Johnson-Cook Failure Parameters

JOHNSON-COOK FAILURE PARAMETERS
CALIBRATION OF THE JOHNSON-COOK FAILURE PARAMETERS AS

THE CHIP SEPARATION CRITERION IN THE MODELLING OF THE
ORTHOGONAL METAL CUTTING PROCESS
By KEYAN WANG, B. ENG.
A Thesis Submitted to the School of Graduate Studies in Partial Fulfilment of

Requirements for the Degree Masters of Applied Science
McMaster University Copyright by Keyan Wang, March 2016

i
MASTERS OF APPLIED SCIENCE (2016)
McMaster University
(Mechanical Engineering)
Hamilton Ontario
TITLE: Calibration of the Johnson-Cook Failure Parameters as the Chip Separation

Criterion in the Modelling of the Orthogonal Metal Cutting Process
AUTHOR: Keyan Wang, B. Eng. (McMaster University)
SUPERVISORS: Dr. Eu-Gene Ng
NUMBER OF PAGES: xx,103
ii
ABSTRACT
The finite element analysis (FEA) is a numerical method widely used to predict
the metal-cutting performance in both academic and industrial studies, avoiding the
high expense and time consumption of experimental methods. The problem is how to
calibrate reliable fracture-parameters as chip-separation criterion are implemented into
FEA modelling. This thesis introduces a calibration method of the Johnson-Cook
fracture parameters used in the orthogonal metal cutting modelling with a positive rake
angle for AISI 1045 steel. These fracture parameters were obtained based on a set of
quasi-static tensile tests, with smooth and pre-notched round bars at room temperature
and elevated temperatures. The fracture parameters were validated by low- and highstrain rate simulations corresponding to tensile tests and orthogonal metal-cutting
processes respectively in ABAQUS/Explicit. Compared to literature calibration
methods, this method is simpler, less expensive but valid.
iii
ACKNOWLEDGMENTS
I cannot express enough thanks to all of those who continued support and
encourage me to complete this thesis.
I would like to express my sincere gratitude to my supervisor Dr. Eu-Gene Ng for
his patience, motivation, and immense knowledge. Without his guidance and support, I
could not have imagined a better study for the researching and writing of this thesis.
Most importantly, I deeply appreciate this opportunity given to me by Dr. Eu-Gene Ng
to complete this important research.
I would like to thank the Department of Mechanical Engineering, McMaster
University, for financing and sponsoring this research.
I would like to thank Dr. Micheal Bruhis for his excellent help and guidance during
the tensile tests.
I would like to thank Mr. Doug Culley for his help and guidance for the metal heattreatment and the measurement of metallography.
I would like to thank the members in my research group and my friends: Youssef
Ziada, Keyvan Hosseinkhani, Amy Short, Wanlin Zhang, Chenhui Niu and Si Sun for
their help with research discussions and the experimental work.
Finally, I would like to thank my parents and my friends for their encouragement,
patience and support during the whole of this research process.
iv
TABLE OF CONTENTS
ABSTRACT ............................................................................................................... III
ACKNOWLEDGMENTS ......................................................................................... IV
TABLE OF CONTENTS ............................................................................................ V
LIST OF FIGURES ................................................................................................... IX
LIST OF TABLE CAPTIONS ................................................................................ XV
NOMENCLATURE AND LIST OF SYMBOLS................................................ XVII
1. INTRODUCTION.................................................................................................... 1
1.1 BACKGROUND ....................................................................................................... 1
1.2 OBJECTIVE ........................................................................................................... 4
2. LITERATURE REVIEW ........................................................................................ 5
2.1 DUCTILE FRACTURE .............................................................................................. 5
2.1.1 Definition and Features of Ductile Fracture ................................................ 5
2.1.2 Effect of Stress State on Ductile Fracture ..................................................... 9
2.1.3 Effects of Temperature and Strain Rate on Ductile Fracture ..................... 14
2.2 DUCTILE FRACTURE MODELS............................................................................... 18
2.2.1 Evaluation Studies of Fracture Models....................................................... 19
2.2.2 Johnson-Cook Fracture Model ................................................................... 22
v
2.2.3 Bao-Wierzbicki Fracture Model .................................................................. 23

2.3 CALIBRATION OF THE JOHNSON-COOK FRACTURE MODEL ................................. 24
3. EXPERIMENTAL PROCEDURES ..................................................................... 30
3.1 MATERIALS ......................................................................................................... 30
3.2 TENSILE SPECIMENS ............................................................................................ 30
3.3 TENSILE TEST METHOD ........................................................................................ 33
3.4 ARAMIS SYSTEM .............................................................................................. 34
3.4.1 Specimen Preparation ................................................................................. 34
3.4.2 Test Conditions ............................................................................................ 34
3.4.3 Data Analysis .............................................................................................. 35
3.5 ANNEALING AND METALLOGRAPHY .................................................................... 38
3.6 GEOMETRY MEASUREMENTS............................................................................... 39
4. MODELLING PROCEDURES ............................................................................ 41
4.1 MODELLING PROCEDURES OF TENSILE TESTS..................................................... 41
4.1.1 Mesh Conditions ......................................................................................... 41
4.1.2 Material Properties ..................................................................................... 42
4.1.3 Boundary Conditions .................................................................................. 48
4.2 LAGRANGIAN CUTTING MODELLING PROCEDURES ............................................. 48
4.2.1 Geometry and Mesh Conditions.................................................................. 49
vi
4.2.2 Material Properties ..................................................................................... 50

4.2.3 Boundary Conditions and Temperature Conditions.................................... 51
4.2.4 Friction Conditions ..................................................................................... 52
4.2.5 Simulation Matrix ....................................................................................... 52
5. RESULTS AND DISCUSSION ............................................................................. 53
5.1 VERIFICATION OF THE PUBLISHED JOHNSON-COOK CONSTITUTIVE AND FRACTURE
CONSTANTS ............................................................................................................... 53
5.2 FLOW-CHART FOR THE CALIBRATION AND VERIFICATION OF THE JOHNSON-COOK
CONSTITUTIVE AND FRCTURE CONSTANTS56
5.3 CALIBRATION AND VERIFICATION OF THE JOHNSON-COOK CONSTITUTIVE
CONSTANTS ............................................................................................................... 58
5.3.1 Calibration and Verification of the Constants A, B, n of the Johnson-Cook
Strain Hardening Component ............................................................................. 58
5.3.2 Calibration and Verification of the Constants m of the Johnson-Cook
Temperature Component .................................................................................... 64
5.4 CALIBRATION AND VERIFICATION OF THE JOHNSON-COOK FRACTURE
CONSTANTS 67
5.4.1 Calibration and Verification of the Johnson-Cook Fracture Constants D1,
D2, D3 ................................................................................................................... 68
5.4.2 Verification of the Calibrated Johnson-Cook Fracture Constant D5 .......... 74
vii
5.4.3 Calibration and Verification of the Johnson-Cook Fracture Constant D4 . 79

5.5 EVALUATION
OF THE
JOHNSON-COOK FRACTURE CONSTANTS
WITH
CUTTING
SIMULATIONS ............................................................................................................ 79
5.5.1 Chip Formation and Stress Contour ........................................................... 79
5.5.2 Temperature of the Tool-chip Interface ....................................................... 83
5.5.3 Cutting Force .............................................................................................. 87
5.5.4 Chip Thickness ............................................................................................ 89
5.5.5 Strain to Fracture of the Layer ................................................................... 91
6. CONCLUSION .................................................................................................... 94
7. FUTURE WORKS .............................................................................................. 96
BIBLIOGRAPHY...97
APPENDIX A ...................................................... .................................................... 103
viii
LIST OF FIGURES
Figure 2.1: Schematic representations of the tensile stress-strain behavior of brittle and
ductile materials loaded to fracture. [William, 2007] .................................................... 5
Figure 2.2: (a) Highly ductile fracture in which the specimen necks down to a point.
(b) Moderately ductile fracture after some necking. (c) Brittle fracture without any
plastic deformation. [ William, 2007] ............................................................................ 6
Figure 2.3: Void nucleation, growth, and coalescence in ductile metals: (a) inclusions
in a ductile matrix, (b) void nucleation, (c) void growth, (d) strain localization
between voids, (e) necking between voids, and (f) void coalescence and fracture.
[Anderson, 2005] ......... ................................................................................................ 7
Figure 2.4: Stages in the cup-and-cone fracture (a) Initial necking. (b) Small cavity
formation. (c) Coalescence of cavities to form a crack. (d) Crack propagation. (e)
Final shear fracture at a 45 angle relative to the tensile direction. [Ralls, 1976] ......... 8
Figure 2.5: (a) Cup-and-cone fracture in aluminum. (b) Brittle fracture in a mild
steel ................................................................................................................................ 8
Figure 2.6: Three types of coordinate systems in the space of principal stresses. [Bai,
2009] ..................................................................... ...................................................... 10
Figure 2.7: Dependence of the equivalents strain to fracture on the stress triaxiality.
[Bao, 2004] ................... .............................................................................................. 12
Figure 2.8: Assumed dependence of the equivalent strain to crack formation on the
stress triaxiality of 1045 steel [Bao, 2005] .................................................................. 13
Figure 2.9: A newly postulated 3D asymmetric fracture locus. [Bai, 2008] ................ 14
Figure 2.10: Average shear strains at fracture for the torsion tests. [Johnson, 1985] .. 15
Figure 2.11: Split Hopkinson device. [Autenrieth, 2009] ............................................ 15
Figure 2.12: Effects of strain rate and temperature on the strain to fracture. [Johnson,
1985] ............................................................................................................................ 16
ix
Figure 2.13: Comparison between experimental data (dotted lines) and model results
(solid lines) for the target material of Weldox 460 E steel: (a) Fracture strain vs. Log
strain rate, (b) Fracture strain vs. Temperature. [Borvik, 2001] .................................. 17
Figure 2.14: Dependences of averaged failure strain , = 2 (0 ) and
normalized failure strain for the tensile specimens at different temperatures.

[Autenrieth, 2009] ........................................................................................................ 18
Figure 2.15: Torsion moment Mt in dependence on title angle for stress triaxiality
p/v=0 and different loading rates . [Autenrieth, 2009] ............................................ 18
Figure 2.16: Comparison of prediction of all seven fracture criteria relatively to the set
of 12 test points (plane stress) on 2014-T351 aluminum specimens. [Wierzbicki,
2005] ............................................................................................................................ 21
Figure 2.17: Fracture strain vs. Pressure-stress ratio for isothermal quasi-static
conditions. [Johnson, 1985] ......................................................................................... 25
Figure 2.18: Comparison between experimental data to model results for Weldox 460
E steel. [Borvik, 2001] ................................................................................................. 27
Figure 2.19: Comparison of ductility curves of aluminum. [Wierzbicki, 2005].......... 28
Figure 3.1: Description of the specimen at necking region or pre-notched area ......... 31
Figure 3.2: Un-notched and pre-notched specimens (dimensions mm)....................... 32
Figure 3.3: Tensile fixture with a threaded round bar specimen. (a) Positions of the
specimens, upper grip, and bottom grip; (b) Dimensions of the upper grip; (c)
Dimensions of the bottom grip. ................................................................................... 33
Figure 3.4: Random-mist-ink-pattern used to tensile specimens for ARAMIS
testing. .......................................................................................................................... 34
Figure 3.5: The position of the camera in the Aramis 2D system ................................ 35
Figure 3.6: Example of results from Aramis system for plotting true stress-strain
curve ............................................................................................................................. 36
x
Figure 3.7: Example of the true strain of the center point with localized strain leading
to necking ..................................................................................................................... 37
Figure 3.8: Example of the gage line of the height of the notch, 2b, in 2D strain map
from ARAMIS system ................................................................................................. 38
Figure 3.9: Comparison of the stress-strain curve under different anneal
temperatures ................................................................................................................. 39
Figure 3.10: Example of the measurement of the geometry of specimen before
testing ........................................................................................................................... 40
Figure 3.11: (a) The area of the central cross section of the fracture surface after
testing. (b) The geometry measurements of the specimen after fracture ..................... 40
Figure 4.1: Meshed models of the tensile tests of un-notched and pre-notched
specimens ..................................................................................................................... 42
Figure 4.2: The initial imperfection in the un-notched model ..................................... 42
Figure 4.3: (a) Schematic representation of tensile test data in stress-displacement
space for elastic-plastic materials [ABAQUS, 2003]; (b) The experimental stress vs.
displacement space result ............................................................................................. 47
Figure 4.4: Linear relationship between d and ..................................................... 48
Figure 4.5: The boundary conditions of the tension model ......................................... 48
Figure 4.6: The conditional link element of the cutting work piece in models ........... 49
Figure 4.7: Geometry and mesh conditions for the work piece with three feed rates . 49
Figure 4.8: Geometry and mesh condition of the cutting tool ..................................... 50
Figure 4.9: The boundary conditions of the cutting model .......................................... 51
Figure 4.10: The contact surfaces as interacted pairs in the cutting models ................ 52
Figure 5.1: The modelling performance with historical constants of the Johnson-Cook
models: (a) Von Misses stress contours with constants from [Borkovec, 2008]; (b)
xi
Temperature contours with constants from [Borkovec, 2008]; (c) Von Misses stress
contours with constants from [Varizi, 2010]; (d) Temperature contours with constants
from [Varizi, 2010] ...................................................................................................... 55
Figure 5.2: Comparison of the published Johson-Cook fracture curves from Borkovec
[2008] and Varizi [2010] with the experimental results from Bai [2009] ................... 55
Figure 5.3: (a) The strain of the central point of the specimen from ARAMIS; (b) d
calculated from the strain map from (a). ...................................................................... 59
Figure 5.4: (a) Length of the chord of the necking region 2b; (b) The arrow height of
the necking region h; (c) Radius of the curvature of the necking region R ................. 60
Figure 5.5: Comparison of stress-strain data for the tension test and the data adjusted
by Bridgman correction factor ..................................................................................... 61
Figure 5.6: Comparison of the experimental data and the calibrated results of the
Johnson-Cook constitutive model ................................................................................ 62
Figure 5.7: Comparison of the elongation between tension test with notched radius as
9.525 mm specimen at room temperature and its simulation: (a) initial stage, (b) after
fracture ......................................................................................................................... 64
Figure 5.8: Plot of thermal softening fraction vs. dimensionless temperature ............ 65
Figure 5.9: Comparison of the elongation between tension tests with un-notched
specimen and their simulations: (a) at 150C; (b) at 285C ......................................... 66
Figure 5.10: Comparison of the Johnson-Cook constitutive curves between the
experimental results, the calibrated results, and the results from Jaspers and
Dautzenberg [2002]and Borkovec [2008].................................................................... 67
Figure 5.11: The measurements of the geometry of the original specimens of all kinds
of specimens. (mm) (a) Un-notched specimen; (b) R=9.525(mm) pre-notched
specimen; (c) R=6.35(mm) pre-notched specimen; (d) R=3.175(mm) pre-notched
specimen; (e) R=1.5875(mm) pre-notched specimen; where R is the radius of the
curvature of the notch .................................................................................................. 69
Figure 5.12: The measurements of all kinds of the specimen after fracture. (mm) (a)
Un-notched specimen; (b) R=9.525(mm)pre-notched specimen; (c) R=6.35(mm) prexii
notched specimen; (d) R=3.175(mm) pre-notched specimen; (e) R=1.5875(mm) prenotched specimen; where R is the radius of the curvature of the notch....................... 70
Figure 5.13: Experimental stress triaxiality vs. strain at fracture and the calibrated
Johnson-Cook failure model ........................................................................................ 72
Figure 5.14: Comparison of the deformation and the radius of the cross section area
between tension tests at room temperature with their simulations .............................. 73
Figure 5.15: Plot of the ratio of fracture strain and dimensionless temperature .......... 76
Figure 5.16: Comparison of the experimental and numerical views of the failure
specimens at 150C ...................................................................................................... 77
Figure 5.17: Comparison of the experimental and numerical views of the failure
specimens at 285C ...................................................................................................... 78
Figure 5.18: von Misses stresses contour maps around the cutting layer of the cutting
simulations at step 125/250: d1=0.1 mm, d2=0.2 mm, d3=0.3 mm; v1=150 m/min,
v2=250 m/min, v3=350 m/min .................................................................................... 81
Figure 5.19: Temperature contour maps of the cutting simulations: d1=0.1 mm,
d2=0.2 mm, d3=0.3 mm; v1=150 m/min, v2=250 m/min, v3=350 m/min ................. 84
Figure 5.20: Paths used as the tool-chip interface in the work piece and in the tool ... 86
Figure 5.21: Temperatures along the tool-chip interface in the work piece and in the
tool ............................................................................................................................... 86
Figure 5.22: Comparison of the temperatures at the tool-chip interface with different
feeds and cutting speeds ............................................................................................... 87
Figure 5.23: The nodes used for obtaining the cutting forces ...................................... 88
Figure 5.24: Cutting forces of the cutting simulations ................................................ 88
Figure 5.25: Measurement of the chip thickness of the cutting model with depth of cut
0.2mm, velocity 250m/min .......................................................................................... 89
Figure 5.26: Chip thickness with three feeds and cutting speeds from the tension
xiii
models .......................................................................................................................... 89
Figure 5.27: The strains to fracture of the layer at time 0.0012s under different cutting
conditions ..................................................................................................................... 91
Figure 5.28: The stress triaxiality of the element in the layer of the cutting simulations
under different cutting conditions at time 0.0012s ...................................................... 92
Figure 5.29: The top and bottom elements used for the calculation of the stress
triaxiality of the element in the layer ........................................................................... 93
Figure 5.30: The temperatures of the elements in the layers of the cutting simulations
under different cutting conditions ................................................................................ 93
xiv
LIST OF TABLE CAPTIONS

