MQL oil machining. Among the concentrations
of 0.1%, 0.3%, and 0.5 wt%, the higher content showed more improvement in the
cutting performance. In turning of hardened AISI 4340 steel with a MQL mixture of 10%

142

graphite added in semi-solid lubricant (composed of calcium and a sodium soap base
emulsified with mineral oil) [Paul, 2013], the machining process presented better cutting
performance (e.g. tool wear, cutting force and temperature, surface finish) compared to
dry machining and turning with pure MQL. In the MQL grinding process of a hardened
tool steel, Alberts et al. [2009] found that the mixture of exfoliated Graphite nanoPlatelet (xGnP) at 1wt% dispersed in isopropyl alcohol (IPA) was the optimal
concentration to reduce cutting forces, specific energy and surface roughness. Park et
al. [2011] showed that by using a very small amount (0.1wt%) of exfoliated graphite
nano-platelets (xGnP) added in MQL Unist Coolube® 2210 vegetable oil, the tool wear
was reduced by up to 50% in MQL ball milling of 1045 steel.
Boric acid (H3BO3)
Boric acid is a popular alternative solid lubricant added to MQL oils which have been
evaluated in a few publications. Rao et al. [2008] investigated turning of EN 28 steel
using compressed air and solid lubricant powders (graphite and boric acid) at different
particle sizes of 50, 100, 150 and 200µm . The reduction of cutting force and flank wear
was found when using solid lubricants compared to dry and wet conditions. He claimed
that a particle size of 50µm would be most effective for both graphite and boric acid.
Krishna [2012] examined the performance of boric acid nano-particle (100um)
suspensions at various concentrations (0.25, 0.5, 0.75, and 1.0 wt%) in coconut oil with
MQL turning AISI 1040 steel. In comparison to dry conditions, the turning process with
MQL oil showed significant reduction in cutting temperature, surface finish and tool flank
wear due to slight improvements in thermal conductivity and the heat transfer

143

coefficient. Among them, the 0.5wt% mixture showed the best cutting performance.
Using the same approach, Ramana [2011] studied MQL mixtures of boric acid particles
(50 nm, 60nm, 80nm and 0.5um) in canola oil in turning tests of AISI 1040 hardened
steel. Surprisingly, the experimental results revealed that MQL oil with boric acid
showed higher cutting force and tool temperature compared to dry and flood cooling
with pure canola oil. Furthermore, the smaller particles showed worse performance in
terms of cutting force, tool wear and surface quality of the machined surface.
IV.1.3 Improved performance of MQL by varying spray configuration
The effectiveness of MQL is directly dependent on the application angle of the
nozzle. Figure 114 depicts the illustration of spray angles (yaw and pitch angle) of an
external MQL nozzle in a ball-milling experiment. The experiment is set for the rotating
mill to travel along the Y axis. In this setup, the pitch angle is the angle between the Y
axis and nozzle spray while yaw angle is the angle between the table feed direction and
nozzle spray. Table 18 summarizes the findings from other works in determining the
optimal spray angles and other spray parameters for external MQL systems with
respect to the definitions of yaw and pitch angle in Figure 114.

Figure 114: Pitch and Yaw angle of the nozzle in End-ball milling

144

Table 18: Publication on optimal spray conditions in MQL machining with definition of
Yaw and Pitch angle as shown in Figure 114
Ref.

Machining
Process

Nozzle
distance

Flow

Yaw
angle

Pitch
angle

Improvement

Mao et Grinding of AISI Deionized 0.75wt%
52100
al. [2012]
water Al2O3particl
es (60nm)

20mm

60 ml/h

180°

15°

50% in surface roughness,
20% in grinding temperature

Yan
al. []

--

20mm

43.8 ml/h

60°

60°

10% in Flank wear and
surface roughness to other
yaw and pitch angles

Lacalle
Milling of
Biodegra
et
al. Aluminum 5083- dable oil
[2006]
H112

--

--

0.06ml/min

45°

--

30% in flank wear to flood
coolant with emulsion 95% of
water

Liu et al. Milling Ti6Al4V Vegetable
[2011]
oil

--

25mm

10 ml/h

45°

--

Little lower cutting force and
temperature to others

Ueda et
al. [2006]

Vegetable
oil

--

40 ml/h

45°

45°

10% in temperature

Syntilo
XPS
Castrol
5%

--

40

100 ml/h

180°

10-12°

Slightly improvement in
grinding force and surface
finish to other positions and
depend on material of wheel

Vegetable
oil

xGnP

50-70
mm

1 ml/min

--

10°

40-50% in Flank
improvements of xGnP
mixture in comparison to
pure oil.

et Milling of forged
steel

Turning and
Milling of AISI
1045

Tawakoli
Grinding of
et
al. hardened steel
100Cr6
[2010]
Park et
al. [2011]

Milling of AISI
1045 steel

Cutting
fluid

Esters

Additive

rate

IV.2 BACKGROUND
Lamellar-type solid lubricants are readily available in the form of platelets. The
diameter and thickness of these platelets are typically to the scale of tens of microns
and only a few microns, respectively, which are classified in this thesis as microplatelets. Some of these platelets are also available as nano-platelets whose thickness
is well below a micron (even to a few nanometers). Clearly, the micro-platelets are not
as inexpensive. However, the mixtures with micro-platelets are not as stable as those
with nano-platelets. Another aspect of nano-particles is the worldwide excitement of
nanotechnology research. Many research findings are reporting on new processing

145

techniques, which will eventually make them cheaper and available for wide varieties of
applications. In this study, these micro- & nano-particles and platelets are used to
reduce friction and/or wear. It is also important to recognize that a wide variety of shape
and size distributions of these particles are now available.
In this chapter, we will explore the use of micro and nano platelets of graphite and
hBN mixed in the vegetable oil, Unist Coolube 2210 obtained from Unist, Inc. (Grand
Rapid, Michigan) in tribometer tests and MQL ball-milling tests. This study defines
‘nano-platelets’ to have a thickness of less than 0.1µm and ‘micro-platelets’ to have a
thickness of more than 0.1µm. The hBN phase is similar to graphite but quite different
from its related compound, cubic boron nitride (cBN). Like graphite, the hBN phase has
a layered structure where each layer is weakly bonded with adjacent layers. With the
application of a shear load, each layer can easily slide against other layers providing
lubricity. Both graphite and hBN can provide the lubricity desperately needed in more
severe machining conditions, even when the oil droplets can dissociate. The
dissociation temperature of hBN is even higher than that of the graphite. The main
advantage of the nano-platelet form of solid lubricants (hBN and graphite) is the aspect
ratio of diameter to thickness. When applied, the larger surface of the nano-platelets
will land on the tool surfaces and additional lubricity from these nano-platelets can be
provided in the MQL-machining. The nano-platelets have the thickness in nano-scale
and the diameter in micro-scale, which offers a unique advantage of being filtered by
nose hair and not being absorbed through human skin. This thesis intends to find the
effect of thickness and diameter on these platelets in MQL machining of AISI 1045 steel
and Ti64.

146

IV.2.1 Micro and nano-platelet characterizations
In the periodic table of elements, boron and nitrogen are the neighbors of carbon (C).
Boron Nitride (BN) was first founded in early 19th century. The compound has the same
number of outer shell electrons as graphite and diamond. BN has very similar properties
as carbon with both hexagonal and cubic crystal structures [Haubner, 2002]. Figure 115
illustrates the similarity in the hexagonal structure of graphite and hBN. While both
forms of carbon exist as graphite and diamond in nature, both forms of BN are synthetic
and exist as cubic boron nitride (cBN) and hexagonal boron nitride (hBN). The hBN
phase is more similar to graphite but quite different from its related compound, cubic
boron nitride (cBN). The electrical properties are recognized as the prominent
differences between carbon and BN. Due to its high electrical resistivity, BN is wellknown as dielectric and thermally conductive whereas carbon is electrically and
thermally conductive. The main difference in appearance is that hBN is white while
graphite is black. Like graphite, the hBN phase has a layered structure where each
layer is weakly bonded with adjacent layers. With the application of a shear load, each
layer can easily slide against other layers providing lubricity. hBN is expected to have
the similar lubricant properties to graphite. The main advantage of the mixtures of
graphite/oil and hBN/oil is that they can provide lubricity even if the oil droplets
dissociate due to a high cutting temperature. The hBN is stable up to 1000°C [Haubner,
2002], while graphite was known to start decomposing at moderately high temperatures
(500°C) [Zhu, 1998]. The hBN has good chemical inertness and is not wetted by most
molten metals [Haubner, 2002]. These advantages of hBN compared to graphite make

147

hBN a more promising solid lubricant in MQL processes where the cutting temperature
is expected to be very high.
Several research investigations on the lubricity of BN for various applications were
published. The early research on tribological properties of cubic, amorphous and
hexagonal boron nitride films by Watanabe et al. [1991] showed that BN films helped to
decrease the friction force and the friction coefficient as a function of normal load. The
curious behavior of hBN was discovered with a ring-on-roller tribometer when added to
lubricating oil as well as the effect of hBN concentration on friction and wear in Kimura
et al. [1999]. The sliding experiment of bearing steel vs. itself showed that the addition
of BN at as little as 1 wt% resulted in the reduction of wear. The wear was reduced as
concentrations of BN increased, although the friction coefficient was slightly increased.
Based on these investigations, BN was considered as a potential additive for the
lubricants.