Table 2.1: Formulas of fracture criteria, where and are the equivalent strain and
stress respectively; is the equivalent strain to fracture; 1 and 2 are the maximum
and intermediate principal tensile stress; m is the hydrostatic stress, n is the hardening
coefficient, =
2 2
1
3 , 1 2 3 are the rate of principle strains. [Bao,
2004b] .......................................................................................................................... 20
Table 3.1: Chemical Composition of AISI 1045 steel (wt%) ...................................... 30
Table 3.2: Dimensions and the stress triaxialities of all five kinds of tensile
specimens ..................................................................................................................... 31
Table 4.1: Physical and mechanical properties of the material used in tension
models .......................................................................................................................... 43
Table 4.2: Some temperature-dependent properties of the material use in tension
models .......................................................................................................................... 43
Table 4.3: The material constants of the Johnson-Cook constrictive model of tension
material used in simulation .......................................................................................... 44
Table 4.4: Fracture constants of the Johnson-Cook fracture model used in the
modelling ..................................................................................................................... 45
Table 4.5: Sample calculation results ........................................................................... 46
Table 4.6: Material properties of the carbide tool ........................................................ 50
Table 4.7: Simulation matrix of cutting models ........................................................... 52
Table 5.1: Published Johnson-Cook constants for AISI 1045 steel. (a) Published
Johnson-Cook constitutive constants; (b) Published Johnson-Cook fracture
constants ....................................................................................................................... 53
Table 5.2: Calibrated constants of the Johnson-Cook constitutive model for strain from
tension tests .................................................................................................................. 62
xv
Table 5.3: Adjusted constants of the Johnson-Cook constitutive model for strain
hardening...................................................................................................................... 63
Table 5.4: Yield stress and the thermal softening fraction of tension tests .................. 65
Table 5.5: Calibrated constants of the Johnson-Cook constitutive models from tension
tests .............................................................................................................................. 67
Table 5.6: Strain at fracture and the stress triaxiality of all five kinds of tensile
specimens. (Dimensions of different types of specimens shown in Table 3.2) ........... 71
Table 5.7: Calibrated D1, D2 and D3 values ................................................................. 72
Table 5.8: Comparison of the experimental and modelling reduction of areas of
tension tests at 25C ..................................................................................................... 74
Table 5.9: Experimental data used to compute Tf and result Tf .............................. 75
Table 5.10: The fracture strains used for obtaining D5 of the Johnson-Cook failure
model from tension test with un-notched specimens at three temperatures ................ 76
Table 5.11: Comparison of the experimental and modelling reduction of areas. (a)
Temperature=150C; (b) Temperature=285C ............................................................ 79
Table 5.12: The cutting conditions used in FE simulations of cutting process ............ 79
Table 5.13: The cutting experimental conditions and results ....................................... 88
Table 5.14: Chip thickness of all cutting simulations .................................................. 90
xvi
NOMENCLATURE AND LIST OF SYMBOLS

2D
3D
a
A0
Aave
Af
AISI
ALE
Atrue
b
Bave
Btrue
C1
C2
CCD
cp
d
D
d
d0
D1
D2
D3
D4
D5
Dcr
df
DIC
E
E
FEA
FEM
: Two dimensional
: Three dimensional
: Radius of the center cross section of the tensile specimen (mm)
: Original central cross section of the tensile specimen
: Average Johnson-Cook material constant adjusted with value (MPa)
: Area of the center cross section of the tensile specimen after fracture (mm2)
: American Iron and Steel Institute
: Arbitrary Lagrangian Eulerian
: Actural Johnson-Cook material constant adjusted with value (MPa)
: (When described the notched tensile specimen as shown in Figure 3.1) Half
of the chord length of the notch of the tensile specimen (mm)
: Average Johnson-Cook material constant adjusted with value (MPa)
: Actural Johnson-Cook material constant adjusted with value (MPa)
: (When used in the function of stress triaxiality and strain to fracture)
Material constant (-)
: Charged-coupled device
: Specific heat capacity (J/kgC)
: (When used in Figure 4.3) overall damage variable (-)
: Accumulated damage (-)
: Diameter of the minimum cross section of the tensile specimen (mm)
: Initial diameter of the center cross section of the tensile specimen (mm)
: (When used in Johnson-Cook fracture equation) Fracture parameter (-)
: Critical accumulated damage (-)
: Diameter of the center cross section of the tensile specimen after fracture
(mm)
: Digital image correlation
: Errors of the cutting force between the cutting experimental result and FEM
(in Figure 5.25)
: Youngs Modulus (MPa)
: Finite element analysis
: Finite element method
xvii
h
[I]
k
KT
L
nave
No.
ntrue
OFHC
p
r
R
R0
RAe
RAm
Rf
[S]
T*
Tc
tc_m
tc_t
Tmelt
to
Troom
Ttf
tu
u
: (When described the notched tensile specimen as shown in Figure 3.1) arrow
height of the notch of the tensile specimen (mm)
: (When used in the deviatoric stress tensor equation) Identity tensor (MPa)
: Thermal Conductivity (W/(mC))
: Thermal softening fraction (-)
: Characteristic length (m)
: Average Johnson-Cook material constant adjusted with value (-)
: Number (-)
: Actural Johnson-Cook material constant adjusted with value (-)
: Oxygen-free high thermal conductivity
: Hydrostatic pressure (MPa)
: (When used in Equation (5.20)) chip thickness ratio (-)
: Radius of the curvature of the profile of the notch of the tensile specimen
(mm)
: Initial radius of the curvature of the profile of the notch of the tensile
specimen (mm)
: Reduction areas from the tension tests (-)
: Reduction areas from the modelling (-)
: Radius of the curvature of the profile of the notch of the tensile specimen
after fracture (mm)
: deviatoric stress tensor (MPa)
: homologous temperature (-)
: Chip thickness after separation (mm)
: Measured chip thickness (mm)
: Theoretical chip thickness (mm)
: Melting temperature ()
: Thickness of the chip prior to chip formation. (mm)
: Room temperature ()
: Transition temperature for failure ()
: Discrete value of the temperature ()
: Discrete value of the homologous temperature (-)
: Uncut chip thickness (mm)
: Displacement space (m)
: Equivalent plastic displacement at fracture (m)
wt%
: Equivalent plastic displacement (m)

: Weight percentage (-)
: Rake angle of the tool ()
: Thermal expansion (C-1106)
xviii

: Increment of the equivalent plastic strain (-)
: (When used in the Johnson-Cook constrictive model) equivalent plastic
strain (-)
: Equivalent strain to fracture (-)
exp
fs
: True fracture strain measured from a smooth specimen test (-)
: Void nucleation strain (-)

: (When used in the Johnson-Cook constrictive model) dimensionless plastic
strain rate (-)
: (When used in the Johnson-Cook fracture model) equivalent strain to
fracture (-)
: Discrete value of the strain rate (s-1)
: Discrete value of the dimensionless strain rate (-)
: Stress triaxiality (-)
: Stress triaxiality of the original specimen (-)
: Stress triaxiality of the specimen after fracture (-)
: Lode angle ()
: Adjusted parameter depends on the strain rate (-)
: Adjusted variable (-)
: Poissons ratio (-)
: Normalized third stress invariant (MPa)
: Density (kg/m3)
: (When used in the Johnson-Cook constrictive model) von Mises flow stress
(MPa)
: Ratio of the average of the three normal stresses to the von Mises equivalent
stress (-)
: Principal stress in the Cartesian coordinate system (MPa)
: Average axial stress at the smallest cross section of the tensile specimen
(MPa)
: Stress corrected with the Bridgman correction factor (MPa)
: (When used in Figure 4.3) undamaged response stress (MPa)
: (When used in the Johnson-Cook fracture model) von Mises equivalent
stress (MPa)
: (When used in the Johnson-Cook fracture model) average of the three
normal stresses (MPa)
: Onset of the damage stress (MPa)
: Discrete value of the stress triaxiality (-)
0
f
*
1
2
3
a
effect
xix
[]
: Stress tensor (MPa)

: Coordinate of the principal stress in the spherical coordinate system (MPa)
: Shear plane angle ()
xx
M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)
1. Introduction
1.1 Background
The finite element method (FEM) is a numerical approach that subdivides a
continuum system to finite simpler elements, and analyzes the element properties
mathematically, assuming the approximate behavior of the entire system. It is widely
applied to predict metal cutting performance such as the wear of the tool and the plastic
deformation of the machined work piece. Klamecki [1973] employed the FEM to model
the 3-dimensional metal cutting process, but only in the initial stages of chip formation.
Usui and Shirakashi [1982] and Iwata et al. [1984] first simulated the chip formation
under steady state cutting by using two dimensional models. Strenkowski and Carroll
[1985] studied the chip separation criteria based on updated Lagrangian modules.
Moreover, Komvopoulos and Erpenbeck [1991] used a distance tolerance criterion to
investigate the chip separation and formation. Lin and Lin [1992] studied the chip
separation, using the argument of strain energy, and analyzed chip geometry, the
residual stresses in the machined surface of the work piece, and the temperature
distributions in the tool and chip and the cutting forces. Ceretti et al. [1996] simulated
the chip separation of the cutting process by deleting elements which reach the critical
value of accumulated damage. In general, four formulations in FEM metal-cutting
simulations are Lagrangian, Eulerian, Arbitrary Lagrangian Eulerian (ALE) and
Smooth Particle Hydrodynamics. Lagrangian has the advantage of more convenient
modelling of the evolution of the chip formation from the incipient stage to the steady
form, in which the mesh follows the material which is attached to the workpiece;
meanwhile, the elements move with the material along the cutting process [Soo, 2007].
Hence, an appropriate chip separation criterion is necessary. Chip separation criteria
include two groups: geometrical and physical. A typical geometrical separation
criterion is the distance tolerance criterion, which is based on a predefined critical
distance between the tool tip and the separation node of the workpiece [Komvopoulos,
1991]. On the other hand, physical separation criteria mean the separation of the node
or element occurs based on a predefined-critical-physical-parameter from the material
properties and cutting conditions, such as strain, stress or strain-energy-density.
Strenkowski and Carroll [1985] used the effective plastic strain criterion, that is the
node removed from the workpiece when its effective plastic strain reaches a predefined
critical value, which was also used by Xie et al. [1998], Liu and Guo [2000] and Hortig
and Svendsen [2007]. In addition, other physical criteria include the strain energy
density criterion by Lin and Lin [1992] Usta [1999] and critical stress criterion by
Huang and Black [1996], Shet and Deng [2000], Mamalis et al. [2001] and Li et al.
[2002]. However, Zhang [1999] evaluated above chip separation criteria and concluded
that a single threshold of separation cannot be used for different cutting conditions,
which means a more feasible and comprehensive criterion needs to be organized for a
reliable simulation of the cutting process.
Furthermore, several models based on accumulative damage mechanics were
proposed as the onset of fracture criteria in the Lagrangian based simulation of metal
2
cutting processes, such as the Cockroft-Latham model [Cockroft, 1968], the Wilkins
model [Wilkins, 1978], and the Johnson-Cook fracture model [Johnson, 1985]. In this
study, the Johnson-Cook fracture model is used. The problem of using the JohnsonCook fracture model is how to acquire and calibrate the fracture constants for the
proposed material, which is AISI 1045 steel in this study. The Johnson-Cook fracture
model expressed the effects of the stress triaxiality, the strain rate, and the temperature
on the strain to fracture respectively in three brackets. The fracture constants relative to
different effective factors were usually calibrated separately. Firstly, to calibrate the
stress triaxiality relative fracture constants D1, D2 and D3, some researchers obtained
the data based on quasi-static tests, such as tensile tests with smooth and notched round
bars, torsion tests, or tests with specially designed specimens [Johnson, 1985], [Borvik,
2001], [Majzoobi, 2011] and [Dzugan, 2013]. As the strain to fracture is influenced not
only by the stress state but also the stress histories, the numerical data of these tests
were also required. Furthermore, the failure parameters were finally calibrated based
on different optimization methods by different researchers. Secondly, the strain-raterelative-fracture-constant D4 is obtained based mostly on the experimental results from
a dynamic test [Johnson, 1985; Borvik, 2001]. Thirdly, D5, as the temperature-relativefracture-constant, is calibrated from the tests with elevated temperatures [Johnson, 1985;
Borvik, 2001]. Vaziri et al. [2010] calibrated all constants in one approach so that the
data from the ALE finite element simulations were calculated by an optimization
method to obtain several sets of fracture constants, and one of the best sets was selected
based on the performance as input in an updated Lagrangian based simulation of metal
3
cutting.
1.2 Objective
The objective of this research is to calibrate reliable fracture constants for the
Johnson-Cook fracture model, which is used as chip separation criteria in the finite
element method simulation of metal cutting for AISI 1045 steel. The sub-objective is
as follows:
1. Design experiments to calibrate the fracture constants of the Johnson-Cook
fracture model with the use of low strain-rate experimental configuration.
2. Build finite element simulations of the corresponding experiments to verify the
obtained fracture constants.
3. Acquire the Johnson-Cook fracture strain-rate coefficient D4 from the metal
cutting simulation and experiment.
4. Perform a parametric metal cutting simulation to verify and identify the
limitation of the approach.
2. Literature Review
2.1 Ductile Fracture
2.1.1 Definition and Features of Ductile Fracture
Ductile fracture, which differs from brittle fracture, is a failure process that the
material sustains in plastic deformation before it separates into pieces due to an imposed
stress at the temperature lower than its melting temperature [Callister, 2007]. Figure 2.1
shows the comparison of the tensile stress-strain curves undergoing with or without a
significant amount of plastic strains loaded to fracture, respectively for ductile and
brittle materials [Callister, 2007]. Ductile fracture surfaces have special features that
present necking regions with rough and irregular surfaces compared to the brittle one,
as shown schematically in Figure 2.2. [Callister, 2007].
Figure 2.1: Schematic representations of the tensile stress-strain behavior of brittle and ductile
materials loaded to fracture. [Callister, 2007]
(a)
(b)
(c)
Figure 2.2: (a) Highly ductile fracture in which the specimen necks down to a point. (b)
Moderately ductile fracture after some necking. (c) Brittle fracture without any plastic
deformation. [Callister, 2007]
Three principal stages were observed in the ductile fracture: nucleation, growth
and coalescence of voids as shown in Figure 2.3. Firstly, micro-voids nucleate at
inclusions or second-phase particle when the interface stress elevates to the enough
level to rupture the inter-facial bonds between the particle or inclusion and the matrix
[Argon, 1975]. Secondly, the voids grow until the matrix attains to the plastic limit-load
state by means of plastic strain and hydro-static stress [Thomason, 1998]. Thirdly, the
inter-voids matrix happens to plastic limit-load failure across the sheet with strain
localization between voids. Then it begins to neck down internally between the adjacent
voids [Cottrell, 1959]. The voids further coalesce and the inter-void matrix cracks with
a knife-edge fracture surface [Thomason, 1998].
(a)
(b)
(c)
(d)
(e)
(f)
Figure 2.3: Void nucleation, growth, and coalescence in ductile metals: (a) inclusions in a ductile
matrix, (b) void nucleation, (c) void growth, (d) strain localization between voids, (e) necking
between voids, and (f) void coalescence and fracture. [Anderson, 2005]
A cup and cone shape was a typical ductile fracture feature obtained under uniaxial tensile tests. Figure 2.4 shows the cup and cone fracture process. After necking,
some small micro-voids first form in the interior of the material as shown in Figure
2.4(b) [Callister, 2007]. The micro-voids then enlarge and coalesce to form a crack as
shown in Figure 2.4(c), and the long axis of this elliptical crack perpendicular to the
stress direction [Callister, 2007]. Finally, the crack propagates rapidly by the outer ring
of the specimen, and shear deformation bands are formed at a 45-degree angle from the
tensile stress axis as shown in Figure 2.4(d) [Anderson, 2005]. The central interior
region of the fracture surface appears fibrous and irregular as shown in Figure 2.4(e)
and Figure 2.4(a) [Anderson, 2005]. This characteristic surface is named cup and cone
fracture because one of the fracture surfaces looked similar to a cup and the adjacent
one looked similar to a cone as shown in Figure 2.5(a).
(b)
(a)
(d)
(c)
(e)
Figure 2.4: Stages in the cup-and-cone fracture (a) Initial necking. (b) Small cavity formation. (c)
Coalescence of cavities to form a crack. (d) Crack propagation. (e) Final shear fracture at a 45
angle relative to the tensile direction. [Ralls, 1976]
(b)
(a)
Figure 2.5: (a) Cup-and-cone fracture in aluminum. (b) Brittle fracture in a mild steel.
Ductile fracture plays an important role in a successful finite element simulation