Figure 115: An illustration of hexagonal crystalline structures of ghraphite and hBN
[Encyclopedia Britannica, Inc., 1995]

148

Table 19: Properties of graphite and hBN
Formula

BN

Graphite (C)

Molecular mass

24.82 g/mol

12.01 g/mol

Density (hexagonal)

2.27 g/cm3

2.23g/cm3

Density (cubic)

3.48 g/cm3

3.51 g/cm3

2700-3000°C

3652 - 3697°C

1-2

1-2

Youngs Modulus (MPa)

20-102

8-15

Coefficient of Friction

0.15 to 0.70

0.2 to 0.6

34

25-470

1000°C

500°C

Melting range
Mohs

Thermal

hardness

conductivity

Temperature

Stability

Figure 116: Temperature range for lubrication of different solid lubricants [Chen N.,
2004]
The solid lubricant platelets examined in this study include micro-platelets (graphite
and hBN) and nano-platelets (xGnP and exfoliated hBN). The graphite micro-platelets
were obtained from Alfa Aesar Com., USA while the micro-platelets of hBN5 were
obtained from Changsung Corp. (South Korea). Four nano-platelets designated as

149

xGnP M5, xGnP M15, xGnP C300, xGnP C750 and one exfoliated hBN300 were
provided by XG Science, Inc., (Lansing, Michigan, USA) using their cost-effective
exfoliation process after being downsized by the pulverization process [Fukushima,
2003]. Among xGnP platelets, the Grade M5 and M15 have the same thickness with
distinct diameters while C300 and C750 have the same diameter with distinct
thicknesses. Table 20 summarizes the diameters and thicknesses of each platelet.
Some of these data have been interpreted based on surface area measurements or
based on SEM images. As presented in Table 20, a typical, commercially available hBN
powder has a surface area of around 2 m2/g while the surface area of XGS hBN is
246 m2/g. The surface area of hBN is quite larger than that of xGnP grade M5 (~120150 m2/g) but is smaller than that of xGnP grade C (300 and 750 m2/g for C300 and
C750, respectively).
Table 20: The diameters and thicknesses of various platelets

Nano-platelets

Microplatelets

Platelets

Diameter Thickness Aspect SBET(m2/g)
ratio

Company

Graphite

7µm

200 nm+

35

4.8**

Alfa Aesar

hBN5

5µm

1-2µm+

3

1.27**

Changsung

xGnP M5

5µm

6-8 nm

714

~120-150

XG Sciences

xGnP M15

15µm

6-8 nm

2143

~120-150

XG Sciences

xGnP C300

2µm

3 nm*

667

300

XG Sciences

xGnP C750

2µm

1.2 nm*

1667

750

XG Sciences

hBN300

11 nm*

8.24 nm

1

246

XG Sciences

+ A estimation based on SEM image (See Figures 5 and 6)
* A calculated value based on the surface area BET (Brunauer–Emmett–Teller theory) data
** The calculated surface area based on diameter and thickness

150

The patent by Kwon and Drazl [2010] has been issued using xGnP platelets to
enhance the effectiveness of MQL machining.

A wide variety of graphite platelets

varying in diameter and thickness are now produced by XG Science, Inc.. The main
advantage of xGnP over MoS2 is that the dissociation of graphite is much higher than
that of MoS2. In the paper by Park et al. [2011], the oil and the nano-platelets have
been mixed in a high-speed mixer to produce stable solutions, which have been
substituted for vegetable MQL oil in the MQL-based ball-milling experiments. The
diameter and thickness of the graphite platelets were approximately 1µm and 10nm in
average, respectively. The results were very encouraging in the reduction of tool wear.
Figure 5 presents the micrographs of some of the platelets at 1000x magnification.
Figure 6 shows the thickness of xGnP and graphite micro-platelets at 30,000x
magnification. Although both platelets have a similar particle diameter, the thicknesses
of xGnP and hBN are much smaller than those of micro-platelets. We expect that the
nano-platelets can facilitate a better lubricity and reduce wear even at the same
concentration levels in oil. Figure 119 presents the diameter versus thickness and the
aspect ratio for graphite platelets.

151

a) micro-platelet graphite

b) nano-platelet xGnP (M5)

c) micro-platelet hBN5
d) nano-platelet hBN300
Figure 117: SEM analysis

a) xGnP M5

b) micro-platelet graphite

c) hBN5

Figure 118: SEM images micro- and nano-platelets

152

Diameter vs. thickness for graphite

2500

Graphite 2-15um

2000

12

Graphite

Aspect ratio

Diameter (µm)

16

xGnP M5
xGnPM15

8

Aspect ratio for graphite

xGnP C300
xGnP C750

1500
1000
500

4
0
0
-20

80
180
Thickness (nm)

280

Figure 119: Diameter, thickness and aspect ratio of graphite platelets.
IV.2.2 Vegetable oil Unist Coolube 2210
The vegetable oil, Unist Coolube® 2210, has the flash point ~200° C from Unist, Inc.,
(Grand Rapid, Michigan, US) and was selected as the lubricant It is not hazardous to
the workers. More importantly, it is formulated from renewable plant-based oils, which is
biodegradable, making it friendly to the environment. With very small amounts of oil
used in MQL machining, its main purpose is a “lubricant”, not a “coolant”. As a lubricant,
it reduces friction between tool and work material, thus decreasing heat generation.
Unlike flood cooling, it does not provide cooling. As shown in Park et al. [2011], the
Unist oil showed the smallest wetting angle among various lubricants tested. The
lubricant with the smallest wetting angle will spread out on the machined surface easily,
covering a larger area by a thin, low friction lubricant between the cutting tool and work
piece. Henceforth, wetting angle was measured on various mixtures of graphite, xGnP
and hBN platelets with vegetable oil. The mixtures were tested in a pin-on-disc type
tribometer with reciprocating motion to study the performance of lubricity and wear

153

resistance. The ball-mill experiment is chosen to evaluate the applicability of the nanoplatelet enhanced MQL lubricant as the cutting edges of the tools are exposed each
revolution.
IV.3 EXPERIMENTAL SETUP AND PROCEDURES
IV.3.1 Suspension Stability of Mixtures
In practice, the stability of MQL mixtures – the segregation of particulates in a mixture
- is important. With the unstable mixture, the addition of solid lubricants will not provide
the intended function. The suspension stability is influenced by diameter, thickness of
platelets and mixing conditions such as mixing speed and mixing time. The suspension
stability test was conducted with the mixtures made with both micro- and nano-platelets
of graphite and hBN with vegetable oil. The mixtures were prepared on a high-speed
mixer DAC 150FVZ-K from FlackTek (Landrum, SC) as shown in Figure 120. The
mixing conditions were set at a speed of 2300rpm and durations of 4, 6 and 8 minutes
as the stabilities of the mixtures were achieved.

Figure 120: High-speed mixer DAC 150FVZ-K

154

IV.3.2 Wetting angle measurement
The wetting angle is one of the significant parameters determining lubricity of the
MQL lubricant. As the MQL droplets are applied to the spinning tool, the wetting angle
indicates how well the droplets will stick to the tool surface. A lubricant with a small
wetting angle shows good wettability on the surface. The wetting angle tests were
conducted with both micro- and nano-platelets of graphite and hBN platelet-mixtures at
various concentrations on the TiAlN-coated WC ball insert to be used in the ball-milling
tests. The mixtures include vegetable oil mixed with 0.1 and 0.5wt% of hBN300 and 0.1
wt% and 1.0 wt% of xGnP grade M5 and xGnP C300 and C750 grades. Figure 121
shows the equipment and setup for the wetting angle measurement. To generate
consistent amounts of droplets, the motorized syringe assembly (from AST Product,
Inc.) was used to dispense a droplet of 0.5, 1, or 2.0 µl on the tool surface. The CCD
camera was used to capture the image of the droplet produced by the LED black light.
The droplet image was processed to obtain the wetting angles based on the boundaries
of the droplet and the tool surface.

155

FVS − FLS = FLV cos θ
F : Solid-Vapor Interfacial Force
VS

F : Liquid-Solid Interfacial Force
LS

F : Liquid-Vapor Interfacial Force
LV

Figure 121: Wetting Measurement system [Park, 2011]
IV.3.3 Surface characterization of two coated inserts using in tribotest and endball milling experiment
In the MQL milling experiments, two different types (denoted as A and B) of ball nose
end mill TiAlN-coated carbide inserts from two manufactures were used. To study the
effect of tool surface quality on the performance of solid lubricant platelet-enhanced
MQL mixtures, the topography of the surface of each tool was investigated. The surface
appearance and profile of these tool inserts were investigated by SEM images. The
roughness parameters such as the ten-point height mean roughness (Rz) and mean
spacing between peaks (Sm) were also measured with the Dektak 6M stylus surface
profilometer with the point-to-point resolution of 1 angstrom as shown in Figure 121.

156

These data were attained at a horizontal resolution of 0.067µm/sample over the
average value of 30,000 data points on the 2mm evaluation length scan in the parallel
and perpendicular directions of the grinding marks on the tool inserts.

Stylus
Came
ra
Sampl
e Stage
Rotat
e Stage
X
Adjustment

Y
Adjustment
Figure 122: Veeco Dektak 6M Surface Profiler
IV.3.4 Tribometer Tests
The friction measurements on the mixtures were attained with the tribometer test.
The tests were performed with a linear ball-on-disc type tribometer (CSM Instruments)
as shown in Figure 123 where a steel 440c ball, with 6.35 mm in diameter, oscillated on
TiAlN-coated carbide flat and rough surfaces (tool A and B) to evaluate frictional
characteristics under various lubrication conditions. The software calculates the
coefficient of friction between two sliding objects at a given normal load. All the tests
were set at a track length of 6mm (with an amplitude of 3mm) while carrying a load of
1, 5 and 10N with a sliding speed of 1.0, 2.5 and 4.0 cm/s at room temperature as
summarized in Table 4. The lubrication conditions include dry, pure vegetable oil and

157

mixtures of vegetable oil with 0.1, 1.0wt% of graphite and xGnP (M5); 0.1, 0.5wt% of
hBN300 and hBN5.
The wear characteristics under each lubricant condition were measured with a WC
ball instead of the 440C steel ball used in the friction tests in order to accelerate the
wear on the surfaces. The wear tests were conducted on both tool A and B, on which
depth and width measurements of the wear tracks were measured after 25,000, 35,000
and 50,000 cycles. However, since it is not possible to continue the wear test on the
exactly the same location, each wear cycle started from scratch. The average depth and
width on the wear tracks were measured on three cross-sections along a wear track
length of 6mm. Wear rate and volume of the wear track were calculated based on the
measurements from Dektak 6M Stylus Surface Profilometer. Figure 124 presents the
profile of a cross-section of the wear tracks using a Dektak 6M Surface Profiler which
the wear depth and width were measured.

Ball holder

Load
Sample
Mounting

W
Steel ball

Ball

L
Wear

Reciprocating
Actuator

Sample
(Tool A & B)

Figure 123: Linear ball-on-disc type tribometer

158

Table 21: Parameter for Tribometer tests
Tool

Ball-nose TiAlN-coated carbide inserts Tool A
and tool B (diameter of 25mm from two
manufacturers)

Ball

Steel (in friction test) & WC (in wear
resistance test), diameter of 6.35mm(1/4”)

Speed

0.25cm/sec, 1.0cm/sec, 2.5cm/sec, 4.0cm/sec

Normal load

1N, 5N, 10N

Length of track

L=6mm

Running time

• 3333cycles (14m with amplitude of 3mm) for
friction tests
• 25000 and 50000 cycles for wear tests

Figure 124: Profile of a cross-section of wear track.
IV.3.5 Ball Milling Experiments with steel AISI 1045
The MQL dispensing device (Uni-MAX) provided by Unist, Inc. (Grand Rapid,
Michigan) was used for MQL process providing the mist to the cutting area in the ball
milling experiments. The device sprays the vegetable oil through an external co-axial
nozzle. The definitions of the pitch and yaw angles for MQL nozzle was presented in
Figure 114.