of machining studied by Bil et al. [2004] and Liu and Guo [2000]. The criteria for
fracture initiation and the evolution of fracture are necessary requirements in the finite
element modelling. Two key factors on the ductile fracture behaviors are the loading
conditions and the material itself. Moreover, the initiation of ductile fracture is affected
by the stress state, temperature, and the strain rate, all of which will be discussed later.
8
2.1.2 Effect of Stress State on Ductile Fracture

The stress state is the most important factor to control the ductile fracture initiation.
The stress state is usually defined based on three invariants of the stress tensor [], as
shown in Equations (2.1), (2.2) and (2.3) respectively [Bai, 2009].
1
1
p m tr([ ]) ( 1 2 3 )
3
3
(2.1)
3
1
[S ] : [S ]
[( 1 2 )2 ( 2 3 )2 ( 3 1 )2 ]
2
2
1
(2.2)
1
9
27
27
r ( [ S ] [ S ] : [ S ]) 3 [ det([ S ])] 3 [ ( 1 m )( 2 m )( 3 m )]3 (2.3)
2
2
2
where [S] is the deviatoric stress tensor as detailed in Equation (2.4)
[ S ] [ ] p[ I ]
(2.4)
where [I] is the identity tensor and 1, 2 and 3 are the principal stresses, assuming
1 2 3 [Bai, 2009]. The stress triaxiality parameter , as shown in Equation (2.5),

as a dimensionless hydrostatic pressure, is defined by Bai et al. [2009].
p m
(2.5)
where m is the average of the three normal stresses, is the von Mises equivalent
stress. The Lode angle is expressed by the relationship to the normalized third stress
invariant by Bai et al. [2009], as detailed in Equation (2.6)
r
q
( )3 cos(3 )
(2.6)
The range of is 1 1 , for the range of the Lode angle is 0 3 [Bai, 2009].
Figure 2.6 shows that the stress state of a point consists of three coordinate systems:
Cartesian coordinate system (1, 2, 3), cylindrical coordinate system (m, , ) and
9
spherical coordinate system ( f , , ) [Bai, 2009]. The relationship between

coordinate and the stress triaxiality is shown in Equation (2.7) [Bai, 2009].
m
2
c tan
3
(2.7)
Moreover, the Lode angle can be obtained by Equation (2.8) [Bai, 2009].
ar cos
(2.8)
Figure 2.6: Three types of coordinate systems in the space of principal stresses. [Bai, 2009]
Bridgman [1952] observed that the ductility of material is highly sensitive to

pressure [Bridgman, 1952]. Using the Bridgman piston-cylinder type hydrostatic
pressure system, different steels or Fe-C materials were investigated, and the strain to
fracture is examined increasing with the higher pressure by Bridgman [1952] and
Davidson and Ansell [1968].
Furthermore, McClintock [1968] and Rice and Tracey [1969] inferred an
expression of the relationship between the stress triaxiality and the strain to fracture by
an analysis of the void growth. The function of f ( ) is shown in Equation (2.9).
10
f f ( ) C1e C
2
(2.9)
where C1 and C2 are material constants, and it is typically found that C2 1.5 . The
studies of the void growth models like Equation (2.7) and the extensions are considered
as the foundation of the modern ductile fracture mechanics [Bai, 2009]. Depending on
these fracture mechanics, the criteria for fracture initiation are proved strongly by
means of stress triaxiality by void growth mechanics modelling [Roy, 1981; Sun, 1991].
Similarly, the same conclusions are obtained based on other mechanics such as the
continuum damage mechanics [Rosa, 2001] and are considered initiation and
propagation-toughness-mechanics [Sun, 1991; Henry, 1997].
Additionally, depending on numbers of tests, more specific relationships between
the stress triaxiality and the strain to fracture for different materials have been studied.
Hancock and Mackenzie [1976] used tensile tests with pre-notched round specimens to
explore a formula of f ( ) as listed in Equation (2.10).

3
2
f n exp( )
(2.10)
where n is the void nucleation strain and is material constant. This fracture model
has been further extended by combining the effect of temperature and strain rate upon
strain to fracture by Johnson and Cook [1985], which is the most popular fracture model
used in the finite element of machining.
Specifically, the effect of stress triaxiality upon the strain to fracture for some
materials, such as aluminum, is not monotonically in a wide range of stress triaxiality
[Bao, 2003; Bao, 2004a]. Hence, a fracture locus describing different relationships of
the stress triaxiality with a strain to fracture in a different range of the stress triaxiality
11
is designed and studied from different tests [Bao, 2004a] as shown in Figure 2.7.
Figure 2.7: Dependence of the equivalents strain to fracture on the stress triaxiality. [Bao,
2004a]
The experimental results dominated by different fracture mechanisms are shown

1
with three branches: 0 , 0 0.4 , 0.4 [Bao, 2004a]. In the range
3
1
0 , results are obtained from the upsetting tests based on shear fracture
3
mechanism, and the fitted curve of the experimental results is presented in Equation
(2.11), which is developed based on the study by Wierzbicki and Werner [1998]. The
other two fitted curves are obtained by simple parabolic fit, and they are presented in
Equation (2.12) and Equation (2.13), respectively [Bao, 2004a].
f 0.1225 ( ) 0.46
1
3
1
for 0
3
(2.11)
f 1.9 2 0.18 0.21
for 0 0.4
(2.12)
f 0.15 1
for 0.4 0.95
(2.13)
It has been mentioned that the stress triaxiality related to the ductile fracture must
be bigger than -0.33, which is so called cut-off value, for no material failure was
12
observed when the stress triaxiality was smaller than -0.33 from the results of the
upsetting tests [Bao, 2004a] and the tension tests [Bridgman, 1952].
However, this kind of fracture locus only contributes to a given material. The other
materials have different fracture loci. For example, the fracture locus of AISI 1045 steel
is monotonical as shown in Figure 2.8 [Bao, 2005]. The function of strain to fracture
relates to the pressure for AISI 1045 steel based on the experimental results from Kao
et al. [1990] as shown in Equation (2.14).
f 1.25 exp(
p
)
1000
(2.14)
Figure 2.8: Assumed dependence of the equivalent strain to crack formation on the stress
triaxiality of 1045 steel [Bao, 2005].
In addition, the effect of the stress state, consisting of both the pressure and the
Lode angle, upon the ductile fracture is studied and described by a 3D (threedimensional) fracture locus as shown in Figure 2.9 [Bai, 2008].
13
Figure 2.9: A newly postulated 3D asymmetric fracture locus. [Bai, 2008]
Equation (2.15) presents the 3D fracture locus of the strain to fracture affected by
both stress triaxiality and Lode parameter [Bai, 2008].
1
2
1
2
f ( , ) [ ( D1e D D5e D ) D3e D ] ( D1e D D5e D ) D3e D (2.15)

2
2.1.3 Effects of Temperature and Strain Rate on Ductile Fracture

In metal cutting, the materials will deform under high strains, high strain rates, and
high temperatures conditions. Hence, the effects of temperature and strain rate on
ductile fracture are studied by Johnson and Cook [1985]. Figure 2.10 shows the
experimental stress-strain curves for three different kinds of materials from the SplitHopkinson tension-bar tests at elevated temperatures [Johnson, 1985]. The SplitHopkinson tension-bar test is a dynamic test with an imposed stress wave on the
specimens, and the device structure is shown in Figure 2.11.
14
Figure 2.10: Average shear strains at fracture for the torsion tests. [Johnson, 1985]
Figure 2.11: Split Hopkinson device. [Autenrieth, 2009]
Figure 2.12 shows the analysis effects of the strain and temperature on the fracture
strain. The fitted straight lines were calculated by the least squares method to cover
the average of temperatures for each test from the beginning to where the fracture
occurs [Johnson, 1985]. Hence, it is shown that the ductility of material increases with
higher temperatures. The effect of the strain rate on ductile fracture is also detailed in
Figure 2.12 when at T*=0where T* is the non-dimensional temperature. The ratio of
the strain to fracture at high-strain-rate to low-strain-rate is greater than 1. In other
words, a higher stain rate will increase the strain to fracture of material.
15
Figure 2.12: Effects of strain rate and temperature on the strain to fracture. [Johnson, 1985]
However, the effects of the temperature and strain rate on ductile fracture are not
uniform for different materials. In other words, the influence of these two factors on
ductility strongly depends on the material itself. Figure 2.13 shows totally different
relationships between the strain-rate or the temperature to the strain to fracture for
Weldox 460 E steel compared to the results from Johnson and Cook [1985]. As shown
in Figure 2.13, the strain to fracture decreases with the strain rate increasing, and the
temperature sensitivity of the fracture strain is not monotonous and is neglected.
Especially, the effect of the temperature upon the strain to fracture only starts from
300C up to higher temperatures [Borvik, 2001].
16
(a)
(b)
Figure 2.13: Comparison between experimental data (dotted lines) and model results (solid lines)
for the target material of Weldox 460 E steel: (a) Fracture strain vs. Log strain rate, (b) Fracture
strain vs. Temperature. [Borvik, 2001]
Besides, a similar phenomenon appears in the study for AISI 1045 steel as shown
in Figure 2.14, so that the temperature has almost no effect on the strain to fracture up
to a so called transition temperature for failure Ttf =176C [Autenrieth, 2009]. Moreover,
the expression of this phenomenon in a failure model is created as shown in Equation
(2.16) [Autenrith, 2009]. Furthermore, the effect of strain rate on failure strain for AISI
1045 steel is also studied by Autenrith et al. [2009]. As shown in Figure 2.15, the change
of the strain to fracture for different strain rates is too small and it is neglected by
Autenrith et al. [2009].
17
f 3 (T ) H ( H 1) (1 ( T Ttf ) H 4 ) H 5
3
3
Tmelt Ttf
forT Ttf
forT Ttf
(2.16)
Figure 2.14: Dependences of averaged failure strain , = ( ) and normalized failure
strain for the tensile specimens at different temperatures. [ Autenrieth, 2009]
Figure 2.15: Torsion moment Mt in dependence on title angle for stress triaxiality p/v=0 and
different loading rates . [Autenrieth, 2009]
2.2 Ductile fracture models

Depending on the above studies of the fracture mechanisms, the fracture models
are formulated to predict initial crack or failure, in particular, are implemented into the
commercial codes in the finite element simulations. Some researchers evaluated
18
different fracture models as guides for choosing a suitable fracture model for different
loading conditions and different materials [Bao, 2004b; Wierzbicki, 2005; Teng, 2006;
Liu, 2014; Duan, 2009].
2.2.1 Evaluation Studies of Fracture Models
Bao and Wierzibcki [2004b] compared and evaluated nine fracture models as the
ductile-crack-formation-criteria for 2024-T351 aluminum alloy. The upsetting tests and
round smooth and pre-notched tension tests and their corresponding finite element
simulations were studied with these nine fracture models [Bao, 2004b]. The material
fracture was in a range of negative and low to high stress triaxiality respectively as
shown in Table 2.1 [Bao, 2004b]. They concluded that none of the fracture models
produce a compatible result for the whole range of the considered stress triaxiality [Bao,
2004b]. The reason is the crack forms based on two totally different mechanisms
imposed by the different range of stress triaxiality: void growth and shear decohesion
[Bao, 2004b]. But some fracture models are suitable for narrowing the range of stress
triaxiality. Concretely, the hydrostatic stress criterion and general Rice-Tracey model
are good for a range of low to high stress triaxiality from the smooth and pre-notched
tension tests, while the Cockcroft-Latham-Oh criterion works well for a range of a
negative stress triaxiality from the upsetting tests [Bao, 2004b].
Furthermore, seven fracture models were evaluated for optimal fracture criteria in
the finite element simulations (also for 2024-T351 aluminum alloy) based on a set of
plane-stress tests and their corresponding finite element simulations [Wierzbicki, 2005].
19
Table 2.1: Formulas of fracture criteria, where and

are the equivalent strain and stress
respectively; is the equivalent strain to fracture; 1 and 2 are the maximum and intermediate
principal tensile stress; m is the hydrostatic stress, n is the hardening coefficient, =
are the rate of principle strains. [Bao, 2004b]
Figure 2.16 shows the comparison results of seven fracture models, which showed
the Xue-Wierzbicki and CrachFEM models are good according to all tests [Wierzbicki,
2005]. The maximum stress fracture model fits well for all tests except the round bar
tension test [Wierzbicki, 2005]. The Wierzbicki criterion only predicts well either in the
range of stress triaxiality of
1 2 or 1 1 , but not in both