Pitch angle is the nozzle angle respected to XY plane. Yaw angle is
159

defined by the angle between nozzle and X axis in the XY plane. The nozzle outlet
pressure and flow rate can be adjusted with the air metering screw and pulse
duration/frequency in the control panel, respectively. According to the previous work
[Park, 2010], the optimum MQL condition was determined to be 8 psi and 1.5ml/min for
outlet pressure and flow rate, respectively. The ball-mill experiments under various MQL
conditions were performed on the 3-axis vertical milling center (Sharnoa CNC mill) as
shown in Figure 125. The 25mm diameter end ball-nose TiAlN coated carbide inserts
(Tool A and B) were used for tool materials. AISI 1045 steel (203.2 mm×127 mm×203.2
mm) was used for work materials. In the experiment, the feed rate and cutting speeds
were kept constant at 2500mm/min and 3500 rpm, respectively. The depth of cut (DOC)
and radial depth of cut (ROC) were 1 mm and 0.6mm, respectively. The machining
conditions are summarized in
Table 22. The cutting started at one corner of the work material in the direction of
203.2mm and continued line by line to finish one layer. A layer is removed when all the
passes are completed at the same vertical height. In each layer (203.2 mm×127 mm),
the tool cuts the total of 212 lines. The tool wear was also measured after cutting each
layer to record the progress of tool wear. Due to the limited supply of work materials, the
maximum of cutting layer was set at 8 layers. The milling was conducted at different
lubrication conditions: dry condition and MQLs (pure MQL oil and nano-platelets
enhanced MQL oil). This study mainly focuses on the steady stated wear. To minimize
the effect of chipping and fracture in our analysis, the insert was considered broken if
the chipping and fracture were larger than twice of the average of flank wear width.
Table 22: Machining conditions for steel AISI 1045

160

Tool

Ball-nose TiAlN-coated carbide inserts Tool A and
Tool B (diameter of 25mm)

Feed Rate

2500 mm/min

Axial DOC

1 mm

Radial DOC

0.6 mm

Cutting Speed

3500rpm (108 m/min)

Lubricants

Dry,
MQL oil
Nano-platelets xGnP enhanced MQL oil at 0.1, 1,
and 5 wt%

MQL
spray Pitch angle: 15°
parameter
Yaw angle: -30, 60, 120 and 180°
Outlet pressure: 8 psi
Flow rate: 1.5ml/min

161

Outlet pressure
gauge
MQL oil

Z axis

Spindle

MQL
Nozzle

Compressed air

Tool
Holder

MQL
Nozzle

Z

Y axis Tool
Insert

Y

X axis

Pitch angle
Yaw angle

Workpiece
Vice

CNC
table
Outlet
Pressure
Gauge
Tool
Holder

Nose Wear

Flank Wear

MQL
nozzle
Workpiece

End ball milling
Insert
Figure 125: Experimental Set up for MQL ball milling

162

IV.4 EXPERIMENTAL RESULTS AND DISCUSSION
IV.4.1 Stability of Mixtures
The tests showed that the graphite mixtures were stable less than one day after
mixing in comparison to more than 3 days for the xGnP mixtures. The results also
showed that the stability of the mixtures was improved with a longer mixing time for
both micro- and nano-platelets. Park et al. [2011] showed that the larger diameter
platelets were segregated quicker than the smaller diameter platelets with the same
thickness. Based on the stability tests with graphite, xGnP M5 and xGnP M15, it
was determined that the thickness of the platelets influenced the stability of the
mixtures more than the diameter. For example, the graphite micro-platelets with
thickness of 200nm and average diameter 7µm were segregated much faster than
both xGnP M5 and M15 nano-platelets (with the same thickness of 6-8nm), which
have the diameter of 5µm and 15µm, respectively. The stability of the mixtures with
0.1wt% platelets after mixing for 3 days are shown in Figure 126, of which xGnP
C750 provided the most stable mixture because of the smaller thickness and
diameter.

Graphite

xGnP M5 xGnP C300 xGnP C750 hBN300

hBN5

Figure 126: The stability of the mixtures with 0.1wt% micro- and nano platelets after 3
days

163

IV.4.2 Wetting angle measurement
Figure 127 compares the wetting angles of water, water mixed with mineral oil
produced by NRG Resources (NRG oil), pure unist oil and mixtures of unist oil with
micro and nano-platelets for droplets of 0.5 µm. It can be seen not only that the
introduction of nano-platelets decreases the wetting angle but also that the wetting
angle of hBN is slightly smaller but comparable to those of the xGnP mixtures. In the
later section, the result from our tribometer test revealing the preponderant wear
resistance of hBN to xGnP will be shown.

Figure 127: Wetting angle of MQL lubricants on the surface of TiAlN coated carbide
inserts (left angle, right angle)
IV.4.3 Surface characterization of two coated inserts
Figure 128 presents the SEM micrographs of both (Tool A and B) tool surfaces
at 1000x and 5000x magnifications. Figure 129 presents Rz and Sm for measurements
over the length of 200 um which were along and perpendicular to the grinding marks of
the tools. Despite the similar spacing parameters, the roughness of tool B is around 1.6

164

µm while the roughness of tool A is 0.6 µm. Thus, tool A has a smoother surface, which
we expected to yield a better sliding contact with less wear.

Grinding marks
a) Tool A at 1000x

b) Tool B at 1000x

c) Tool A at 5000x

d) Tool B at 5000x.

Figure 128: SEM surface images of tool surfaces
Sm (Mean Spacing Between Peaks) over
the lengh of 2000um
50

Rz (Ten Point Height Average) over the
lengh of 2000um
2000

Tool A

Tool B 1688.12

Tool A

1200

800

40

Sm(um)

Rz(nm)

1600

Tool B 44.93

1515.48

601.81

29.52

30

24.83

20
686.84

400

10

0

0

Parallel

26.59

Parallel

Perpendicular

Perpendicular

Figure 129: Roughness (Rz) and spacing parameters (Sm)

165

IV.4.4 Tribometer Tests
Because of the space limitations, only a few selected results are presented in this
thesis to convey the effectiveness of this technology. Figure 130 depicts the friction
coefficients at sliding speeds of 2 and 4 cm/sec under various loads. The friction results
show that the oil in these mixtures primarily controls the friction behavior as the friction
coefficients with oil and oil mixtures do not vary much. Also, no significant change in the
frictional behavior can be found between nano-platelet-enhanced and micro-plateletenhanced lubricants. As noted in Park et al. [2011], the friction behavior of the
xGnP(M5)/oil mixtures did not differ much with that of the hBN/oil mixtures and did not
change much compared to pure oil. With the increase in load, the friction increased
slightly as shown in Figure 131. At elevated temperatures, solid lubricants are expected
to provide a better lubricant performance [Allam, 1991] as expected in MQL machining.

166

0.4

Friction Coefficient

a) At sliding speed of 2.5cm/s
0.314

0.312

0.312

0.3

0.2

0.1

0.060

0.064

0.062

0.0
1N

5N

10N

Dry

Load (N)
Pure Unist oil

Graphite 0.1wt%

xGnP-M5 0.1wt%

xGnP- M5 1.0wt%

xGnP-C300 0.1wt%

xGnP-C750 0.1wt%

hBN300 0.1wt%

hBN300 0.5wt%

hBN5 1.0wt%
0.4

Friction Coefficient

b) At sliding speed of 4.0 cm/s
0.3

0.2
0.123

0.114

0.097
0.1
0.012

0.020

0.016

0.0
1N

5N

10N

Load (N)
Figure 130: Friction coefficients of various mixtures on tool A

167

0.14
0.128

hBN 0.1wt

0.120
0.12

Friction Coefficient

xGnP 0.1wt
0.10

At load of 10N
0.08
0.061

0.062

0.06

0.04
0.019

0.02

0.018

0.00
1cms

2.5cms

4.0cms

Figure 131: Friction coefficients of mixtures with xGnP (M5) and hBN300 as function of
sliding speeds.
Despite the similar friction behaviors of xGnP (M5) and hBN300 enhanced lubricants,
the wear behavior has been substantially improved with the hBN300 enhanced
lubricants. Figure 132 shows the comparison in appearance of wear tracks between
micro- and nano-platelet/oil mixtures. The comparative results in wear resistance
performance between micro and nano-platelet/oil mixtures were summarized in Figures
133 - 134 where the wear depth and width were presented. The wear tracks generated
with the nano-platelets mixture were smaller than those with the micro-platelets mixture
in terms of both depth and width on Tool A. Among three grades of xGnP, M5, C300
and C750, the best result was obtained with C750 grade with its thickness of 1.2nm,
which indicates that the thickness is an important factor. The effect is much more

168

pronounced on the track depth than on the track width. Because of the thickness of the
nano-platelets, they can effectively cover the large area of the tool surface, which
facilitate sliding in their shear planar direction between the WC ball and tool surface. On
the other hand, due to its larger thickness and the associated low coverage area, the
micro-platelets did not perform as well. The wear track data on Tool B showed slight
improvement in wear resistance using nano-platelet mixtures compared to microplatelet mixtures as shown in Figure 133. One common lubrication condition between
Figures 133 and 134 is a mixture with nano-platelets of hBN300 at 0.5wt%. The wear
depth of Tool A is one fourth that of tool B, which cannot be completely explained by the
difference in surface roughness. This may be possibly due to the coating quality.