3
3
3
3
[Wierzbicki, 2005]. And the other three fracture criteria are only useful in narrow ranges
of the stress triaxiality [Wierzbicki, 2005]. In addition, it emphasized that all the
evaluations of the fracture models are only made for the material used. In other words,
the choice of the fracture model is strongly dependent on the material itself [Wierzbicki,
20
2005].
Figure 2.16: Comparison of prediction of all seven fracture criteria relatively to the set of 12 test
points (plane stress) on 2014-T351 aluminum specimens. [Wierzbicki, 2005]
Moreover, as a continuation of the project, the applicability of six fracture models

used in the dynamic loading condition, a high velocity perforation, were evaluated for
Weldox 460 E steel and 2024-T351 aluminum alloy by applying a finite element
simulation of a high-velocity perforation test [Teng, 2006]. It concluded that the BaoWeizbicki fracture model was good at predicting the fracture phenomenon in
perforation and that the Johnson-Cook fracture model was described as able to predict
though not for the shear/compression test while the other four models were not able to
predict this high velocity perforation test [Teng, 2006]. Additionally, a similar
conclusion is obtained from another evaluation of six fracture models in the finite
element simulation of metal cutting processes for 2024-T351 aluminum [Liu, 2014].
However, for AISI 1045 steel, the Johnson-Cook fracture model was proven able to
predict chip separation in the finite element simulation of high speed cutting [Duan,
21
2009]. As the expected study is used in the finite element simulation of metal cutting,
two optimum fracture models, the Johnson-Cook fracture model and the BaoWierzibcki fracture model, evaluated by [Liu, 2014] and [Duan, 2009] will be further
discussed later.
2.2.2 Johnson-Cook Fracture Model
The Johnson-Cook fracture model is performed including effects of stress
triaxiality, strain rate, and temperature on the strain to fracture respectively, as shown
in Equation (2.17) [Johnson, 1985].
f [ D1 D 2 exp D3 * ][1 D 4 ln * ][1 D5T * ]
(2.17)
where D1, D2, D3, D4, and D5 are material parameters, * is the ratio of the average of
the three normal stresses to the von Mises equivalent stress assumed * 1.5 , * is
the dimensionless strain rate, and T* is the homologous temperature that
T*
T Troom
where Tmelt and Troom are melting and room temperatures respectively.
Tmelt Troom
The first set of brackets of the formula expresses the effect of stress triaxiality on strain
fracture monotonously, which follows the expression by [Hancock, 1976], and is traced
back to the study of the void growth by McClintock [1968], Rice and Tracy [1969].
This model is based on damage-accumulated criteria assumed in a linear way as
shown in Equation (2.18) [Johnson, 1985].
(2.18)
where is the increment of the equivalent plastic strain, and f is the equivalent strain
to fracture. The fracture occurs when D = 1.0.
22
The Johnson-Cook fracture model is applied widely because of the simplicity of

formulation, the ease of calibration and a number of the material parameters provided
by Johnson and Holmquist [1989]. However, Johnson and Cook [1985] only
determined the positive range of the stress triaxiality based on some tensile tests and
shear tests, and no small or negative value of stress triaxiality is expressed. For steel,
Johnson and Cook [1985] even neglected the data of torsion testing in the fitting curve.
Perhaps one function is not able to express two different fracture mechanisms over the
range of stress triaxiality from negative to positive: shear decohesion and void growth.
It is proved that ductility of aluminum is not a monotonic function of stress triaxiality
[Wierzbicki, 2005]. To effectively apply the Johnson-Cook fracture model, researchers
extended the model in different ways. Liu et al. [2014] proved that Johnson-Cook
fracture model can be used as damage initiation coupled with damage evolution in metal
cutting simulations. Moreover, the damage evolution combines the two different
fracture modes effects. Bao and Wierzbicki [2004b] extended the Johnson-Cook
fracture model in their fracture locus so that the part with high stress triaxiality is similar
to Johnson-Cook fracture model while the other parts are expressed in different ways.
2.2.3 Bao-Wierzbicki Fracture Model
Bao and Wierzbicki [2004b] performed an empirical fracture model as the ductility
of material based on the stress triaxiality. They found a cut-off value where the stress
triaxiality equals -1/3, which means that no fail occurs if the stress triaxiality is less
than this cut-off value. They also expressed the fracture locus in three branches for
two different fracture mechanisms: void growth and shear decohesion as shown in the
23
Equation (2.19).
0.46
H 1
0
.
1225

fh 1.9( H ) 2 0.18 H 0.21
H
0.15
1 H
0
3

0 H 0. 4

0.4 H 0.95

(2.19)
This model performed more comprehensive fracture behavior over a wide range
of stress triaxiality that almost cover all kinds of fracture process. However, it is still
required to extend with the strain rate and temperature effects on the ductile fracture.
Hence, Liu et al. [2014] extended the Bao-Wierzibcki fracture model with the parts
expressed strain rate and temperature effects in Johnson-Cook fracture model in
Equation (2.20)
pl
T Troom
1 D5
f fh 1 D4 ln
T
room
0
melt
(2.20)
2.3 Calibration of the Johnson-Cook Fracture Model

The most important problem for implementing the Johnson-Cook fracture model
into metal cutting simulation or other finite element analysis is to calibrate fracture
constants for any proposed material.
Johnson and Cook [1985] calibrated the material constants for three kinds of
materials: OFHC copper, Armco iron, and 4340 steel. Firstly, the fracture constants D1,
D2, D3 are calibrated based on quasi-static tensile tests with smooth and notched round
bars and torsion tests as shown in Figure 2.17, and then adjusted the results to * 1.0
using following obtained D4. The strains to fracture were calculated by
24
f 2 ln(d 0 / d f ) , where d0 and df are the diameters of the initial and fractured cross
section of the necking surface, which were measured by using a photograph of the
specimens. The stress triaxiality used for fitting the curve is the average value of the
stress triaxiality along the loading process. This stress triaxiality is obtained based on
the stress history data of the point in the center of the necking zone from the numerical
simulation of these quasi-static tensile tests. Specifically, the torsion data of 4340 steel
was ignored, for it is so far from the proposed curve. D4 and D5 were calibrated based
on the ratio of the tensile Hopkinson bar test to quasi-static tensile tests as shown in
Figure 2.17. The tensile Hopkinson bar test is a dynamic test with high strain rates and
with elevated temperatures. However, the temperature effects were expressed with a
low degree of accuracy here by Johnson and Cook [1985].
Figure 2.17: Fracture strain vs. Pressure-stress ratio for isothermal quasi-static conditions.
[Johnson, 1985]
Borvik et al. [2001] determined the fracture parameters D1, D2, D3 for Weldox 460
25
E steel based on quasi-static tensile tests with smooth and notched round bars without
the use of numerical simulations, and using the method of least squares to minimize the
residual of the fracture strain model, which is expressed in Equation (2.21).
( f )
N N N
1 1 1
exp
f
( , , T ) ( D1 D2 exp( D3 ))(1 ) (1 D5T ) (2.21)

*
* D4
where * , , * , T , T* are the discrete value of the stress triaxiality, the strain rate,
the dimensionless strain rate, the temperature and the homologous temperature
respectively. The material data at * 0 is estimated based on Equation (2.22)
[Lemaitre, 1992].
2
3
*
2 1
f exp
fs [ (1 ) 3(1 2 )( max ) ]
(2.22)
where exp
fs is the true fracture strain measured from a smooth specimen test. However,
the extent of the curve with the results of D1, D2, D3 to the range * 0 only depended
on the hydrostatic tests. D4 was calibrated based on data from dynamic tensile tests at
room temperature and minimized the residual of Equation (2.21). D5 was calibrated
based on data from the quasi-static tensile tests at elevated temperatures with smooth
specimens, and D5 0 at range from 0C to 300C as shown in Figure 2.13(b). Figure
2.13(a), 2.13(b), and Figure 2.18 show the comparison between the experimental data
and the model results based on the calculated material parameters, respectively.
26
Figure 2.18: Comparison between experimental data to model results for Weldox 460 E steel.
[Borvik, 2001]
Tomasz Wierzbicki et al. [2005] only calibrated the D1, D2, D3 for 2024-T351
aluminum alloy based on the quasi-static round smooth and notched tensile tests as
shown in Figure 2.19. However, the calibrated curve only provides an upper bound
curve for material ductility, and the fracture curve of the same material from Johnson
and Holmquist [1989] provides a low bound curve for material ductility as shown in
Figure 2.19. That means different results were obtained dependent on different
experiments. The Johnson-Cook fracture model is only reliable in a certain range of
stress triaxiality, and the material parameters are calibrated by using the experiments
covering this range of stress triaxiality.
27
Figure 2.19: Comparison of ductility curves of aluminum. [Wierzbicki, 2005]
All above calibrations are dependent on experimental data, and some combined
with the finite element simulated data. However, Vaziri et al. [2010] provided a new
calibration for AISI 1045 steel that used the data from ALE based finite element
simulation of metal cutting. The best set of material parameters were obtained by a
nonlinear least-squares optimization procedure that the objective function is detailed in
Equation (2.23) [Vaziri, 2010].
4
LSE min fi 2 ( x )
x
(2.23)
i 1
where x= [D1, D2, D3, D4, D5] as a vector, and i expresses the different cutting conditions
tests 1, 2, 3 and 4 respectively. The Johnson-Cook fracture model is based on the
assumption that the elements fail when the accumulate damage value is D = Dc r= 1.0,
as shown in Equation (2.24).
f i ( x ) 1. 0
f ,i
1
d p ,i
[ D1 D2 exp( D3 )][1 D4 ln i* ][1 D 5Ti * ]
*
i
(2.24)
where fi is numerically integrated based on the trapezoidal method. Then the optimized
28
sets of fracture parameters were obtained by using a MAPLE code to the program. The
best set of fracture parameters is identified based on an updated Lagrangian simulation
of metal cutting by comparing the predicted cutting forces and thrust forces to
experimental results.
Majzoobi and Dehgolan [2011] calibrated the material parameters of the JohnsonCook constitutive model and fracture model together. A method combining the
experimental, numerical, and optimization technique is used. D1, D2, and D3 were
obtained based on a quasi-static tensile test with a notched bar. D4 was obtained based
on a dynamic test with a notched bar by using a high rate testing device Flying wedge.
However, no test of temperature effects on the fracture is studied here.
Dzugan et al. [2013] calibrated the fracture parameters based on the fracture locus
f ( , ) for the typical steel used in the nuclear power plant industry. They used three
kinds of tests including tensile tests with notched round bars, punch tests, and a
specially designed test with a specimen with a double curvature. All the tests are quasistatic tests at room temperature. Hence, only D1, D2, and D3 were calibrated.
29
3. Experimental Procedures
3.1 Materials
The workpiece material investigated was AISI 1045 steel with chemical
composition detailed in Table 3.1.The materials were cold rolled and annealed with a
tensile strength of 758 MPa. The annealing process carried out was 923 K held at two
hours followed by cooling down to room temperature within two hours in the oven.
After the annealing process, the tensile strength was reduced to about 586 MPa.
Table 3.1: Chemical Composition of AISI 1045 steel (wt%).
Materials
Si
Mn
AISI 1045 steel 0.42~0.50 0.17~0.37 0.50~0.80
Ni
Cu
Cr
0.30
0.25
0.25
3.2 Tensile Specimens

The tensile specimens were un-notched and four unique geometries were prenotched round bars. These specimens with five kinds of geometries will induce different
stress triaxiality states according to Bridgmans [1952] analysis as shown in Equation
(3.1).
1
= 3 + ln(1 + 2)
(3.1)
where is the stress triaxiality state, R is the radius of curvature of the profile of the
neck, and a is the radius of its minimum cross section [Bridgman, 1952]. The
description of the specimen after necking or with pre-notched is shown in Figure 3.1.
30
Figure 3.1: Description of the specimen at the necking region or pre-notched area.
The four different pre-notched specimens were designed with the same a and
different R to get four different ratios, a/2R, and the corresponding stress triaxiality
states as shown in Table 3.2: Dimensions and the stress triaxialities of all five kinds of
tensile specimens. The ends of the specimens shoulders were the threaded shoulders
which eliminate slippage of the specimen during the test comparing to smooth
shoulders with wedge grips.
Table 3.2: Dimensions and the stress triaxialities of all five kinds of tensile specimens.
Type of
specimen
Radius of the center

crosses section
Local radius of a
notch
Ratio
Stress
triaxiality
a (mm)
R (mm)
a/2R
Un-notched
6.35
0.33
Notched 1
3.81
9.525
0.2
0.52
Notched 2
3.81
6.35
0.3
0.60
Notched 3
3.81
3.175
0.6
0.80
Notched 4
3.81
1.5875
1.2
1.12
The dimensions and the geometries of all five un-notched and pre-notched round
bars are shown in Figure 3.2.
31
Un-notched
Notched No.1
Notched No.2
Notched No.3
Notched No.4
Figure 3.2: Un-notched and pre-notched specimens (dimensions in mm).
32
3.3 Tensile test method

The tensile tests were performed at room and elevated temperatures (150C and
285C). An Instron SFL2702 Environmental Chamber was used on the MTS Series 810
hydraulic testing machine. The load cell system of the MTS testing machine was chosen
at 250 kN. All the tests were performed with the same crosshead velocity of 0.016
mm/sec to ensure quasi-static conditions. Deformation behavior was viewed through
the window of the environmental chamber by the CCD (Charge-coupled device)
cameras. The digital image correlation (DIC) method was used to analyze the
deformation patterns within gauge section with ARAMIS 6.1 software.
The tensile tests were performed at 25C, 150 C, and 285 C. The high
temperature tests were warmed up to the set value 10 minutes prior to testing.
The tensile specimen is a threaded round bar fixed with two grips. The positions
of the tensile specimen, upper grip, and bottom grip, are shown in Figure 3.3. The
design of the grips is explained in APPENDIX A.
Figure 3.3: Positions of the specimens, upper grip, and bottom grip.
33
3.4 ARAMIS System

The deformation histories of the specimen throughout the tests were recorded with
the ARAMIS system. This system performed the in-situation strain maps and analyzed
the deformation of the specimens from beginning to fracture.
3.4.1 Specimen Preparation
The specimens were printed prior to testing as the preparation for ARAMIS based
testing. The surface of the specimen was sprayed with a fine mist of black and white
ink with a 0.5 mm nozzle, which deforms along with the specimen. The ink was chosen
to withstand temperatures up to 650C, to prevent peeling off during the high
temperature tests. The ink pattern was a high-contrast-random-marking, as shown in
Figure 3.4, to determine the 2D or 3D coordinates of the surface of the specimen during
the testing, and to get the fine accuracy of the deformation measurements.
Figure 3.4: Random-mist-ink-pattern used to tensile specimens for ARAMIS testing.
3.4.2 Test Conditions

In the ARAMIS 2D system, one CCD camera with a resolution of 1280 x 1024
pixels was used. The position of the camera is shown in Figure 3.5: The camera was
focused on the vertical center line of the specimen (Figure 3.5 (a)), and the focus
distance was around 217 mm (Figure 3.5(b)). Optimal illumination of the cylinder
34
surface was reached using two lamps, to avoid constant brightness on the center line of
the specimens surface. The fast-image recording was used with the upper limit of the
ARAMIS system, defined as 1000 images for each test. The first image processing steps
defined the macro-image facets. These facets were tracked in each successive image
with a sub-pixel accuracy of 0.02 pixels (0.04 mm per pixel). The facet size was 1313
pixels (with a spatial resolution of 0.04 mm per pixel) with a step size of 11 pixels. In
some cases, the facet size was 911 or 1113 pixels with a step size of 9 pixels to reach
the optimal calculation. The images were calibrated with the diameter of the specimens
6.35 mm before the calculation of the strain map.
(b)
(a)
Figure 3.5: The position of the camera in the Aramis 2D system.
3.4.3 Data Analysis

The projects were computed following the guidelines in the ARAMIS user manual.
To obtain the true stress-strain curve, a point with a localized strain in the center of the
specimen was selected. First, a section was taken along the center line of the specimen
on the 2D strain map as shown in Figure 3.6(a). The true strain plot of the section at all
stages before the fracture was shown in Figure 3.6(b). Figure 3.6 shows the localization
35
of the specimen leading to the necking under the tensile tests. The point in the center
with this strain localization was selected.
(a) 2D strain map with a section line to obtain data at different tensile displacements.
(b) True strain plot of the section line with all stages before fracture.
Figure 3.6: Example of results from Aramis system for plotting true stress-strain curve.
Figure 3.7(a) shows the true strain of the selected point through all stages from
ARAMIS system. The data between stages 360 to 402 were unstable because of the
error in the calculation of the grey values of the facet in the center area, which came
from the large deformation after necking. The data were adjusted to be a smoother curve
36
as shown in Figure 3.7(b). This adjusted strain epsilon Y (solid line) was used to plot
the true stress-strain curve. Also, this strain was used to calculate the processing
diameter of the center cross section area, d.
(a) Strain of the selected center point from the Aramis system
0.8
Local strain of the center point in the necking

section
X-strain
Y strain
Strain
0.6
0.4
0.2
0.0
0.00
100.00
200.00
300.00
400.00
-0.2
-0.4
Time
(b)Strain of the selected center point adjusted to smooth curve depending on the results
from Aramis system
Figure 3.7: Example of the true strain of the center point with localized strain leading to necking.
To obtain the elongation of the tensile specimens, a line (Line 1) was taken along
the vertical center line at the height of the notch, 2b, as shown in Figure 3.8. Also, this
length, 2b, was used to calculate the radius of the notch, R.
37
(a)
Line 1 defined as the height of the notch at the 2D strain map.
(b)
The length of Line 1(yellow) through all strain stages.
Figure 3.8: Example of the gage line of the height of the notch, 2b, in 2D strain map from
ARAMIS system.
3.5 Annealing and Metallography

To soften the materials, annealing was used at different temperatures from 450C
to 650C. Comparison of the yield stresses with different anneal conditions is shown in
Figure 3.9. These plots show the true stress-strain curve adjusted with the power law,
where the small circle shows the initiation of necking. The optimal anneal condition
was selected as heating the specimen for two hours at 650C and after a cooling down
to room temperature within two hours in the oven.
38
Figure 3.9: Comparison of the stress-strain curve under different anneal temperatures.
3.6 Geometry Measurements

The fracture properties of the materials were characterized by the strain to fracture
and stress triaxialities, and both of these values could be obtained from the dimensions
of the geometries of the specimens. The measurements of the geometries of the
specimens used the results from the ARAMIS system and the microscope before testing,
as shown in Figure 3.10 and Figure 3.11, respectively. The diameter d0 and the area Af
of the center cross section of the specimen before testing and after fracture were
measured to compute the strain to fracture. The diameter df of the center cross-section
of the specimen and the radius Rf of the notch after fracture were measured to compute
the stress triaxialities, and also the reductions of areas were obtained depending on these
measurements.
39
Figure 3.10: Example of the measurement of the geometry of specimen before testing.
(a)
(b)
Figure 3.11: (a) The area of the central cross section of the fracture surface after testing; (b)
The geometry measurements of the specimen after fracture.
40
4. Modelling Procedures
4.1 Modelling Procedures of Tensile Tests
To verify the method and the calibration of the obtained constants of the JohnsonCook constitutive and fracture models, the tension tests were simulated by using
ABAQUS/Explicit version 6.10. A dynamic coupled temperature-displacement
analysis was used.
4.1.1 Mesh Conditions
Due to the symmetric nature of the specimens, the models are built as
axisymmetric models as a quarter of the specimens under tension configuration. The
element module CAX4RT is used, which is a coupled temperature-displacement
element with four bilinear nodes and the ability to reduce integration with hourglass
control. The mesh conditions for all the tension tests with different notched type are
shown in Figure 4.1. Un-notched, R0=9.525 mm, R0=6.35 mm, R0=3.175 mm, and
R0=1.5875 mm specimen meshes were generated and comprised of 1640, 490, 490, 460,
412 elements, respectively. The bottom elements of all models are meshed finer, for
considering the fracture performance. The characteristic lengths of these elements were
0.2245 mm, 0.172 mm, 0.172 mm, 0.176 mm, 0.176 mm, respectively.
In a model with un-notched specimens, an initial imperfection is needed to get the
necking performance. The initial imperfection used as 0.1% of the initial radius of the
cross-section area is shown in Figure 4.2. The mesh conditions and element types are
41
converged by using the force-displacement data and the reductions of area compared to
the experimental data.
R0=9.525
R0=6.35
R0=3.175
R0=1.5875
Figure 4.1: Meshed models of the tensile tests of un-notched and pre-notched specimens. (mm)
Figure 4.2: The initial imperfection in the un-notched model.
4.1.2 Material Properties

All the tension models are built using the uniformed units with m-kg-s. Some of
the material properties of the AISI 1045 steel with annealing at 650C are shown in
Table 4.1. Some other properties are temperature dependent data as shown in Table 4.2.
42
Table 4.1: Physical and mechanical properties of the material used in tension models.
Density (kg/m3)
7850
Young's modulus (Pa) 2.05e11
Poissons Ratio
0.29
Specific Heat (J/kg C)
486
Inelastic Heat Fraction
0.9
Table 4.2: Some temperature-dependent properties of the material use in tension models.
Temperature Thermal conductivity Expansion coefficient

(C)
(W/m C)
(m/m C)
0
52
20
1.01e-5
100
51
200
1.2e-5
300
46
400
1.3e-5
500
38
600
1.53e-5
700
30
1000
27
The criterion of plastic property used is Johnson-Cook constrictive model as
shown in Equation (4.1) [Johnson, 1983].
= [A + B ][1 + ][1 ]
(4.1)
where is the von Mises flow stress, is the equivalent plastic strain, = /0
is
the dimensionless plastic strain rate, and 0 = 1.0 1 , T* is the dimensionless

temperature, and = ( )/( ) and is the melting
temperature of the material, is the room temperature [Johnson, 1983].
The material constants needed for this model are obtained by fitting the
experimental data as shown in Table 4.3.
43
Table 4.3: The material constants of the Johnson-Cook constrictive model of tension material used
in the simulation.
A
(Pa)
B
(Pa)
n
(-)
m
(-)
615.8e6 667.7e6 0.255 1.078
Melting
temperature
Tmelt
(C)
1350
Transition
temperature
Troom
(C)
25
Epsilon
C
dot zero
(-)
(-)
0.0134
1
The criterion of the damage property is built as the damage initiation and evolution
criteria in ABAQUS. The damage initiation criterion used was the Johnson-Cook
fracture model as shown in Equation (4.2) [Johnson, 1985].
f = [D1 + D2 exp(D3 )][1 + D4 ln ][1 + D5 T ]
(4.2)
where is the equivalent strain to fracture, and the dimensionless pressure-stress

ratio (stress triaxiality) is defined as = m /
, where is the average of the three
normal stresses, and is the von Mises equivalent stress. The dimensionless strain
rate, , and homologous temperature, , are identical to those used in the JohnsonCook constitutive model of Equation 4.1. 1 , 2 , 3 , 4 , and 5 are material
dependent fracture constants, which were acquired by fitting the experimental results
as detailed in Table 4.4. Note that the sign of the constant D3 should be changed from
negative to positive, for the definition of the stress triaxiality in ABAQUS is opposite
to the definition in the Johnson-Cook fracture model. The data in Table 4.4 were those
used in the modelling using ABAQUS software.
Table 4.4: Fracture constants of the Johnson-Cook fracture model used in the modelling.
D1
D2
D3
D4
D5
0.04 1.03 1.39 0.002 0.46
Melting
temperature
(C)
1350
Transition
temperature
(C)
25
Reference
Strain rate
1
The damage evolution criterion used is the onset of the displacement of fracture.
44
The softening of the yield stress was assumed to be linear. The maximum displacement
is calculated with the following steps.
Step 1: Determine the characteristic length L.
The definition of the characteristic length as shown in Equation (4.3) depends on
the element geometry and formulation; it is a typical length of a line across an element
for a first-order element [ABAQUS, 2003]. The size of the smallest element is 1.33e-4
by 1.51e-4, so L is obtained as shown in Equation (4.3).
L = (1.33 4 )2 + (1.51 4 )2 = 2.01 4 ()
(4.3)
Step 2: Acquire the true stress vs. displacement space as shown in Figure 4.3(b).
Using Equation (4.4), the undamaged part of this curve is transferred from the
experimental true stress-true strain curve. The damaged part of this curve will be
discussed in step 4.
u = L
(4.4)
Figure 4.3(a) shows the representation of the stress-displacement space curve,

where u is the displacement space, d is the overall damage variable, and 0 is the
onset of the damage stress, and E is the Youngs modulus, and
is the undamaged
response stress, and is the equivalent plastic displacement, and is the
equivalent plastic displacement at fracture, and line 1 is the elastic property along the
onset of the damage.
Step 3: Determine the equivalent plastic displacement of fracture, as shown in
Equation (4.5),
45
= 0
where = = (2.01 4) 0.6 = 1.2 4() ,
(4.5)
0 is the onset of the
damage on the line 1 shown in Figure 4.3 (a). Assume the expression of line 1 using
Equation (4.6).
(u) = +
where
2.0511
2.014
(4.6)
= 1.0215 , C is a constant.
The point when it starts necking is in line 1, as shown in Figure 4.3(a). Take the
coordinate value of this point, (0.0982.01e-4, 646.8e6), into the expression of the line,
to get the value of C. Table 4.5 shows a sample calculation.
Table 4.5: Sample calculation results.
0 (m)
(m)
C
1.94510 1.907 5 1.009e-4
Step 4: Acquire the damaged stress-displacement curve.
The damaged stress-displacement space curve is calculated with Equation (4.7):
= (1 d)
(4.7)
where is the experimental results, as shown in Figure 4.3(b), d is obtained from

the linear relationship with the equivalent plastic displacement, , as shown in
Figure 4.4.
Hence, Equation (4.8) and (4.9) detailed the function for d and , respectively.
d=
(4.8)
= 1.009e4 = 9.9113
= (1
46
(4.9)
= (1 9.9113 )
= [1 9.9113 ( 0 )]
= [1 9.9113 ( 0 )]
= [1 9.9113 (2.01 4 1.907 5)]
= (1.189 1.992 )
where and
are from the experimental result, then the damaged curve is obtained
as shown in Figure 4.3(b).
(a)
(b)
Figure 4.3: (a) Schematic representation of tensile test data in stress-displacement space for
elastic-plastic materials [ABAQUS lecture]; (b) The experimental stress vs. displacement space
result.
.
Figure 4.4: Linear relationship between d and
4.1.3 Boundary Conditions

As shown in Figure 4.5, the left side and the bottom of the model are fixed, as
these regions were constrained by the other parts of the specimen as a quarter of the
47
specimen was modeled. The tensile load was set at 4 m/s on the top edge of the model.
The initial temperatures modeled were 25 C, 150 C, and 285 C.
Figure 4.5: The boundary conditions of the tension model.
4.2 Lagrangian Orthogonal Cutting Modelling Procedures

To validate the constants for Johnson-Cook constitutive and failure models, an
orthogonal cutting is simulated by using ABAQUS/Explicit version 6.10. A dynamic
coupled temperature-displacement analysis was used. To study the material properties
of cutting a work piece, the Lagrangian method was used. In addition, the conditional
link element is built as shown in Figure 4.6, and the cutting tool was assumed to be
elastic.
Figure 4.6: The conditional link element of the cutting work piece in models.
48
4.2.1 Geometry and Mesh Conditions

Due to plane strain assumption, the modelling is built as a 2D simulation. Figure
4.7 shows the geometry and mesh conditions of the models of the cutting work piece
with three different feed rates. All the models had a cutting length of 10 mm, where the
width of cut is 3 mm and the height of the conditional link element is 2 m. The feed
rates as shown are 0.1 mm, 0.2 mm and 0.3 mm.
The element type used in all models of the cutting work piece is CPE4RT, which
is the element including four bilinear nodes with the ability to reduce integration and
hourglass control using coupled temperature displacement module. The characteristic
length of the element in the cutting layer is 40.05 m, and the one in the uncut chip is
40.557 m.
Figure 4.7: Geometry and mesh conditions for the work piece with three feed rates.
Figure 4.8 shows the geometry of the model of the cutting tool. The rake angle is
5 deg and the clearance angel is 10 deg, assuming that the tool cutting edge is sharp
without the nose radius. The number of elements and nodes in the tool are 689 and 742
respectively. The cutting tool is set up as an elastic body, for the studied object is the
material property of the cutting work piece. Additionally, the wear of tool assumed is
49
not considered.
Figure 4.8: Geometry and mesh condition of the cutting tool.
4.2.2 Material Properties

The material properties used in the cutting work piece are built separately into two
kinds, one where damage property is used with the cutting layer as the main studied
object, whereas the other one is not. The material used for the work piece is the same
as the one used for tensile models, which are shown in Table 4.3 and Table 4.4.
The material used for the cutting tool is a carbide tool [Kalhori, 2001], and its
properties are shown in Table 4.6.
Table 4.6: Material properties of the carbide tool [Kalhori, 2001].
Youngs Modulus, E (GPa)

Poissons ratio, (-)
Density, (kg/m3)
Specific heat capacity, cp (J/kgC)
560
0.22
14500
220
5.4(20 C)
5.3(200 C)
Thermal expansion, (C-1106)
5.4(400 C)
5.6(600 C)
34(20 C)
38(250 C)
42(500 C)
Thermal Conductivity, k (W/(mC))
45(750 C)
47(1000 C)
49(1250 C)
50
4.2.3 Boundary Conditions and Temperature Conditions
Figure 4.9: The boundary conditions of the cutting model.
Figure 4.9 shows the boundary conditions of the work piece and the cutting tool
in the cutting models. The part under the cutting layer of the work piece is fixed. The
cutting tool is fixed in a vertical direction and given a velocity to push left in a
horizontal direction that is parallel to the original upper surface of the work piece. The
velocities used in the cutting models are 150 m/min, 250 m/min, and 350 m/min. The
initial temperature for both the work piece and the cutting tool is 25C.
4.2.4 Friction Conditions
Figure 4.10: The contact surfaces as interacted pairs in the cutting models.
Figure 4.10 shows the contact surfaces used in the cutting models. Heat is
generated from the sliding friction in the tool and chip interface. The Coulomb friction
law is applied, and the friction coefficient is 0.22. All the other contact pairs are
51
assumed with a friction coefficient equaling 0.