Figure 132: The wear track appearance with 0.1wt% of nano-platelets at a speed of
2.5cms and load of 10N after 35000cycles (Left: xGnP M5, Right: hBN300)

169

Graphite (0.1wt%)

xGnP M5 (0.1wt%)

xGnP C300 (0.1wt%)

xGnP C750 (0.1wt%)

hBN300 (0.1wt%)
hBN5 (0.5wt%)
300

Wear Track Width (um)

1600

Wear Track Depth (nm)

hBN300 (0.5wt%)

1400
1200
1000
800
600
400

250
200
150
100
50

200
0

0

25000cycles

25000cycles

50000 cycles

a) Depth of wear tracks

50000 cycles

b) Width of wear tracks

Wear Track Depth(nm)

1800

hBN300
(0.1wt%)
hBN300
(0.5wt%)
hBN5
(1.0wt%)
hBN5
(5.0wt%)

1600
1400
1200
1000
800
600
400

Wear Track Width (µm)

Figure 133: Depth and Width of wear track under various lubricant conditions on tool A
(Normal load: 10N, Speed: 2.5cm/s)

350
300
250
200
150
100
50

200
0

0
25000cycles

50000cycles

25000cycles

a) Depth of wear tracks

50000cycles

b) Depth of wear tracks

Figure 134: Depth and Width of wear track under various lubricant conditions on tool B
(Normal load: 10N, Speed: 2.5cm/s)

170

Figure 135 presents two platelets, hBN5 and xGnP M5 in relation to the surface
characteristics of Tool A and B. The difference in the vertical and horizontal scales,
especially for the hBN5 platelet, should be noted. The xGnP M5 is very thin compared
to the surface characteristics of the tools. Only a few (1-2) hBN5 micro platelets filled in
a valley on tool surface. The larger area coverage by the nano-platelets enables them to
provide better lubricity and resistance to wear. Based on this primary finding, MQL ball
mill tests were conducted with both Tool A and B to evaluate the performance of the
nano- and micro-platelet lubricants in an experiment with MQL ball milling.
Micro-platelets hBN5 on the tool surfaces

Nano-platelets xGnP M5 on the tool surfaces

1
0.1
0.8

Tool A

0.6

Tool B

0
Tool A
-0.1

Platelets
0.4

Tool B

H e ig th ( µ m )

H e ig th ( µ m )

-0.2
0.2
0
-0.2

-0.3
-0.4
-0.5

Rz
-0.4

-0.6
-0.6

Platelets

-0.7
-0.8
-0.8

Mean spacing (Sm)
-1
-30

-20

-10
0
10
The cross-sections of Tool surfaces(µm)

20

a)

tool

hBN

5

on

30

-15

-10
-5
The cross-sections of Tool surfaces(µm)

0

b) xGnP M5 on tool surfaces

Figure 135: Geometric Relationships of micro and nano-platelets on the tool surfaces
IV.4.5 Tool Wear with Ball Milling Experiments
IV.4.5.1 Optimal MQL spray angles for ball milling
Based on the cutting geometry calculation with tool diameter and DOC, the pitch
angle of 11° is the minimum angle that the oil-mist flow can access the full cutting zone
171

at any yaw angle as shown in Figure 136. Thus, the pitch angle was fixed at 15° and the
MQL ball milling tests were conducted with Tool B to find the optimal yaw angle for our
MQL experiments. The vegetable oil mixed with 1.0wt% of xGnP (M5) was used for
these experiments. Various yaw angles, -30, 60, 120 and 180°, were chosen for the
experiment. Figure 137 shows the resulting flank wear at various yaw angles which
showed that the negative 30° yaw angle setup provided a relatively good lubrication
conditions for the cutting region while the 120° yaw angle configuration showed the
worst conditions. The results showed 180° is the optimal yaw angle contributing to
adequate lubrication. This finding is corroborated by the geometry calculation from
experiment setup with 30° and 120° yaw angle as shown in Figure 138. At 30°yaw
angle, the rake face, flank face and nose were lubricated by the oil-mist during the MQL
cutting process. In other words, the lubricant oil-mist was not blocked by tool as tool
entered into the cutting region. Based on this finding, a pitch angle of 15° and a yaw
angle of 180° were set for the MQL ball-mill tests.

Figure 136: Minimum pitch angle of MQL nozzle for oil mist entering the cutting zone

172

180

120deg

-90deg

60deg

-150deg

-30deg

180deg

Flank wear (µm)

160
140
120
100
80
60
40
25.81

51.61

77.42

103.23

129.03

Cutting Volume (cm3)
Figure 137: Flank wear at 15° pitch angle and various yaw angles with tool B (dash-line:
Tool chipping)

Figure 138: Top View of MQL Experiment: The distribution of lubricant at 120° and -30°
yaw angle
IV.4.5.2 Effectiveness of thickness and diameter of platelets to tool wear
A series of MQL ball-milling tests was carried out on AISI 1045 steel with the
vegetable oil mixed with nano- and micro-platelets to study the effects of the thickness
on the wear performance. The lubrication conditions include pure vegetable oil,
vegetable oil mixed with micro-platelet additives (0.1wt% of graphite, 0.5 and 5.0wt% of

173

hBN5) and vegetable oil mixed with nano-platelet additives (0.1 and 1.0 wt% of xGnP
M5, xGnP C300 and xGnP C750 and 0.5wt% of hBN300) used with Tool A and B.
Figures 139-140 showed nose wear and flank wear at various lubrication conditions. In
appearances, the nose wear and flank wear with MQL mixtures were much lower than
those of dry machining. Due to nose of the insert being in contact with the workpiece
during the cutting process and not exposed to MQL oil mist, the improvement in nose
wear of MQL mixtures with nano-platelets to MQL with pure oil was not significant and
consistent as summarized in Figure 141. To help discussion on flank wear without too
much information, Figure 142 for Tool A and Figure 143 for Tool B present some of the
flank wear results.

Nose wear
a) Dry

b) 0.1wt% xGnP M5

c) 0.1 wt% hBN300

d) 0.5 wt% hBN300

Figure 139: Central wear with MQL nano-platelet enhanced mixtures after milling 3
layers

174

a) wear
Flank

a) Dry

b) 0.1wt% xGnP M5 c) 0.1 wt% hBN300
hBN300
Figure 140: Flank wear after milling 6 layers.

d) 0.5 wt%

70
Pure Oil

hBN300 (0.1 wt%)

hBN300 (0.5 wt%)

xGnPC750 (0.1wt%)

xGnPC300 (0.1wt%)

xGnPM5 (0.1wt%)

60

Nose wear (µm)

50
40
30
20
10
0
1

2

3

4 Layers 5

6

7

8

Figure 141: Nose wear at 3500 rpm after ball milling six layers with tool A

175

120

Flank wear (µm)

100

Graphite (0.1wt%)

hBN5 (0.5wt%)

Pure Oil

xGnP M5 (0.1wt%)

xGnP C300 (0.1wt%)

xGnP C750 (0.1wt%)

hBN300 (0.5 wt%)

80
60
40
20
0
1

2

3

4

5

6

Layers
Figure 142: Flank Wear at 3500rpm after ball milling six layers with tool A (dash-line:
Tool chipping)

176

Flank wear (µm)

120

100

80

Pure Oil
hBN5 (5.0 wt%)

60

hBN5 (0.5wt%)
xGnP M5 (1.0wt%)

40
25.81

51.61

77.42

103.23

129.03

Cutting Volume (cm3)
Figure 143: Flank Wear at 3500 rpm after ball milling six layers with tool B
The micro-platelet enhanced lubricants increased tool wear than pure oil lubricant
(traditional MQL), exposing the ineffectiveness of the micro-platelets. Comparing three
xGnP grades, M5, C300 and C750, The flank wear with both C300 and C750 grades,
despite chipping after cutting 5 layers, was reduced much more than M5. The flank
wear with M5 was at the level of pure oil. Similarly, the comparison between hBN5
(micro-platelet) and hBN300 (nano-platelet) also indicates the importance of the nanoscale thickness. The MQL-ball mill experiments with nano-platelet enhanced lubricants
performed better than pure oil lubricants (traditional MQL).

177

Chapter 5:

Tool wear improvement in machining of Ti64 with
nano-platelets enhanced MQL

From the previous chapter (or Chapter 4), the thickness of both graphite and hBN
need to be in nano scale in order to be effective in MQL. Obviously, the cutting
conditions in machining Ti64 are more aggressive with higher cutting forces and
temperatures than machining steel AISI1045. The nano-platelet enhanced lubricant
found to be very effective in machining steel AISI1045 may not be effective. In this
chapter, thus, the performance of MQL with xGnP C750, which performed the best
among xGnP grade with steel AISI1045, was evaluated in machining Ti64. The
performance of the lubricant mixed with nano xGnP C750 could be better for crater
wear where TiC reaction layer act as a barrier to dissolution and diffusion of tool in the
chip. These wear mechanisms may not be operating in flank wear, which is more critical
in milling.
V.1 EXPERIMENTAL SETUP AND PROCEDURES
V.1.1 Ball Milling Experiments with Ti64
The MQL ball-milling setup with Ti64 was the same as that with AISI 1045 steel as
shown in Figure 125. The 25mm diameter end ball-nose TiAlN coated carbide inserts
(Tool B) were used for tool materials. The Ti workpiece was TIMETAL Ti64 block
(203.2mm×101.6mm×203.2mm) from Titanium Metals Corporation (Tomroto, OH),
whose ingot chemical analysis of the Ti64 sample is shown in Table 23. In the
experiment, the cutting speeds were set at 2500 rpm and 3500 rpm. The feed rate and
the radial depth of cutting (ROC) were kept constant at 2500mm/min and 0.6mm,
178

respectively. Due to the poor machinability of Ti64, the depth of cut (DOC) was set at
0.5mm, which is a half of the DOC used in milling of AISI 1045 steel. The machining
conditions for Ti64 are summarized in Table 24. The status of tool wear was captured
after cutting each layer (203.2mm×101.6mm) (170 lines of cutting for each layer) to
record the progress of tool wear. The maximum of cutting layer was set at 8 layers. The
milling was conducted at different lubrication conditions: dry, pure MQL oil and MQL oil
enhanced with nano-platelets xGnP C750. To study mainly the steady stated wear with
minimized effect of chipping and fracture in our analysis, the insert was considered
broken if the chipping and fracture were larger than twice of the average of flank wear
width.
Table 23: Ingot chemical analysis of Ti64 work material
Fe

V

Al

C

O

N

Y

Top

0.15

4.00

6.23

0.025

0.18

0.008

<0.0005

Bottom

0.16

3.89

6.19

0.026

0.19

0.008

<0.0005

179

Table 24: Machining conditions for Ti64
Tool

25mm diameter end ball-nose TiAlN coated
carbide inserts (Tool B)

Feed Rate

2500 mm/min

Axial DOC

0.5mm

Radial DOC

0.6 mm

Cutting Speed

2500 rpm (55 m/min)
3500rpm (77m/min)

Lubricants

Dry,
MQL oil
Nano-platelets xGnP enhanced MQL oil at
0.1, 1, and 5wt%

MQL spray
parameter

Pitch angle: 15°
Yaw angle: 180°
Outlet pressure: 8 psi
Flow rate: 1.5ml/min

V.1.2 Tool wear measurements
Because chipping and tool fracture are common in machining of titanium alloys, flank
wear is characterized by capturing four areas,VB1, VB2, VB3, VB4 of the cutting edges,
as shown in Figure 144. Seven data points were taken for the flank wear width on each
area as shown in Figure 145. The maximum and average values of the flank wear width
(VBmax, VBavg) are reported based on a total of 28 data points.

180

Nose Wear
V

V
B4

B3

V
B2

V
B1

TiAlN coated
WC insert
Crater Wear

Figure 144: Types of tool wear were measured on end-ball nose inserts

a)

b)

Figure 145: The flank wear measurement in a) VB2 and b) VB1 on the cutting edge
before and after etching the adhesion layer of titanium.