4.2.5 Simulation Matrix
The simulation matrix allows the verification of the Johnson-Cook failure criteria
and the performance of the chip formation, cutting forces, temperatures, stresses, and
strains as shown in Table 4.7.
Table 4.7: Simulation matrix of cutting models.
Cutting Speed
(m/s)
Feed rate (mm)
0.1
0.2
0.3
150
Simulation 1
Simulation 4
Simulation 7
52
250
Simulation 2
Simulation 5
Simulation 8
350
Simulation 3
Simulation 6
Simulation 9
5. Results and Discussion

5.1 Verification of the Published Johnson-Cook Constitutive
and Fracture Constants
The objective of this phase of the research was to employ published constants of
the Johnson-Cook constitutive and fracture models to simulate the orthogonal cutting
process. Tables 5.1 (a) and 5.1(b) show the published constants for Johnson-Cook
constitutive and fracture models respectively by Borkovec [2008], Jaspers and
Dautzenberg [2002], and Varizi et al. [2010].
Table 5.1: Published Johnson-Cook constants for AISI 1045 steel.
(a) Published Johnson-Cook constitutive constants.
Published Constitutive Constants

[Borkovec, 2008]
[Jaspers, 2002]
(b)
A
B
n
C
M
(MPa) (MPa)
(-)
(-)
(-)
375.0 552.0 0.457 0.020 1.400
553.1 600.8 0.234 0.0134 1.000
Published Johnson-Cook fracture constants.
Published Fracture Constants D1

D2
D3
D4
D5
[Borkovec, 2008]
0.25 4.38 -2.68 0.002 0.61
[Varizi, 2010]
0.05 4.22 -2.73 0.0018 0.55
The orthogonal-cutting finite-element-model is based on the Lagrangian
formulation method. The fixed cutting parameters used were 250 m/min cutting speed,
0.2 mm un-deformed thickness, +5 deg rake angle, and 3 mm width of cut. The
coefficient of friction was assumed to be 0.22, and the tool was assumed to be infinitely
sharp and have no cutting-edge-radius. Two simulations were carried out: one based on
Borkovecs [2008] constitutive and fracture constants, and the other with Jaspers and
Dautzenbergs [2002] constitutive together with Varizi et al. [2010] fracture constants.
Figures 5.1(a) and 5.1(b) show the Von Mises and temperature contour plots
53
respectively when using fracture constants published by Borkovec [2008]. The

simulation with the Borkovecs [2008] fracture constants terminated at 4% of the total
simulation time. The termination was due to the entanglement of the workpiece
elements with the cutting edge. This phenomenon was very likely because the fracture
strain determined by Borkovec [2008] was too high. Therefore, the chip could not
separate from the workpiece. A similar trend was observed when using the fracture
constant determined by Varizi et al. [2010], as shown in Figure 5.1(c) and 5.1(d).
However, the simulation with Varizi et al. [2010] fracture constants was terminated at
41% of the total simulation time. Element entanglement was also observed from Figure
5.1(c) and 5.1(d). That Varizi et al. [2010] fracture constants could simulate longer
cutting time was likely due to a substantially lower D1 value when compared to those
published by Borkovec [2008]. The preliminary results detailed in Figure 5.1(a) to 5.1(d)
concluded that the fracture constants published by Borkovec [2008] and Varizi et al.
[2010] cannot be used to simulate orthogonal cutting. From the cutting model with
Varizis [2010] fracture constants, the predicted hydrostatic pressure and Von Mises
stress were -342 MPa and 1110 MPa, respectively. Therefore, the triaxiality ratio was
equal to 0.31. The fracture strain computed with Varizi et al. [2010] fracture constants
required a strain of 1.9 to induce fracture.
54
(a)
(b)
(c)
(d)
Figure 5.1: The modelling performance with historical constants of the Johnson-Cook models: (a)
Von Misses stress contours with constants from [Borkovec, 2008]; (b) Temperature contours with constants
from [Borkovec, 2008]; (c) Von Misses stress contours with constants from [Varizi, 2010]; (d) Temperature
contours with constants from [Varizi, 2010].
However, experimental work published by Bai et al. [2009] showed in Figure 5.2
that the fracture strain measured was between 0 and 0.5. Therefore, the values
determined by Borkovec [2008] and Varizi et al. [2010] were too large.
Figure 5.2: Comparison of the published Johnson-Cook fracture curves from Borkovec
[2008] and Varizi et al. [2010] with the experimental results from Bai et al. [2009].
55
The stress triaxiality magnitudes along the conditional link elements were in the
range of 0.3 to 0.5 during metal cutting simulation. With the range of stress triaxiality
values mentioned earlier, uniaxial tensile tests have to be performed to acquire the
Johnson-Cook fracture constant with a different temperature at low strain rates.
Before performing experiments to determine Johnson-Cook fracture constants,

uniaxial tensile tests have to be carried out to determine the Johnson-Cook strain
hardening component constants and to compare the constants with those acquired by
Borkovec [2008] and Jaspers and Dautzenberg [2002].
5.2 Flow-chart for the Calibration and Verification of the

Johnson-Cook Constitutive and Fracture Constants
The flow-chart for the calibration and the verification of the whole Johnson-Cook
constitutive and fracture constants are shown below.
56
Start
Calibration of the Johnson-Cook constitutive constants
A, B, n, and m by using the tensile tests.
No
Does the elongation of the tensile simulation using the
obtained A, B, n and m agree with the tensile test?
Yes
Calibration of the Johnson-Cook fracture constants
D1, D2, D3 by the tensile tests at room temperature.
No
Does the reduction of area of the tensile
simulation using the obtained D1, D2, D3 agree
with the tensile tests at room temperature?
Yes
Calibration of the Johnson-Cook fracture constant
No
D5 by the tensile tests at three temperatures.
No
Does the reduction of area of the tensile
simulation using the obtained D5 agree with the
tensile tests at elevated temperature?
Yes
Calibration of the Johnson-Cook fracture
constant D4 by the cutting process simulation.
Do these obtained Johnson-Cook constitutive and

fracture constants work in the cutting simulations?
Yes
The Johnson-Cook constitutive and fracture
constants were calibrated and verified.
End
57
5.3 Calibration
and
Verification
of
the
Johnson-Cook
Constitutive Constants
The constants of Johnson-Cook constitutive model for the material are unique.
These constants can be different for the same materials with different heat treatments.
Hence, the constants of the material used in this research should be determined first.
The Johnson-Cook constitutive model is shown in Equation (4.1). In the Johnson-Cook
constitutive equation, there are three brackets which express the strain effect with A, B,
n, the strain rate effect with C, and the temperature effect with m. In this research,
calibrated constants of the Johnson-Cook constitutive model are obtained from uniaxial
tensile test configurations. To verify the constants of the Johnson-Cook constitutive
model, the comparison of the deformation and elongation is made between the
simulations and experiments.
5.3.1 Calibration and Verification of the Constants A, B, n of the JohnsonCook Strain Hardening Component
The Johnson-Cook hardening term is shown in Equation (5.1), where = 1.0
and = 0 of room temperature.
= A + B
(5.1)
To obtain the constants A, B, and n, stress-strain data are acquired from tension
test with the un-notched specimen at room temperature. The equivalent tension stresses
are computed by using the Bridgman correction factor as shown in Equation (3.1)
[Bridgman, 1952]. The diameter of the central cross section d is computed using
58
Equation (5.2), where the local strain map of the central point in the specimen acquired
from the ARAMIS system is shown in Figure 5.3.
0
d=
(5.2)
2.718
(a)
(b)
Figure 5.3: (a) The strain of the central point of the specimen from ARAMIS;
(b) d calculated from the strain map from (a).
The radius of curvature of the profile of the neck is R calculated using Equation
(5.3). The length of the chord, 2b, is measured as the length of the necking region
(Line 1) in the 2D strain map from ARAMIS report. The arrow height,h =
(0 )
2
, is
calculated from the original diameter of the cross section of the specimen, D0, and d as
shown in Figure 5.3. Figure 5.4 shows the data 2b, h, and R during the experiment.
59
R=
2 +2
2
(5.3)
(a)
(b)
(c)
Figure 5.4: (a) Length of the chord of the necking region 2b; (b) The arrow height of the
necking region h; (c) Radius of the curvature of the necking region R.
60
Then, the true stress, a, and the stress corrected with the Bridgman correction
factor [Bridgman, 1952], effect, are computed from the ARAMIS report using Equation
(3.1), as shown in Figure 5.5, which showed the true stress and the Bridgman corrected
effective stress.
Figure 5.5: Comparison of stress-strain data for the tension test and the data adjusted by the
Bridgman correction factor.
Firstly, the constants of Equation (5.1) in the Johnson-Cook constitutive model are
computed depending on the stress-strain data with the Bridgman correction factor, as
shown in Figure 5.5. A is the yield stress; B and n are obtained depending on the strain
hardening curve by using a sum of squares. The comparison of the experimental stressstrain data and the calibrated results of the Johnson-Cook constitutive model are shown
in Figure 5.6.
61
Figure 5.6: Comparison of the experimental data and the calibrated results of the JohnsonCook constitutive model.
The constants of the Johnson-Cook constitutive model of the material for tension
test are shown in Table 5.2.
Table 5.2: Calibrated constants of the Johnson-Cook constitutive model for the strain from tension
tests.
A (MPa) B (MPa) n (-)

614.2
599.4
0.36
These constants are obtained by using the stress from the tensile tests for =
0.000484, and constants A, B, and n must be adjusted for = 1.0 using Equation
(5.4).
= 1 + Cln
(5.4)
Hence, the stress of the tensile test can be expressed using Equation (5.5).
= A + B = ( + )
(5.5)
where A, B, n are the constants obtained from the tensile test, while Atrue, Btrue, ntrue are
the actual true constants that were adjusted by using value. Hence, the real constants
of the Johnson-Cook constitutive model are expressed using Equation (5.6).
62
= , = , =
(5.6)
The value of is calculated by Equation (5.4), where C=0.0134 from Jaspers

[2002], = 0.000484 from tensile test. So =0.8977, and Atrue=684.2 MPa,
Btrue=667.7 MPa, that are calculated by Equation (5.6). Considering the shear stress
flow of the material, the average flow stress of the tensile and torsion tests is used.
Hence, the calibrated constants are the average constants adjusted by an assumed 10
percent difference by using the tension data. So assuming the adjusted variable is =
0.9, the average constants Aave, Bave, nave of the Johnson-Cook constitutive model is
shown in Equation (5.7).
= , = , = +
(5.7)
Therefore, Table 5.3 shows the final results as the constants of the Johnson-Cook
constitutive model for strain hardening.
Table 5.3: Adjusted constants of the Johnson-Cook constitutive model for strain hardening.
Aave (MPa) Bave (MPa) nave (-)

615.8
667.7
0.255
To verify the constants A, B, n in the Johnson-Cook constitutive model, a

comparison of a tensile test and its simulation is used.
The tensile test used is the test
with pre-notch-radius of 9.525 mm specimen at room temperature and low strain rate.
Figure 5.7 shows the comparison of the deformation between this experimental
geometry by the camera and its computed geometry from strain contours of the
simulation. Figure 5.7 shows the deformation of the specimen at the initial stage and
the end of the test, respectively. The left parts of Figure 5.7(a) and 5.7(b) are the actual
63
photos of the specimen from the experiment, and the right parts are from the simulation.
The gage in red represents the experimental result, and the gage in green represents the
computed result. The gage length and the reduction in the cross-sectional area were
similar between both simulation and experiment. That result means the simulation
agrees with the experimental result, which validates the constants A, B, n in the
Johnson-Cook constitutive model.
(a)
(b)
Figure 5.7: Comparison of the elongation between tension tests with a notched radius as
9.525 mm specimen at room temperature and its simulation: (a) initial stage, (b) after fracture.
5.3.2 Calibration and Verification of the Constants m of the Johnson-Cook

Temperature Component
The third bracket of Equation (3.1) respects the thermal effect on the stress.
Assuming 0 is the stress with 0 = 1.0 1 at room temperature = 0 , the
Equation (3.1) could express using Equation (5.8).
= 0 [1 ]
A thermal softening fraction KT is defined with Equation (5.9) and (5.10).
64
(5.8)
KT =
yield stress at elevated temperature

yield stress at room temperature
y0
K T = 1 T m
(5.9)
(5.10)
To obtain the constant m from Equation (5.10), the yield stress-dimensionless

temperature data are considered from tension tests with un-notched specimens at three
temperatures. The experimental data are shown in Table 5.4. As expressed in Equation
(5.10), constant m is calculated with experimental data in Table 5.4 fitting the least
squares method by using the solver analysis in Microsoft office Excel, and the value
of m was found to be 1.078. The experimental data and the fitting line are shown in
Figure 5.8.
Table 5.4: Yield stress and the thermal softening fraction of tension tests.
T* (-) y (MPa) KT (-)

0
583.38
1
0.0943 533.52 0.915
0.1962 485.30 0.832
Figure 5.8: Plot of the thermal softening fraction vs. dimensionless temperature.
65
To verify the constants m in the Johnson-Cook constitutive model, the tension test
with un-notched specimens at 150C and 285C was carried out and compared with the
simulation with the exact conditions. Figure 5.9 (a) and (b) show the comparison
between experiment and simulated results with tests carried out at 150C and 285C
respectively. The gage length and the reduction in the cross-sectional area were similar
to both experiment and simulation. Therefore, the m constant is valid.
(a)
(b)
Figure 5.9: Comparison of the elongation between tension tests with un-notched specimens
and their simulations: (a) at 150C; (b) at 285C.
Figure 5.10 shows the experimental plastic behavior of the AISI 1045 steel
performed in this research and compared with the plasticity constants acquired by
Borkovec [2008] and Jaspers and Dautzenberg [2002]. The experimental plastic
behavior was carried out at room temperature and at a low strain rate. The Jaspers and
Dautzenberg [2002] constants agreed better with the experimental results when
compared to Borkovecs [2008] constants. Therefore, this result shows that it is valid
to apply the Jasper and Dautzenberg [2002] constitutive constant to the current research.
66
Table 5.5 shows the calibrated constants acquired from the experimental stress-strain
curve.
Figure 5.10: Comparison of the Johnson-Cook constitutive curves between the experimental
results, the calibrated results and the results from Jaspers and Dautzenberg [2002]and Borkovec
[2008].
Table 5.5: Calibrated constants of the Johnson-Cook constitutive models from tension tests.
Constants of the Johnson-Cook

constitutive model
Calibration results
A
(MPa)
615.8
B
(MPa)
667.7
C (-)
n (-) m (-)
0.0134 0.255 1.078
5.4 Calibration and Verification of the Johnson-Cook