181

V.2 EXPERIMENTAL RESULTS AND DISCUSSION
V.2.1 Flank wear and chipping on cutting edges
V.2.1.1 Flank wear at low cutting speed (2500rpm)
As expected, the milling inserts used in MQL conditions cut more layers than under
dry conditions. Figures 146-149 presents the flank wear at VB1 on the inserts after
cutting the 1st to 8th layers under each lubrication condition at a cutting speed of
2500rpm. The maximum flank wear’s width, VBmax, and the average values of the flank
wear widths, VBavg are summarized in Figures 150 and 151. The MQL mixture with
1wt% of xGnP C750 is the best followed by the mixture with 0.1wt% of xGnP C750. In
the case of dry machining, the flank wear was not even due to the numerous chippings
and tool fractures. Consequently, the insert was only able to cut 5 layers. The flank
wear was more uniform when using the MQL mixture with 1wt% of xGnP C750. Only a
few minor chippings at the cutting edge were present with pure oil and the mixture with
0.1wt% of xGnP C750. Therefore, a certain concentration of nano-platelets of xGnP are
needed to substantially reduce both flank wear and chipping on the tool. In steady state
wear (VBavg), the mixture with 1wt% of xGnP C750 was most effective, which yielded
15% and 25% reduction in flank wear after cutting 5 layers. More importantly, machining
was stopped before reaching the life of each insert.

182

st

1 layer

2

nd

rd

layer

3 layer

th

4 layer

The milling cannot continue due to chipping reaches
the limit (insert was considered as broken one)
th

5 layer
Figure 146: The flanks wear in VB1 on insert with Dry at 2500rpm.

1st layer

5th layer

2nd layer

3rd layer

6th layer

7th layer

4th layer

8th layer

Figure 147: The flank wear in VB1 on the insert with pure Unist oil at 2500rpm.
In [Park, 2011], the optimum concentration of xGnP platelets was 0.1wt% in
machining of AISI 1045 steel. The higher concentration, 1wt%, did not perform as well
as. In the machining experiment with Ti alloys, 1.0wt% is the optimum weight content for
at a cutting speed of 2500rpm. The optimum concentration may be affected by the
dimensions of xGnP platelets and the surface quality of the tool.
183

1st layer

2nd layer

3rd layer

4th layer

5th layer

6th layer

7th layer

8th layer

Figure 148: The flank wear in VB1 on insert with xGnP C750 0.1wt% at 2500rpm.

1st layer

2nd layer

3rd layer

4th layer

5th layer

6th layer

7th layer

8th layer

Figure 149: The flank wear in VB1 on insert with xGnP C750 1wt% at 2500rpm.

184

The maximun Flank wear VBmax (µm)

200

160

120

80

40

Dry_Max

Unist_Max

xGnP750 1wt%_Max

xGnP750 0.1wt%_Max

0
1

2

3

4

5

6

7

8

Cutting layer
Figure 150: Maximum flank wear with (VBmax) at 2500rpm (after etching)

The average Flank wear VBavg (µm)

90

80

70

60

50

Dry_Avg

Unist_Avg

xGnP750 1wt%_Avg

xGnP750 0.1wt%_Avg

40

30
1

2

3

4

5

6

7

8

Cutting layer
Figure 151: The average flank wear width (VBavg) at 2500rpm (after etching)

185

V.2.1.2 Flank wear at high cutting speed (3500rpm)
As shown in Figures 152 and 153, the enhancement of MQL solution is not as
pronounced because of the chipping under high impact conditions. It is important to
note that chipping occurred substantially more when cutting of the first few layers. The
higher concentration (5wt%) of xGnP nano-platelets was tested with an expectation of
more lubricity and/or damping due to more xGnP platelets in the lubricant. However, the
experimental results did not agree with the expectation. Among the MQL mixtures, the
mixture with 1wt % of xGnP C750 is the best followed by dry, Unist oil and 5wt% as
shown Figures 15 and 16. The reason why enhanced MQL solution at 5wt% did not
perform well is not clear at this time. It could be the poor wettability with the tool
surfaces or, the hindrance of the sliding actions at the high concentrations.

Dry

Unist

C750 -

C750 -

Figure 152: The flank wear in VB1 after 1st layer with different lubrication conditions at
3500rpm.

Dry

Unist

C750 -

C750 -

Figure 153: The flank wear in VB1 after 2st layer with different lubrication conditions at
3500rpm.

186

170
150

VBmax(µm)

130
110
90
70
50
30

Dry_Max

Unist_Max

xGnP750 1wt%_Max

xGnP750 5wt%_Max

1

2

3

4

Cutting layer

Figure 154: Maximum flank wear width (VBmax) at 3500rpm (after etching)
100
90

VBavg (µm)

80
70
60
50

Dry_Avg

Unist_Avg

xGnP750 1wt%_Avg

xGnP750 5wt%_Avg

40
30
1

2

3

4

Cutting layer
Figure 155: The average flank wear width (VBavg) at 3500rpm (after etching)

187

V.2.1.3 Micro-chipping and tool fracture
Chipping and tool fracture are the common modes of tool damage in machining Ti
alloys. Especially for the consistently interrupted cutting process of milling, the tool
experiences high stress cycles and high impact leading to the cutting edges being
chipped frequently. In this work, the tool damage were classified into two categories:
micro-chipping and fracture, as shown in Figure 156. The elliptical shape represents
micro-chipping

while

the

rectangular

shape

represents

fracture.

They

were

distinguished based on the size; micro-chippings if the size was smaller than half of the
average flank wear width and tool fractures otherwise. To be more confident, the type of
chipping was evaluated based on the images from the flank face and the rake face. At
the low cutting speed, the MQL solutions (pure oil and the mixture with nano-platelets)
helped to reduce micro-chipping and prevent the propagation of micro-chipping into tool
fractures. Figures 157 and 158 represent the cutting layer on which the first damage
occurred at the cutting speeds of 2500rpm and 3550rpm, respectively, for various
lubrication conditions. The largest damage on tools under various lubrications was
presented in Figures 159 and 160 for cutting speeds of 2500rpm and 3550rpm
respectively. At low cutting speed, the tool fractures occurred at the first cutting layer
with dry milling while only micro-chippings were found with MQL mixtures even after
cutting 8th layers. At the cutting speed of 3500rpm, due to the high impact condition,
both micro-chippings and tool fractures occurred even with the MQL conditions. Tool
fractures happened after cutting only one layer under dry condition were more severe
than those under MQL mixtures which occurred after cutting 3-4 layers. The higher

188

concentration of nano-platelets did not help to reduce chipping. The concentration of
1wt% outperformed the others in preventing tool damage.
Micro-chipping

Micro-chipping

VBavg

VBavg

a) Micro-chipping
Fracture

Fracture

VBavg

VBavg

Fracture

b) Tool fracture
Figure 156: Types of tool damage on ball nose end mill insert.
st

Dry 1 layer

C750st(0.1wt%)
1 layer

Unist 1st layer

C750 (1wt%)
2nd layer

Figure 157: The first damage at the cutting edge with different lubrication conditions at
2500rpm.

189

Dry 1st
layer

Unist 1st
layer

C750
(1wt%) 1st
layer

C750
(5wt%) 1st
layer

Figure 158: The first damage at the cutting edge under various lubrication conditions at
3500rpm.
Dry 5th
layer

th

C750
(0.1wt%)
8th layer

Unist 8

C750
(1wt%)
8th layer

Figure 159: The largest damage at the cutting edge under various lubrication conditions
at 2500rpm.
Dry 3rd

Unist 2nd
layer

C750
(1wt%) 4th
layer

C750
(5wt%) 3rd
layer

Figure 160: The largest damage at the cutting edge with different lubrication conditions
at 3500rpm.
V.2.2 Nose wear of insert
The cutting speed at the nose of the insert is very low due to the small effective
diameter (represented by De shown in Figure 161). Furthermore, the nose of the insert
being in contact with the workpiece during the cutting process is the least lubricated
region by the MQL oils in the contact zone. As expected, the effectiveness of MQL
mixture on nose wear was not significant as on flank and rake face. Figure 162 and
190

Figure 163 present the center wear under certain lubrication conditions for low and high
cutting speeds, respectively. At low cutting speed, the nose wear under dry cutting was
higher than those MQL pure oil and MQL mixtures with xGnP C750. However, the
outperformance of MQL lubrication conditions to dry was reduced as high cutting
speeds increased.

Figure 161: The effective diameter of the cut, De, for ball-nose end-mill insert

Dry
st

1 layer

Unist
st

1 layer

C750
(1wt%)
1st layer

Nose
wear
Dry

Unist

5th layer

8th layer

C750
(1wt%)
8th layer

Figure 162: The nose wear at the first and the last cutting layer under different
lubrication conditions at 2500rpm.

191

Dry
1

Unist

st

1st layer

C750
(1wt%)

Nose wear

Dry

Unist

3rd

2nd layer

C750
(1wt%)

Figure 163: The nose wear at the first and the last cutting layer under different
lubrication conditions at 3500rpm.
V.2.3 Crater wear
In the milling process, the rake face of the tools did not experience the dramatic wear
unlike those in turning. In the milling experiment, only slight wear with a thin adhesion
layer was observed in the rake face of the inserts under all lubrication conditions.
Therefore, the MQL solution does not affect the rake face. The chipping at the cutting
edge had a more significant impact on the rake face than on the crater wear as shown
in Figures 164 and 165 for cutting speeds of 2500rpm and 3500rpm, respectively.
Fracture at the cutting edge was more frequent and larger in dry condition than the MQL
conditions at both low and high cutting speeds. Among MQL conditions, the lubricant
enhanced by nano-platelets helps to reduce the chipping at the cutting edge. The low
crater wear revealed the deformation of the tool is not as severe as in turning. A reason
for low crater wear is the cutting temperatures in milling were fairly lower than those in

192

turning which limited dissolution/diffusion. The abrasion of chip to rake face was less
compared with turning due to discontinuous chip in milling. In milling of Ti64 at low
cutting speed, cooling is not as critical as lubrication, which enhances the effectiveness
of MQL.

chipped
chipped

chipped

crater
wear

crater
wear

crater
wear

crater
wear

C750 (0.1wt%) 8th
C750 (1wt%) 8th
Unist 8th
Dry 3rd
layer
layer
layer
layer
Figure 164: The crater wear and damage on the tool at the last cutting layer with
different lubrication conditions at 2500rpm.

chipped

chipped

chipped

crater
wear

chipped
crater
wear
crater
wear

crater
wear
nd

th

Unist 2 layer
C750 (1wt%) 4
C750 (5wt%) 3rd
Dry 3rd
layer
layer
layer
Figure 165: The crater wear and damage on the tool at the last cutting layer with
different lubrication conditions at 3500rpm.

193

Chapter 6: Conclusions
In

the

turning

experiment

of

Ti64

with

carbide

and

PCD

inserts,

the

dissolution/diffusion wear is the dominant wear mechanism on the rake face of carbide
insert while grain-pulled out is the main mode of tool wear on rake face of PCD insert.
These conclusions are corroborated by the smooth craters found on carbide inserts and
the rough and uneven crater on PCD inserts. Because the crater wear started away
from the cutting edge and progressed rapidly to cutting edge on carbide inserts, tool
fracture on cutting edge was more excessive than the PCD inserts. The cutting edge on
PCD inserts was intact, which enable them to have longer tool lives. Abrasion by the
hard α-clusters and tool fractures were the main causes for flank wear on both carbide
and PCD inserts. At the high cutting speed, PCD inserts were additionally damaged by
notch wear. Obviously, for both carbide and PCD inserts, the wear in low cutting speeds
is not as extensive as those from high cutting speeds. The thermal conductivity and
hardness of PCD are much higher compared to those of the carbide inserts, resulting in
lower cutting temperatures and superior resistance to abrasion wear. Therefore, the
wear rates of PCD inserts were much smaller than those of carbide inserts at each
cutting speed. PCD showed a better performance in resisting flank wear independent of
cutting speed. However, the performance of PCD is even more superior at high cutting
speed.
The OIM results of the microstructure analysis of work material and chip in the
turning experiments of Ti64 proved the evident of phase change and root cause flank
wear and scoring marks on flank face. The chips generated with carbides and PCD
inserts contain the higher β-phase faction and higher peaks at 60° and 90°
194

misorientations compared to the original work material indicated that the phase change
α→β happened during machining. This result agreed well with the FEM simulation,
which showed the cutting temperature reaching beyond the transus temperature. The
microstructure of the original work material revealed the heterogeneity in the
microstructure coming from the presence of both α and β grains and the cluster of the αgrains in the hard orientations which abrade and fracture the tool surfaces,. In addition,
when the adhesion layers are detached from the inserts, some grains are pulled out of
the tool materials, further damaging the inserts. The hard α-clusters caused large,
steady and smoother scoring marks on ‘softer’ carbide surface and sometime fracture
with large ‘hard’ clusters. The hard α-clusters can damage the inserts in a more brittle
manner and the detached tool fragments further damaged mainly evident by the surface
patterns showing the grain fractured out from the flank surface with the scoring mark on
PCD inserts.
To enhance the machinability, the ball milling experiment of 1045 steel and Ti64 was
conducted with the MQL oil mixed with micro- and nano-platelets of graphite and hBN to
compare with dry and regular MQL cutting. Despite the fact that the friction of these
mixtures with micro and nano- platelets regardless of the thickness and concentrations
did not change from that of pure oil, the tribo-test clearly showed the advantage of
mixing nano-platelets of xGnP and hBN in reducing wear. These findings have been
corroborated with the experimental results from the ball milling experiment with 1045
steels. The MQL mixtures with micro-platelets (Graphite and hBN5) did not present any
advantage over the pure oil (traditional MQL) in reduction of flank wear while the
mixtures with only a small amount (0.1w% for xGnP M5, xGnP C300, xGnP C750 and
195

0.5wt% for HBN300) of nano-platelets presented a significant improvement in wear
performance. Among the xGnP grades, the thinner platelets showed better resisting
wear performances in both tribotest and MQL-ball mill test. In the milling of Ti64,
however, the content of nano-platelet C750 needed to be increased (up to 1wt%) to
provide significant improvement. MQL with nano-lubricants substantially outperforms to
dry and regular MQL machining by reducing flank wear and tool damage which consist
of micro chippings and fractures. At the cutting speed of 2500rpm, the tool fractured in
dry machining while only micro chippings occurred when the MQL mixtures were used.
At given cutting speed, cutting condition is more aggressive in machining Ti64 than
those in machining steel. As expected, the effectiveness of mixtures with nano-platelets
in MQL machining Ti64 was less significant at cutting speed of 3500rpm while those
mixtures showed outperformance at that cutting speed in machining 1045 steel.
Machining of Ti64 at the lower cutting speed, the nano-platelets enhanced MQL
presented much more effective. This nano-lubricant technology works because it
provides the necessary lubricity even when the oil dissociate due to the high cutting
temperature.

196

BIBLIOGRAPHY

197

BIBLIOGRAPHY
Alberts, M., Kalaitzidou, K., & Melkote, S. (2009). An investigation of graphite
nanoplatelets as lubricant in grinding. International Journal of Machine Tools and
Manufacture, 49(12-13), 966–970.
Allam, I. (1991). Solid lubricants for applications at elevated temperatures. Journal of
Materials, 26(15), 3977–3984.
Amrita, M., Srikant, R., Sitaramaraju, a., Prasad, M., & Krishna, P. V. (2013).
Experimental investigations on influence of mist cooling using nanofluids on
machining parameters in turning AISI 1040 steel. Proceedings of the Institution of
Mechanical Engineers, Part J: Journal of Engineering Tribology, 227(12), 1334–
1346.
Andriya, N., Rao, P. V., & Ghosh, S. (2012). Dry Machining of Ti-6Al-4V using PVD
Coated TiAlN Tools. Proceedings of the World Congress on Engineering, III, 2–7.
Arrazola, P.-J., Garay, a., Iriarte, L.-M., Armendia, M., Marya, S., & Le Maître, F. (2009).
Machinability of titanium alloys (Ti6Al4V and Ti555.3). Journal of Materials
Processing Technology, 209(5), 2223–2230.
Attanasio, A., Gelfi, M., Pola, A., Ceretti, E., & Giardini, C. (2013). Influence of Material
Microstructures in Micromilling of Ti6Al4V Alloy. Materials, 6(9), 4268–4283.
Bartz, W. J. (1971). Solid lubricant additives - effect of concentration and other additives
on anti-wear performance. Wear, 17(5-6), 421–432.
Bayoumi, a. E., & Xie, J. Q. (1995). Some metallurgical aspects of chip formation in
cutting Ti-6wt.%Al-4wt.%V alloy. Materials Science and Engineering: A, 190(1-2),
173–180.
Bermingham, M. J., Kirsch, J., Sun, S., Palanisamy, S., & Dargusch, M. S. (2011). New
observations on tool life, cutting forces and chip morphology in cryogenic
machining Ti-6Al-4V. International Journal of Machine Tools and Manufacture,
51(6), 500–511.
Boothroyd, G. (1975). Fundamentals of Metal Machining and Machine Tools.
Washington. DC: Scripta Book Company.
Boyer, R. R. (1996). An overview on the use of titanium in the aerospace industry.
Materials Science and Engineering: A, 213(1-2), 103–114.

198

Braga, D. U., Diniz, A. E., Miranda, G. W. ., & Coppini, N. L. (2002). Using a minimum
quantity of lubricant (MQL) and a diamond coated tool in the drilling of aluminum–
silicon alloys. Journal of Materials Processing Technology, 122(1), 127–138.
Brandao, L. C., Coelho, R. T., & Lauro, C. H. (2011). Contribution to dynamic
characteristics of the cutting temperature in the drilling process considering one
dimension heat flow. Applied Thermal Engineering, 31(17-18), 3806–3813.
Brinksmeier, E., Diersen, P., Zillmer, A., & Janssen, R. (1999). Cooling Lubrication
Reduction when Machining Advanced Materials. Aspects of Cooling Lubrication
Reduction in Machining Advanced Materials, 213(08), 769–778.
Britton, T. B., Liang, H., Dunne, F. P. E., & Wilkinson, a. J. (2009). The effect of crystal
orientation on the indentation response of commercially pure titanium: experiments
and simulations. Proceedings of the Royal Society A: Mathematical, Physical and
Engineering Sciences, 466(2115), 695–719.
Calamaz, M., Coupard, D., & Girot, F. (2008). A new material model for 2D numerical
simulation of serrated chip formation when machining titanium alloy Ti–6Al–4V.
International Journal of Machine Tools and Manufacture, 48(3-4), 275–288.
Cetin, M. H., Ozcelik, B., Kuram, E., & Demirbas, E. (2011). Evaluation of vegetable
based cutting fluids with extreme pressure and cutting parameters in turning of AISI
304L by Taguchi method. Journal of Cleaner Production, 19(17-18), 2049–2056.
Chen N. (2004). Coefficients of Friction for Graphite.
Cook, N. H., & Nayak, P. N. (1986). The thermal mechanics of tool wear. Journal of
Manufacturing Science and Engineering, 88(1), 93–100.
Corduan, N., Himbart, T., Poulachon, G., Dessoly, M., Lambertin, M., Vigneau, J., &
Payoux, B. (2003). Wear Mechanisms of New Tool Materials for Ti-6AI-4V High
Performance Machining. CIRP Annals - Manufacturing Technology, 52(1), 73–76.
Dearnley, P. A., & Grearson, A. N. (1986). P. A. Dearnley A. N. Grearson, 2(January),
47–58.
Dhar, Ahmed, M. T., & Islam, S. (2007). An experimental investigation on effect of
minimum quantity lubrication in machining AISI 1040 steel. International Journal of
Machine Tools and Manufacture, 47(5), 748–753.
Dhar, N. R., Paul, S., & Chattopadhyay, A. B. (2002). Machining of AISI 4140 steel
under cryogenic coolingÐtool wear, surface roughness and dimensional deviation.
Journal of Materials Processing Technology, 123(3), 483–489.