Fracture Constants
Since there are three brackets in the Johnson-Cook fracture formula, Equation
(4.2), that expresses the stress triaxiality effect with D1, D2, D3, the stress rate effect
with D4, and the temperature effect with D5, the calibration, and verification of these
constants are conducted respectively. To verify the Johnson-Cook fracture constants,
67
the comparison of the deformation and reductions of areas is made between the
simulations and experiments carried out under tension configuration.
5.4.1 Calibration and Verification of the Johnson-Cook Fracture Constants
D1, D2, D3
First, the effect of the stress triaxiality is considered. When = 1.0 and T =
0, the strain at fracture expressed with the Johnson-Cook fracture model is shown in
Equation (5.11).
f = D1 + D2 expD3
(5.11)
where , as the strain at fracture, is computed from the area of original central cross
section of specimen A0 and the area of central cross section of specimen after fracture
Af, as shown in Equation (5.12).
A
f = ln ( A0 ) = 2ln( d0 )
f
(5.12)
The diameters of the minimum central cross sections were measured using a
microscope. Figure 5.11 showed the measurements of the original size of the specimens
from the tension tests with one un-notched and four different notched specimens under
room temperature conditions.
68
(a)
(b)
(c)
(d)
(e)
Figure 5.11: The measurements of the geometry of the original specimens of all kinds of
specimens. (mm) (a) Un-notched specimen; (b) R=9.525(mm) pre-notched specimen; (c)
R=6.35(mm) pre-notched specimen; (d) R=3.175(mm) pre-notched specimen; (e) R=1.5875(mm)
pre-notched specimen; where R is the radius of the curvature of the notch.
After tensile specimens were fractured, the parts of the specimens are matched
back and measured by using a microscope. Figure 5.12 shows the measurements of the
geometry of all five kinds of the specimen after tension tests under room temperature.
69
(a)
(b)
(c)
(d)
(e)
Figure 5.12: The measurements of all kinds of the specimen after fracture. (mm)
(a) Un-notched specimen; (b) R=9.525(mm) pre-notched specimen; (c) R=6.35(mm) prenotched specimen; (d) R=3.175(mm) pre-notched specimen; (e) R=1.5875(mm) prenotched specimen; where R is the radius of the curvature of the notch.
To get D1, D2, D3 of the Johnson-Cook failure model, both of the stress triaxialities
for the original specimens, 0, and the specimens fracture after tension tests, f, are
required. All these stress triaxialities are computed depending on the measurements by
70
using a microscope as shown in Figure 5.13 and 5.14. The experimental data of f, 0,
and f are shown in Table 5.6.
Table 5.6: Strain to fracture and the stress triaxiality of all five kinds of tensile specimens.
(Dimensions of different types of specimens shown in Table 3.2)
Type of
specimen
Un-notched
Un-notched
Notched 1
R0=9.525mm
Notched 2
R0=6.35mm
Notched 3
R0=3.175mm
Notched 4
R0=1.5875mm
Strain at
fracture,
0.69
0.66
Original stress
triaxiality, 0
0.33
0.33
Fracture Stress
triaxiality, f
0.50
0.50
0.66
0.43
0.41
0.50
0.47
0.44
0.45
0.57
0.53
0.39
0.82
0.70
Figure 5.13 shows the results of the experimental data used for the Johnson-Cook
failure model, and the calibrated curve of the Johnson-Cook failure model fitted the
least squares method. The calibrated curve was obtained by using the solver analysis
in Microsoft office Excel. Table 5.7 shows the D1, D2, D3 values.
Figure 5.13: Experimental stress triaxiality vs. strain at fracture and the calibrated JohnsonCook failure model.
71
Table 5.7: Calibrated D1, D2, and D3 values.
D1 D2
D3
0.04 1.03 -1.39
To verify the constants D1, D2, D3 of the Johnson-Cook failure formula, the tension
test with un-notched and four different notched specimens at room temperature and
corresponding simulations are used to compare with experimental results. Figure 5.14
shows the comparison between the deformation geometries acquired from experiments
and simulations. Figure 5.14 shows the deformations of these specimens after fracture
of these tensile processes. The left column of the Figure 5.14 shows the actual photos
of the specimen from the experiment, and the right column is from the simulation.
Comparing with the length of the radius of the cross section of the fracture surface, the
results are almost the same. That means the simulation agrees with the experimental
result, which validates the constants D1, D2, D3 in the Johnson-Cook failure model.
72
(a) Un-notched specimen
(b) Specimen with notch 1 (R0=9.525mm)
(c) Specimen with notch 2 (R0=6.35mm)
(d) Specimen with notch 3 (R0=3.175mm)
(e) Specimen with notch 4 (R0=1.5875mm)
Figure 5.14: Comparison of the deformation and the radius of the cross section area between
tension tests at room temperature with their simulations.
Additionally, the reduction areas of the tension tests at room temperature and the
corresponding simulations are compared, as shown in Table 5.8. RAe (%) means the
reduction areas from the tension tests, and RAm (%) is the one from the simulations.
Errors (%) =
100%, the errors of the models, are all smaller than 5 %.
That means the models are in good agreement with the tension tests, and the constants
D1, D2, D3 of the Johnson-Cook fracture model were valid.
73
Table 5.8: Comparison of the experimental and modelling reduction of areas of tension tests at
25C.
Specimen
Type
RAe (%)
RAm (%)
Errors (%)
Unnotched
48.5
47.2
-2.7
Notched 1
Notched 2
Notched 3
Notched 4
R0=9.525mm R0=6.35mm R0=3.175mm R0=1.5875mm
48.5
39.4
36.4
32.6
46.5
40.0
37.2
32.1
-4.1
1.5
2.3
-1.6
5.4.2 Calibration and Verification of the Johnson-Cook Fracture Constant

D5
Considering the effect of the temperature on the strain at fracture, we assume a
strain at fracture defined at = 1.0 and T = 0 ,f0 , and the used strain rate is also
= 1.0. Then the strain at fracture expressed in Johnson-Cook failure model is shown
in Equation (5.13)
f = f0 [1 + D5 T ]
(5.13)
where T* is the dimensionless temperature. To get the Johnson-Cook fracture constant

D5, the Equation (5.13) could be changed to Equation (5.14).
f
D5 =
where the ratio is
f
f0
f0
(5.14)
Fracture strain 0,=1.0

. The fracture stain from a tension test
strain =0,=1.0
= Fracture
with a un-notched specimen at room temperature is assumed as 0 . As the temperature

increases along the tension test process, in Equation (5.14) is assumed as the
average temperature of this process. Hence, the temperature used is expressed as shown
in Equation (5.15)
=
=
74
0 +
2
(5.15)
The original temperature, 0 , is set prior to the tensile loading surrounding the
specimen in an oven. The fracture temperature, , are computed using Equation (5.16)
changed from Equation (5.10) as shown in the plot of thermal softening fraction vs.
dimensionless temperature in Figure 5.8.
= (1 )
(5.16)
The experimental data needed in Equation (5.9) and (5.16) to compute is in

Table 5.9.
Table 5.9: Experimental data used to compute and result .
No.
Original
Temperature,
T0 ()
1
2
3
4
5
25
25
150
150
285
Yield
Force,
Area,
stress,
F (N)
A (m2)
y
(MPa)
19450.95 15.9 1223.33
19780.25 16.315 1212.40
17657.6 16.08 1098.11
18587.65 17.085 1087.95
18930.3 17.075 1108.66
Thermal
softening
fraction,
(-)
1
0.991062412
0.897639462
0.889335871
0.906260583
Fracture
temperature,
(-)
0
0.0126
0.1207
0.1298
0.1113
The fracture strains from tension tests with un-notched specimens at three
temperatures and the ratios of these fracture strains are used as shown in Table 5.10.
The result of the No. 2 test is smaller than at original temperature, so this value is
ignored and not used for obtaining D5 value.
75
Table 5.10: The fracture strains used for obtaining D5 of the Johnson-Cook failure model from
tension test with un-notched specimens at three temperatures.
Original
temperature
No.
1
2
3
4
5
T0 ()
Original
temperature
0
25
25
150
150
285
0
0
0.0943
0.0943
0.1962
Fracture
temperature
Average
temperature
Fracture
strain
0
0.0126
0.1207
0.1298
0.1113
0
0.0063
0.1075
0.1120
0.1538
0.69
0.66
0.74
0.68
0.68
Ratio
1.0455
1
1.1212
1.0303
1.0303
The plot of the ratio of fracture strains and the dimensionless temperature is shown
in Figure 5.15. The straight line in the plot fits the experimental results with the least
squares method. The value of the slope of this line is D5. Hence, 5 0.46 .
Compared to the published D5 value as shown in Table 5.1 (b), this data is smaller and
more close to the value by Varizi [2010].
Figure 5.15: Plot of the ratio of fracture strain and dimensionless temperature.
To verify the constant D5 of the Johnson-Cook failure formula, simulations were

coming out at similar conditions with the experimental work with an operating
temperature at 150C and 285C. The percentages of the elongation and the reduction
in the area of these experiments were compared. Figure 5.16 and Figure 5.17 show the
76
comparison of the deformation between these experimental geometries by the camera

and its computed geometry from strain contours of the simulations, respectively at
150C and 285C. The length of the radius of the cross section after fracture of the
experimental and modelling results are compared and was found to be similar. That
finding means the simulation agreed with the experimental result, which validates the
Johnson-Cook fracture constants D5.
(c)
Specimen with notch 2 (R0=6.35mm)
(e) Specimen with notch 4 (R0=1.5875mm)
Figure 5.16: Comparison of the experimental and numerical views of the failure specimens
at 150C.
77
(c) Specimen with notch 2 (R0=6.35mm)

Figure 5.17: Comparison of the experimental and numerical views of the failure specimens
at 285C.
To prove the constant D5 additionally, the reduction area from the tension tests at
150C and 285C and the corresponding simulations are compared as shown in Table
5.11. The errors between the experimental and modelling results are all in a small scale
2%, that further validate the D5 constant.
78
Table 5.11: Comparison of the experimental and modelling reduction of areas.

(a) Temperature=150C
Type of
the tensile
specimen
RAe (%)
RAm (%)
Errors (%)
UnNotched 1
Notched 2
Notched 3
Notched 4
notched R0=9.525mm R0=6.35mm R0=3.175mm R0=1.5875mm
49.3
49.1
-0.5
48.9
48.5
-0.8
42.2
42.3
0.2
38.3
38.2
-0.4
32.0
31.9
-0.3
(b) Temperature=285C
Type of the tensile

specimen
RAe (%)
RAm (%)
Errors (%)
Unnotched
49.5
49.6
0.3
Notched 1
Notched 2
Notched 3
R0=9.525mm R0=6.35mm R0=3.175mm
40.4
35.7
33.4
40.1
36.3
33.5
-0.7
1.8
0.3
5.4.3 Verification of the Calibrated the Johnson-Cook Fracture Constant D4

The JohnsonCook fracture constant D4 relates to the effect of the strain rate on
the fracture property. The D4 value used here is 0.002 from Borkovec [2008]. A cutting
simulation is employed to verify the D4 value.
5.5 Evaluation of the Johnson-Cook Fracture Constants with

Cutting Simulations
To validate the calibrated constants of the Johnson-Cook constitutive and failure
models, the cutting process is simulated with three cutting-velocities and three feedrates, as shown in Table 5.12.
Table 5.12: The cutting conditions used in FE simulations of cutting process.
Cutting velocity
(m/min)
150 250 350
Feed rate
(mm)
0.1 0.2 0.3
79
5.5.1 Chip Formation and Stress Contour

Figure 5.18 shows the von Misses stress contour maps around the cutting region
at step 125/250 with three feed rates and three cutting velocities: 0.1 mm, 0.2 mm, 0.3
mm and 150 m/min, 250 m/min, 350 m/min. The continuous chips were observed for
all the simulations. The shapes of the chips with the same velocity but different feed
rates are compared, and the higher the feed rate is, the larger the chip curl radius, which
agrees with results published by Reddy [2011].
The crack performance was checked by visual inspection. If the crack tip
propagates ahead of the cutting edge, the Johnson-Cook fracture constant
underestimated the actual fracture strain. However, when the crack tip is at the same
location on the cutting edge, which is a function of time, then the Johnson-Cook fracture
constant is accurately modelling the fracture strain as a function of hydrostatic pressure,
temperature, and strain rate, which is the case in the results shown in Figure 5.18.
In addition, Figure 5.18 also shows the simulated von Mises stress contours during
cutting. Maximum stresses were observed in the workpiece material at the cutting edge
region. High stress magnitude was also observed in the primary shear zone and
secondary deformation zone, and the high stress also supplies enough energy for the
plastic deformations.
80
(a)f1 v1
(b)f1 v2
(c)f1 v3
(d)f2 v1
81
(e)f2 v2
(f)f2 v3
(g)f3 v1
82
(h)f3 v2
(i)f3 v3
Figure 5.18: von Misses stresses contour maps around the cutting layer of the cutting
simulations at step 125/250:
f1=0.1mm, f2=0.2mm, f3=0.3mm; v1=150m/min,
v2=250m/min, v3=350m/min.
5.5.2 Temperature of the Tool-chip Interface

Figure 5.19 shows the temperature contour plots of all the cutting simulations. As
shown in Figure 5.19, at the same feed rate, the temperature of the tool-chip interfaces
increases when the cutting speed increases. This finding agreed well with experimental
results, as a higher cutting speed offers more energy for heat generation [Dhar, 2007].
Also, at the same cutting speed, the temperature of the area where the tool contacted
with the chip also increases as the feed rate increases. This phenomenon is because the
83
volume of material to be removed increased, resulting in higher mechanical energy

being converted to heat energy.
(a) f1 v1
(b) f1 v2
(c) f1 v3
(d) f2 v1
(e) f2 v2
(f) f2 v3
84
(g) f3 v1
(h) f3 v2
(i) f3 v3
Figure 5.19: Temperature contour maps of the cutting simulations:
f1=0.1 mm, f2=0.2 mm,
f3=0.3 mm; v1=150 m/min, v2=250 m/min, v3=350 m/min.
Additionally, to compare the temperatures of the tool-chip interface at different

feeds and cutting speeds, a path along the tool-chip interface on both the chip surface
and tool surface are set. Figure 5.20 shows an example of setting the paths of the toolchip interface in the work piece in the model with a feed rate of 0.2 mm and a cutting
speed of 250 m/min. Figure 5.21 shows the temperatures along the tool-chip interface
in the work piece and in the tool with a feed rate of 0.2 mm and a cutting speed of 250
m/min. The temperature in the tool decreases from the cutting edge to the point where
the chip separated from the tool along the tool-chip surface while the temperature trend
was reversed in the chip. To compare the temperatures in the chip along the tool-chip
85
interface with different feeds and the cutting speeds, the average temperatures are
shown in Figure 5.22. The temperature increases as the feed rate increase at the same
cutting speed. A similar trend was observed with increasing cutting speed and constant
feed rate [Dhar, 2007]. Furthermore, the range shown in the Figure 5.22 is the
oscillation of the temperature compared to the average data, and the oscillation
increases as the feed rate and the cutting speed increases.
In the work piece
In the tool
Figure 5.20: Paths used as the tool-chip interface in the work piece and in the tool.
Figure 5.21: Temperatures along the tool-chip interface in the work piece and in the tool.
86
Figure 5.22: Comparison of the temperatures in the tool-chip interface with different feeds
and cutting speeds.
5.5.3 Cutting Force

To consider the performance of the cutting simulations at three feed rates and
velocities, the average cutting forces are compared. The cutting forces are obtained by
adding up the forces of the nodes along the right vertical side of the tool as shown in
Figure 5.23, and the results of the average cutting forces are shown in Figure 5.24. To
compare with the same feed rate, the cutting forces decrease as the cutting speed
increase. Moreover, compared to the same cutting speed, the cutting forces increase
with the feed rate increase, as more energy is needed from the cutting forces to the
thicker chip deformation. Besides, the range was shown in the Figure 5.24 to indicate
that the oscillation of the cutting forces along the cutting progress, and that the
oscillation is stronger as the cutting depth increases. The cutting force from a cutting
experimental result from Hosseinkhani and Ng [2013] is also compared in Figure 5.24,
and the cutting experimental conditions and results are shown in Table 5.13. The error,
87
E as shown in Figure 5.24 and Table 5.13, obtained from the comparison of the cutting
force between the experimental result and FEM result is 5%.
Table 5.13: The cutting experimental conditions and results.
Cutting speed (m/min)

250
Feed (mm)
0.2
Cutting force (N)
1224.62
Normal to cutting force(N)
450
Thick of the chip (mm)
0.4
Error (%) (Compared with FEM)
5
Figure 5.23: The nodes used for obtaining the cutting forces.
Figure 5.24: Cutting forces of the cutting simulations.
88
5.5.4 Chip Thickness