199

Donnet, C., & Erdemir, a. (2004). Solid Lubricant Coatings: Recent Developments and
Future Trends. Tribology Letters, 17(3), 389–397.
Emami, M., Sadeghi, M. H., Sarhan, A. A. D., & Hasani, F. (2014). Investigating the
Minimum Quantity Lubrication in grinding of Al2O3 engineering ceramic. Journal of
Cleaner Production, 66, 632–643.
Ezugwu, E. O., & Wang, Z. M. (1997). Materials Titanium alloys and their machinability.
Journal of Materials Processing Technology, 68, 262–274.
Fang, Z. Z. (2005). Correlation of transverse rupture strength of WC–Co with hardness.
International Journal of Refractory Metals and Hard Materials, 23(2), 119–127.
Fukushima, H. (2003). Graphite Nano-reinforcements in polymer Nanocomposites.
Ph.D. Thesis, Michigan State University, East Lansing, MI, USA.
Gabriel, S. B., de Almeida, L. H., Nunes, C. a, Dille, J., & Soares, G. a. (2013).
Maximisation of the ratio of microhardness to the Young’s modulus of Ti-12Mo13Nb alloy through microstructure changes. Materials Science & Engineering. C,
Materials for Biological Applications, 33(6), 3319–24.
Gaitonde, V. N., Karnik, S. R., & Davim, J. P. (2008). Selection of optimal MQL and
cutting conditions for enhancing machinability in turning of brass. Journal of
Materials Processing Technology, 204(1-3), 459–464.
Gente. (2001). Chip Formation in Machining_Ti64 (chip morphology vs CS).pdf. CIRP
Annals - Manufacturing Technology, 50, 49–52.
Groover, M. P. (2007). Fundamentals of modern manufacturing: materials processes,
and systems (3rd ed.). John Wiley & Sons.
Hartung, P. D., & Kramer, B. M. (1982). Tool Wear in Titanium Machining. Ann. CIRP,
31(1), 75–80.
Hasçalık, A., & Çaydaş, U. (2007). Optimization of turning parameters for surface
roughness and tool life based on the Taguchi method. The International Journal of
Advanced Manufacturing Technology, 38(9-10), 896–903.
Haubner, R., Wilhelm, M., Weissenbacher, R., & Lux, B. (2002). Boron Nitrides Properties, Synthesis and Applications (Vol. 102, pp. 1–45). Springer Berlin
Heidelberg.
Heinemann, R., Hinduja, S., Barrow, G., & Petuelli, G. (2006). Effect of MQL on the tool
life of small twist drills in deep-hole drilling. International Journal of Machine Tools
and Manufacture, 46(1), 1–6.
200

Hickman B.S. (1969). The Formation of Omega Phase in Titanium and Zirconium
Alloys: A Review, 4, 554–563.
Hosseini, A., & Kishawy, H. A. (2014). Machining of Titanium Alloys. In J. P. Davim
(Ed.), Machining of Titanium Alloys. Berlin, Heidelberg: Springer Berlin Heidelberg.
Hsu, H.-C., Wu, S.-C., Hsu, S.-K., Lin, T.-F., & Ho, W.-F. (2013). Structure and
mechanical properties of as-cast Ti–5Nb–xCr alloys. Materials & Design, 51, 268–
273.
Hsu, S. M. (2004). Nano-lubrication: concept and design. Tribology International, 37(7),
537–545.
Hua, J., & Shivpuri, R. (2005). A Cobalt Diffusion Based Model for Predicting Crater
Wear of Carbide Tools in Machining Titanium Alloys. Journal of Engineering
Materials and Technology, 127(1), 136.
Huang, H. D., Tu, J. P., Gan, L. P., & Li, C. Z. (2006). An investigation on tribological
properties of graphite nanosheets as oil additive. Wear, 261(2), 140–144.
Hughes, J. I., Sharman, A. R. C., & Ridgway, K. (2004). The effect of tool edge
preparation on tool life and workpiece surface integrity. Proc. IMechE Part B: J.
Engineering Manufacture, 220, 1113–1123.
Ibrahim, G. A., CheHaron, C. H., & Ghani, J. A. (2009). Surface integrity of ti-6al-4v eli
when machined using coated carbide tools under dry cutting condition. International
Journal of Mechanical and Materials Engineering (IJMME), 4(2), 191–196.
Jaffery, S. I., & Mativenga, P. T. (2008). Assessment of the machinability of Ti-6Al-4V
alloy using the wear map approach. The International Journal of Advanced
Manufacturing Technology, 40(7-8), 687–696.
Jawaid, a, Che-Haron, C. ., & Abdullah, a. (1999). Tool wear characteristics in turning of
titanium alloy Ti-6246. Journal of Materials Processing Technology, 92-93, 329–
334.
Jon, M. C., Engineering, M., Science, M., & York, N. (1972). Omega-Phase Formation
in Beta III Titanium Alloy, 150, 139–150.
Jones, N., Dashwood, R., Jackson, M., & Dye, D. (2009). Development of chevronshaped α precipitates in Ti–5Al–5Mo–5V–3Cr. Scripta Materialia, 60(7), 571–573.
Joshi, S., Pawar, P., Tewari, A., & Joshi, S. S. (2014). Effect of β phase fraction in
titanium alloys on chip segmentation in their orthogonal machining. CIRP Journal of
Manufacturing Science and Technology, 7(3), 191–201.
201

Kajdas, C., Kanpsika, A., & Kulczycki, A. (2010). Chemistry and Technology of
Lubricants. In R. M. Mortier, M. F. Fox, & S. T. Orszulik (Eds.), Chemistry and
technology of lubricants (pp. 239–292). Dordrecht: Springer Netherlands.
Kalita, P., Malshe, A. P., Arun Kumar, S., Yoganath, V. G., & Gurumurthy, T. (2012).
Study of specific energy and friction coefficient in minimum quantity lubrication
grinding using oil-based nanolubricants. Journal of Manufacturing Processes,
14(2), 160–166.
Khettabi, R., Songmene, V., & Masounave, J. (2009). Effects of Speeds, Materials, and
Tool Rake Angles on Metallic Particle Emission During Orthogonal Cutting. Journal
of Materials Engineering and Performance, 19(6), 767–775.
Khettabi, R., Songmene, V., Masounave, J., & Zaghbani, I. (2008). Understanding the
Formation of Nano and Micro Particles During Metal Cutting.pdf. International
Journal of Signal System Control and Engineering, 1(3), 203–210.
Kimura, Y., Wakabayashi, T., Okada, K., Wada, T., & Nishikawa, H. (1999). Boron
nitride as a lubricant additive. Wear, 232(2), 199–206.
Konig, W. (1979). Applied research on the machinability of titanium and its alloys. Proc.
AGARD Conf. Advanced Fabrication Processes, Florence, Italy.
Kramer, B. M., & von Turkovich, B. F. (1986). A Comprehensive Tool Wear Model.
CIRP Annals - Manufacturing Technology, 35(1), 67–70.
Kwon, J., Brandes, M. C., Sudharshan Phani, P., Pilchak, a. P., Gao, Y. F., George, E.
P.,
Mills, M. J. (2013). Characterization of deformation anisotropies in an α-Ti
alloy by nanoindentation and electron microscopy. Acta Materialia, 61(13), 4743–
4756.
Kwon, P., & Drzal, L. T. (2010). Nanoparticle graphite-based minimum quantity
lubrication method and composition. US.
Lammer, A. (1988). Mechanical properties of polycrystalline diamonds. Materials
Science and Technology, 4(November), 949–955.
Lee, C.-G., Hwang, Y.-J., Choi, Y.-M., Lee, J.-K., Choi, C., & Oh, J.-M. (2009). A study
on the tribological characteristics of graphite nano lubricants. International Journal
of Precision Engineering and Manufacturing, 10(1), 85–90.
List, G., Nouari, M., Géhin, D., Gomez, S., Manaud, J. P., Le Petitcorps, Y., & Girot, F.
(2005). Wear behaviour of cemented carbide tools in dry machining of aluminium
alloy. Wear, 259(7-12), 1177–1189.

202

Liu, Z. Q., Cai, X. J., Chen, M., & An, Q. L. (2011). Investigation of cutting force and
temperature of end-milling Ti-6Al-4V with different minimum quantity lubrication
(MQL) parameters. Proceedings of the Institution of Mechanical Engineers, Part B:
Journal of Engineering Manufacture, 225(8), 1273–1279.
López de Lacalle, L. N., Angulo, C., Lamikiz, a., & Sánchez, J. a. (2006). Experimental
and numerical investigation of the effect of spray cutting fluids in high speed milling.
Journal of Materials Processing Technology, 172(1), 11–15.
Lütjering, G., & Williams, J. C. (2003). Titanium (2nd ed.). Berlin: Springer.
Machado, A. R., & Wallbank, J. (1990). Machining of titanium and its alloys-a.
Proceedings of the Institution of Mechanical Engineers, Part B: Journal of
Engineering Manufacture, 204(1), 53–60.
Machai, C., Iqbal, A., Biermann, D., Upmeier, T., & Schumann, S. (2013). On the effects
of cutting speed and cooling methodologies in grooving operation of various
tempers of β-titanium alloy. Journal of Materials Processing Technology, 213(7),
1027–1037.
Maciej Motyka, K. K. J. S. and W. Z. (2012). Titanium Alloys - Towards Achieving
Enhanced Properties for Diversified Applications. (A. K. M. N. Amin, Ed.). InTech.
Mao, C., Tang, X., Zou, H., Huang, X., & Zhou, Z. (2012). Investigation of grinding
characteristic using nanofluid minimum quantity lubrication. International Journal of
Precision Engineering and Manufacturing, 13(10), 1745–1752.
Mao, C., Zou, H., Huang, X., Zhang, J., & Zhou, Z. (2012). The influence of spraying
parameters on grinding performance for nanofluid minimum quantity lubrication.
The International Journal of Advanced Manufacturing Technology, 64(9-12), 1791–
1799.
Meyer, L. W., Krüger, L., Sommer, K., Halle, T., & Hockauf, M. (2008). Dynamic
strength and failure behavior of titanium alloy Ti-6Al-4V for a variation of heat
treatments. Mechanics of Time-Dependent Materials, 12(3), 237–247.
Molinari, a, Musquar, C., & Sutter, G. (2002). Adiabatic shear banding in high speed
machining of Ti–6Al–4V: experiments and modeling. International Journal of
Plasticity, 18(4), 443–459.
Motonishi, S., Isoda, S., Itoh, H., Tsumori, Y., & Terada, Y. (1987). Study on Machining
of Titanium and Its Alloys. Kobelto Technology Review, 2, 28–31.
Nabhani, F. (2001). Machining of aerospace titanium alloys. Robotics and ComputerIntegrated Manufacturing, 17(1-2), 99–106.
203

Nageswara Rao, D., & Vamsi Krishna, P. (2008). The influence of solid lubricant particle
size on machining parameters in turning. International Journal of Machine Tools
and Manufacture, 48(1), 107–111.
Nandy, a. K., & Paul, S. (2008). Effect of coolant pressure, nozzle diameter,
impingement angle and spot distance in high pressure cooling with neat oil in
turning Ti-6Al-4V. Machining Science and Technology, 12(4), 445–473.
Narutaki, N., Murakoshi, A., Motonishi, S., & Takeyama, H. (1983). Study on Machining
of Titanium Alloys. CIRP Annals - Manufacturing Technology, 32(1), 65–69.
Narutaki, N., Murakoshi, A., Motonishi, S., & Takeyama, H. (1983). Study on Machining
of Titanium Alloys. CIRP Annals - Manufacturing Technology, 32(1), 65–69.
Nguyen, T. K., Do, I., & Kwon, P. (2012). A tribological study of vegetable oil enhanced
by nano-platelets and implication in MQL machining. International Journal of
Precision Engineering and Manufacturing, 13(7), 1077–1083.
Olortegui-Yume, J. A., & Kwon, P. Y. (2007). Tool wear mechanisms in machining.
Nternational Journal of Machining and Machinability of Materials, 2(3), 316–334.
Olortegui-Yume, J. a., & Kwon, P. Y. (2010). Crater wear patterns analysis on multilayer coated carbides using the wavelet transform. Wear, 268(3-4), 493–504.
Oosthuizen, G. A., Akdogan, G., Dimitrov, D., & Treurnich, N. F. (2010). A Review of the
Machinability of Titanium Alloys. R & D Journal of the South African Institution of
Mechanical Engineering, 26(September), 43–52.
Ozcelik, B., Kuram, E., Huseyin Cetin, M., & Demirbas, E. (2011). Experimental
investigations of vegetable based cutting fluids with extreme pressure during
turning of AISI 304L. Tribology International, 44(12), 1864–1871.
Park, K.-H., & Kwon, P. Y. (2011). Flank wear of multi-layer coated tool. Wear, 270(1112), 771–780.
Park, K.-H., Ewald, B., & Kwon, P. Y. (2011). Effect of Nano-Enhanced Lubricant in
Minimum Quantity Lubrication Balling Milling. Journal of Tribology, 133(3), 031803.
Paul, P. S., & Varadarajan, a. (2013). Performance evaluation of hard turning of AISI
4340 steel with minimal fluid application in the presence of semi-solid lubricants.
Proceedings of the Institution of Mechanical Engineers, Part J: Journal of
Engineering Tribology, 227(7), 738–748.

204

Prasad, B. K. (2005). Effectiveness of an externally added solid lubricant on the sliding
wear response of a zinc?aluminium alloy, its composite and cast iron. Tribology
Letters, 18(2), 135–143.
Prasad, B. K., Rathod, S., Yadav, M. S., & Modi, O. P. (2010). Effects of Some Solid
Lubricants Suspended in Oil Toward Controlling the Wear Performance of a Cast
Iron. Journal of Tribology, 132(4), 041602.
Rabinowic, E., L.A., D., & Russell, P. G. (1961). A study of abrasive wear under threebody conditions. Wear, 4(5), 345–355.
Rabinowicz, E. (1977). Abrasive wear resistance as a materials test. Lubr. Eng., 33(7),
378–381.
Rahim, E. A., Kamdani, K., & Sharif, S. (2008). Performance Evaluation of Uncoated
Carbide Tool in High Speed Drilling of Ti6Al4V. Journal of Advanced Mechanical
Design, Systems, and Manufacturing, 2(4), 522–531.
Rahman, M., Wong, Y. S., & Zareena, A. R. (2003). Machinability of titanium alloys.
JSME International Journal Series C, 46, 107–115.
Rahman, M., Wong, Y. S., & Zareena, A. R. (2003). Machinability of titanium alloys.
JSME International Journal Series C, 46, 107–115.
Rahmati, B., Sarhan, A. a. D., & Sayuti, M. (2013). Investigating the optimum
molybdenum disulfide (MoS2) nanolubrication parameters in CNC milling of
AL6061-T6 alloy. The International Journal of Advanced Manufacturing
Technology, 70(5-8), 1143–1155.
Ramesh, S., Karunamoorthy, L., & Palanikumar, K. (2008). Fuzzy Modeling and
Analysis of Machining Parameters in Machining Titanium Alloy. Materials and
Manufacturing Processes, 23(4), 439–447.
Reddy, N. S. K., & Rao, P. W. (2006). Experimental investigation to study the effect of
solid lubricants on cutting forces and surface quality in end milling. International
Journal of Machine Tools and Manufacture, 46(2), 189–198.
Rioul, O., & Vetterli, M. (1991). No Title. IEEE Signal Processing Magazine, 8, 14–38.
Sadeghi, M. H., Haddad, M. J., Tawakoli, T., & Emami, M. (2008). Minimal quantity
lubrication-MQL in grinding of Ti–6Al–4V titanium alloy. The International Journal of
Advanced Manufacturing Technology, 44(5-6), 487–500.

205

Sarhan, A. a. D., Sayuti, M., & Hamdi, M. (2012). Reduction of power and lubricant oil
consumption in milling process using a new SiO2 nanolubrication system. The
International Journal of Advanced Manufacturing Technology, 63(5-8), 505–512.
Schrock, D. J., & Kwon, P. (2012). Evidence of Phase Dependent Tool Wear in Ti-6Al4V Turning Experiments Using PCD and Carbide Inserts. In ASME 2012
International Manufacturing Science and Engineering Conference (p. 371). ASME.
Schrock, D. J., Kang, D., Bieler, T. R., & Kwon, P. (2014). Phase Dependent Tool Wear
in Turning Ti-6Al-4V Using Polycrystalline Diamond and Carbide Inserts. Journal of
Manufacturing Science and Engineering, 136(4), 041018.
Semiatin, S. L., Seetharaman, V., & Weiss, I. (1996). The Thermomechanical
Processing of Alpha/Beta Titanium Alloys. JOM, 49(6), 33–40.
Shahan, A. R., & Taheri, A. K. (1993). Adiabatic shear bands in titanium and titanium
alloys: a critical review. Materials & Design, 14(4), 243–250.
Shen, B., Malshe, A. P., & Shih, A. J. (2008). Performance of novel mos 2 nanoparticles
based grinding fluids in minimum quantity lubrication grinding. Transaxtions of
NAMRI/SME, 36, 357–364.
Shen, B., Shih, A. J., & Tung, S. C. (2008). Application of Nanofluids in Minimum
Quantity Lubrication Grinding. Tribology Transactions, 51(6), 730–737.
Sima, M., & Özel, T. (2010). Modified material constitutive models for serrated chip
formation simulations and experimental validation in machining of titanium alloy Ti–
6Al–4V. International Journal of Machine Tools and Manufacture, 50(11), 943–960.
Singh, G., Singh, S., Singh, M., & Kumar, A. (2013). Experimental investigations of
vegetable & mineral oil performance during machining of en-31 steel with minimum
quantity lubrication. International Journal of Research in Engineering and
Technology, 02(06), 1030–1037.
Srikant, R. R., Rao, D. N., Subrahmanyam, M. S., & Vamsi Krishna, P. (2009).
Applicability of cutting fluids with nanoparticle inclusion as coolants in machining.
Proceedings of the Institution of Mechanical Engineers, Part J: Journal of
Engineering Tribology, 223(2), 221–225.
Suh, N. P. (1986). Tribophysics. (Pretice-Hall, Ed.).
Tai, B. L., Dasch, J. M., & Shih, A. J. (2011). Evaluation and Comparison of Lubricant
Properties in Minimum Quantity Lubrication Machining. Machining Science and
Technology, 15(4), 376–391.

206

Tawakoli, T., Hadad, M. J., & Sadeghi, M. H. (2010). Influence of oil mist parameters on
minimum quantity lubrication–MQL grinding process. International Journal of
Machine Tools and Manufacture 50, 50(06), 521–531.
Thepsonthi, T., Hamdi, M., & Mitsui, K. (2009). Investigation into minimal-cutting-fluid
application in high-speed milling of hardened steel using carbide mills. International
Journal of Machine Tools and Manufacture, 49(2), 156–162.
Toh, C. K., & Kanno, S. (2004). Surface integrity effects on turned 6061 and 6061-T6
aluminum alloys. Proc. IMechE,Vol. 225, Part N: J. of Nano Engineering and Nano
Systems, 225(1), 3–16.
Trent, E. M. (1991). Metal cutting. Butterworth-Heinemann Ltd, Oxford.
Ueda, T., Hosokawa, A., & Yamada, K. (2006). Effect of Oil Mist on Tool Temperature in
Cutting. Journal of Manufacturing Science and Engineering, 128(1), 130.
Varadarajan, a. S., Philip, P. K., & Ramamoorthy, B. (2002). Investigations on hard
turning with minimal cutting fluid application (HTMF) and its comparison with dry
and wet turning. International Journal of Machine Tools and Manufacture, 42(2),
193–200.
Venugopal, K. a., Paul, S., & Chattopadhyay, a. B. (2007). Growth of tool wear in
turning of Ti-6Al-4V alloy under cryogenic cooling. Wear, 262(9-10), 1071–1078.
Wakabayashi, T., Inasaki, I., & Suda, S. (2006). Tribological Action and Optimal
Performance: Research Activities Regarding Mql Machining Fluids. Machining
Science and Technology, 10(1), 59–85.
Watanabe, S., Miyake, S., & Murakawa, M. (1991). Tribological properties of cubic,
amorphous and hexagonal boron nitride films. Surface and Coatings Technology,
49(1-3), 406–410.
Weinert, K., Inasaki, I., Sutherland, J. W., & Wakabayashi, T. (2004). Dry Machining
and Minimum Quantity Lubrication. CIRP Annals - Manufacturing Technology,
53(2), 511–537.
Wong, T., Kim, W., & Kwon, P. (2004). Experimental support for a model-based
prediction of tool wear. Wear, 257(7-8), 790–798.
Wright, P. K., & Bagchi, a. (1981). Wear mechanisms that dominate tool-life in
machining. Journal of Applied Metalworking, 1(4), 15–23.
Yang, X., Liu, C. R., & Lafayette, W. (1999). Machining titanium and its alloys.
Machining Science and Technology, 3(1), 107–139.
207

Yukhno, T. ., Vvedensky, Y. ., & Sentyurikhina, L. . (2001). Low temperature
investigations on frictional behaviour and wear resistance of solid lubricant
coatings. Tribology International, 34(4), 293–298.
Zhang, X. P., Shivpuri, R., & Srivastava, a. K. (2014). Role of phase transformation in
chip segmentation during high speed machining of dual phase titanium alloys.
Journal of Materials Processing Technology, 214(12), 3048–3066.
Zhang, X., Wu, S., Wang, H., & Liu, C. R. (2011). Predicting the Effects of Cutting
Parameters and Tool Geometry on Hard Turning Process Using Finite Element
Method. Journal of Manufacturing Science and Engineering, 133(4), 041010.
Zhang, Y. C., Mabrouki, T., Nelias, D., & Gong, Y. D. (2011). Chip formation in
orthogonal cutting considering interface limiting shear stress and damage evolution
based on fracture energy approach. Finite Elements in Analysis and Design, 47(7),
850–863.
Zhu, Q., Qiu, X., & Ma, C. (1998). The oxidation resistance improvement of matrix
graphite of spherical fuel elements by slip-gelation process. Journal of Nuclear
Materials, 254(2-3), 221–225.

208