Figure 5.25 shows an example of the measurement of the chip thickness from the
cutting model with a feed rate of 0.2 mm and a cutting velocity of 250 m/min. Where
tu is the uncut chip thickness equalling the cutting depth, and tc is the chip thickness
perpendicular to the rake face of the tool. All of the chip thicknesses of the cutting
models are shown in Table 5.14 and Figure 5.26. The chip thickness decreases as the
cutting speed increases, and the chip thickness increases as the feed increases, which
two phenomena agree with the reality. Hence, the calibrated material properties are
valid.
Figure 5.25: Measurement of the chip thickness of the cutting model with feed of 0.2mm,
velocity of 250m/min.
Figure 5.26: Chip thickness with three feeds and cutting speeds from the tension models.
89
Table 5.14: Chip thickness of all cutting simulations.
Cutting
speed
(m/min)
150
250
350
Feed rate
(mm)
Chip
thickness
measured
from
FEM,
tc_m (mm)
0.1
0.2
0.3
0.1
0.2
0.3
0.1
0.2
0.3
0.33
0.53
0.74
0.28
0.49
0.70
0.27
0.48
0.68
Shear
plane
angle
measured
from
FEM,
(deg)
18
20.8
23.7
20.8
23.5
24.5
21.8
23.4
24.5
Theoretical
chip
Difference,
thickness,
e (%)
tc_t (mm)
0.32
0.54
0.71
0.27
0.48
0.68
0.26
0.48
0.68
4.68
-2.18
4.70
3.36
3.04
2.68
4.77
0.48
-0.26
To further prove that the obtained chip thickness of the cutting model is dependent
on the calibrated material properties, the measured chip thickness tc_m is compared with
the theoretical chip thickness tc_t, which is computed depending on Merchants [1945a;
1945b] model using the equations below.
r=
(5.17)
(5.18)
tan = 1
= ( + )
(5.19)
where r is the chip thickness ratio, to is the thickness of the chip prior to chip formation,
tc is the chip thickness after separation, is the rake angle of the tool, is the shear
plane angle. To obtain the theoretical chip thickness tc_t by using Equation (5.19), is
needed and measured from the equivalent strain contour as shown in Figure 5.18. The
results of , tc_t and the differences (%), which equals
_ _
_
100% , are in Table
5.14. The differences are in 5%, which means the chip thickness of the cutting models
90
depending on the calibrated material properties agree well with the analytical model.
5.5.5 Strain to Fracture of the Layer

Figure 5.27 shows the effects of the cutting conditions on the fracture strain
induced in the conditional link of the cutting models. All the fracture strain magnitude
was obtained at 0.0012 s. As shown in Figure 5.27, the fracture strain of the element in
the layer increased with the higher cutting speed. While the fracture strain of the
element in the layer also increases with the larger feed rate. This phenomenon was
because at higher cutting speeds together with larger feed rates provide more heat
energy that was evolved during the chip separation.
Figure 5.27: The strains to fracture of the layer at time 0.0012s under different cutting
conditions.
Figure 5.28 shows the stress triaxiality of the condition link elements in the layers
of the cutting simulations at 0.0012 s. The stress triaxiality of the element in the layer
is hard to obtain directly from the cutting models, as the element was deleted when it
91
failed as the pressure and von Mises stress of the element are both zero. Hence, the
stress triaxiality of the element in the layer assumed equals the sum of the stress
triaxialities of the top element and bottom element as shown in Figure 5.29. As shown
in Figure 5.28, the stress triaxiality of the element in the layer decreases with higher
cutting speed, and it also decreases with the larger feed rate. In other words, the trends
of the stress triaxiality of the element in the layer with the cutting speeds and feed rates
are reversed to the trends of its strains to fracture. With references to both Figures 5.27
and 5.28, with a higher stress triaxiality ratio, the strain to fracture decreases. This trend
was also observed by Johnson and Cook [1985], as detailed in Figure 2.17.
Figure 5.28: The stress triaxiality of the element in the layer of the cutting simulations under
different cutting conditions at time 0.0012s.
92
Figure 5.29: The top and bottom elements used for the calculation of the stress triaxiality of
the element in the layer.
Figure 5.30 shows the temperatures of the elements of the layers in the cutting
simulations under different cutting conditions at the same time 0.0012s. As shown in
Figure 5.30, the temperature of the element in the layer increases with the increasing of
the cutting speed and feed rate. Comparing with Figure 5.27, the trends of the
temperatures of the elements in the layers have similar trends of the strains to fracture
of the elements in the layers. Comparing with Figure 5.27, at a higher temperature of
the elements in the layer, the strain to fracture also increases. This finding was due to
higher ductility with higher temperature. This trend was also observed by Borvik et al.
[2001] as shown in Figure 2.13 (b).
Figure 5.30: The temperatures of the elements in the layers of the cutting simulations under
different cutting conditions.
93
6. Conclusion
In this thesis, the Johnson-Cook constitutive and fracture parameters were
calibrated for AISI 1045 steel based on a set of quasi-static tensile tests with smooth
and notched round bars at room temperature and elevated temperatures respectively.
The results were validated by the FEM simulations of tensile tests and the orthogonal
cutting process, and both have good agreement with the corresponding experimental
results.
Stress triaxiality is the principal influence factor on the ductile fracture. Depending
on three different fracture mechanisms, including the shear decohesion, the void growth
and the combination of the shear decohesion and void growth, the range of stress
triaxiality was divided into three branches respectively. However, the Johnson-Cook
fracture model is a monotonic function that is only valid in a narrow range of stress
triaxiality with void growth fracture mechanisms. Hence, it may decrease the accuracy
of the Johnson-Cook fracture parameters if more other tests are considered, such as
torsion tests and compression tests. The conclusion is that the tensile tests with smooth
and notched round bars are good enough to calibrate the Johnson-Cook fracture
parameters.
In the metal cutting process, two fracture mechanisms occur in the fracture
elements: shear decohesion and void growth. As the Johnson-Cook fracture model is
only used to predict the onset of fracture, the fracture evolution with these two fracture
mechanisms is necessarily used in the FEM simulations of cutting.
94
Compared to the literature calibrations of the Johnson-Cook fracture parameters

for AISI 1045 steel, the calibration in this thesis does not require high strain rate
experimental results acquired from Split-Hopkinson bar tests.
95
7.
Future Works
Future works extended from this thesis include the following:

1.
To verify the obtained Johnson-Cook fracture parameters for AISI 1045 steel
in FE simulations of metal cutting with other cutting conditions, such as the cutting
with a negative or zero rake angle or with high cutting velocity.
2.
To explore this calibration method for Johnson-Cook fracture parameters for
other materials. Furthermore, to implement the fracture parameters of different

components based on this simple calibration method into the heterogeneous FE cutting
model.
96
BIBLIOGRAPHY
[ABAQUS, 2003] Hibbit, H.D., Karlsson, B.I., & Sorrensen (2003). ABAQUS Online
Documentation: Version 6.4-1, ABAQUS Inc.
[Anderson, 2005] Anderson, T. L., & Anderson, T. (2005). Fracture mechanics:
fundamentals and applications: CRC press. ISBN: 978-1-4200-5821-5
[Argon, 1975] Argon, A., Im, J., & Safoglu, R. (1975). Cavity formation from
inclusions in ductile fracture. Metallurgical Transactions A, 6(4), 825-837.
[Arias, 2008] Arias, A., Rodrguez-Martnez, J. A., & Rusinek, A. (2008). Numerical
simulations of impact behaviour of thin steel plates subjected to cylindrical, conical and
hemispherical non-deformable projectiles. Engineering Fracture Mechanics, 75(6),
1635-1656.
[Autenrieth, 2009] Autenrieth, H., Schulze, V., Herzig, N., & Meyer, L. W. (2009).
Ductile failure model for the description of AISI 1045 behavior under different loading
conditions. Mechanics of Time-Dependent Materials, 13(3), 215-231.
[Brvik, 2005] Brvik, T., Hopperstad, O., Dey, S., Pizzinato, E., Langseth, M., &
Albertini, C. (2005). Strength and ductility of Weldox 460 E steel at high strain rates,
elevated temperatures and various stress triaxialities. Engineering Fracture Mechanics,
72(7), 1071-1087.
[Brvik, 2001] Brvik, T., Hopperstad, O. S., Berstad, T., & Langseth, M. (2001). A
computational model of viscoplasticity and ductile damage for impact and penetration.
European
Journal
of
Mechanics
A/Solids,
20(5),
685-712.
doi:http://dx.doi.org/10.1016/S0997-7538(01)01157-3
[Bai, 2009] Bai, Y., Teng, X., & Wierzbicki, T. (2009). On the application of stress
triaxiality formula for plane strain fracture testing. Journal of Engineering Materials
and Technology, 131(2), 021002.
[Bai, 2008] Bai, Y., & Wierzbicki, T. (2008). A new model of metal plasticity and
fracture with pressure and Lode dependence. International Journal of Plasticity, 24(6),
1071-1096.
[Bao, 2003] Bao, Y. (2003). Prediction of ductile track formation in uncracked bodies.
Ph.D. thesis, Massachusetts Institute of Technology.
[Bao, 2004a] Bao, Y., & Wierzbicki, T. (2004a). A comparative study on various ductile
97
crack formation criteria. Journal of Engineering Materials and Technology, 126(3),

314-324.
[Bao, 2004b] Bao, Y., & Wierzbicki, T. (2004b). On fracture locus in the equivalent
strain and stress triaxiality space. International Journal of Mechanical Sciences, 46(1),
81-98.
[Bao, 2005] Bao, Y., & Wierzbicki, T. (2005). On the cut-off value of negative triaxiality
for fracture. Engineering Fracture Mechanics, 72(7), 1049-1069.
[Bil, 2004] Bil, H., Kl, S. E., & Tekkaya, A. E. (2004). A comparison of orthogonal
cutting data from experiments with three different finite element models. International
Journal of Machine Tools and Manufacture, 44(9), 933-944.
[Borkovec, 2008] Borkovec, J. (2008). Computer simulation of material separation
process. Ph.D. thesis, Brno: Brno University of Technology, Institute of Solid
Mechanics, Mechatronics and Biomechanics.
[Bridgman, 1952] Bridgman, P. W. (1952). Studies in large plastic flow and fracture
(Vol. 177): McGraw-Hill New York.
[Callister, 2007] Callister, W. D., & Rethwisch, D. G. (2007). Materials science and
engineering: an introduction (Vol. 7): Wiley New York.
[Ceretti, 1996] Ceretti, E., Fallbhmer, P., Wu, W., & Altan, T. (1996). Application of
2D FEM to chip formation in orthogonal cutting. Journal of Materials Processing
Technology, 59(1), 169-180.
[Cockcroft, 1968] Cockcroft, M., & Latham, D. (1968). Ductility and the workability
of metals. Journal of Institute of Metals, 96(1), 33-39.
[Cottrell, 2012] Cottrell, A. (2012). Theoretical aspects of fracture. Paper presented at
the ICF0, Swampscott-MA (USA) 1959.
[Davidson, 1968] Davidson, T., & Ansell, G. (1968). The structure sensitivity of the
effects of pressure upon the ductility of Fe-C materials. ASM TRANS QUART, 61(2),
242-254.
[Dhar, 2007] Dhar, N.R., 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 & Manufacture, (47),748-753.
[Duan, 2009] Duan, C., Dou, T., Cai, Y., & Li, Y. (2009). Finite element simulation and
98
experiment of chip formation process during high speed machining of AISI 1045
hardened steel. International Journal of Recent Trends in Engineering, 1(5), 46-50.
[Dzugan, 2013] Dzugan, J., Prantl, A., Spaniel, M., Konopik, P., Ruzicka, J., & Kuzelka,
J. (2013). Identification of ductile damage parameters for pressure vessel steel.
SMiRT22, 18, 30.
[Hancock, 1976] Hancock, J., & Mackenzie, A. (1976). On the mechanisms of ductile
failure in high-strength steels subjected to multi-axial stress-states. Journal of the
Mechanics and Physics of Solids, 24(2), 147-160.
[Henry, 1997] Henry, B., & Luxmoore, A. (1997). The stress triaxiality constraint and
the Q-value as a ductile fracture parameter. Engineering Fracture Mechanics, 57(4),
375-390.
[Hortig, 2007] Hortig, C., & Svendsen, B. (2007). Simulation of chip formation during
high-speed cutting. Journal of Materials Processing Technology, 186(1), 66-76.
[Hosseinkhani, 2013] Hosseinkhani, K., & Ng, E. (2013). Analysis of the Cutting
Mechanics Under the Influence of Worn Tool Geometry. Procedia CIRP, 8, 117-122.
doi:http://dx.doi.org/10.1016/j.procir.2013.06.075
[Huang, 1996] Huang, J., & Black, J. (1996). An evaluation of chip separation criteria
for the FEM simulation of machining. Journal of Manufacturing Science and
Engineering, 118(4), 545-554.
[Iwata, 1984] Iwata, K., Osakada, K., & Terasaka, Y. (1984). Process modelling of
orthogonal cutting by the rigid-plastic finite element method. Journal of Engineering
Materials and Technology, 106(2), 132-138.
[Jaspers, 2002] Jaspers, S., & Dautzenberg, J. (2002). Material behaviour in conditions
similar to metal cutting: flow stress in the primary shear zone. Journal of Materials
Processing Technology, 122(2), 322-330.
[Johnson, 1989] Johnson, G., & Holmquist, T. (1989). Test data and computational
strength and fracture model constants for 23 materials subjected to large strains, high
strain rates, and high temperatures. Los Alamos National Laboratory, Los Alamos, NM,
Report No. LA-11463-MS.
[Johnson, 1983] Johnson, G. R., & Cook, W. H. (1983). A constitutive model and data
for metals subjected to large strains, high strain rates and high temperatures. Paper
presented at the Proceedings of the 7th International Symposium on Ballistics,541-547.
99
[Johnson, 1985] Johnson, G. R., & Cook, W. H. (1985). Fracture characteristics of three
metals subjected to various strains, strain rates, temperatures and pressures.
Engineering Fracture Mechanics, 21(1), 31-48.
[Kalhori, 2001] Kalhori, V. (2001). Modelling and simulation of mechanical cutting.
Ph.D. these, Lulea University of Technology.
[Kao, 1990] Kao, A., Kuhn, H., Richmond, O., & Spitzig, W. (1990). Tensile fracture
and fractographic analysis of 1045 spheroidized steel under hydrostatic pressure.
Journal of Materials Research, 5(01), 83-91.
[Klamecki, 1973] Klamecki, B. E. (1973). Incipient Chip Formation in Metal Cutting-a Three-dimension Finite Element Analysis. Ph.D. thesis, University of Illinois at
Urbana-Champaign.
[Komvopoulos, 1991] Komvopoulos, K., & Erpenbeck, S. (1991). Finite element
modelling of orthogonal metal cutting. Journal of Manufacturing Science and
Engineering, 113(3), 253-267.
[Rosa, 2001] La Rosa, G., Mirone, G., & Risitano, A. (2001). Effect of stress triaxiality
corrected plastic flow on ductile damage evolution in the framework of continuum
damage mechanics. Engineering Fracture Mechanics, 68(4), 417-434.
[Roy, 1981] Le Roy, G., Embury, J., Edwards, G., & Ashby, M. (1981). A model of
ductile fracture based on the nucleation and growth of voids. Acta Metallurgica, 29(8),
1509-1522.
[Lemaitre, 1992] Lemaitre, J. (1992). A short course in damage mechanics: SpringerVerlag, New York. ISBN: 978-3-642-18255-6.
[Li, 2002] Li, K., Gao, X.-L., & Sutherland, J. (2002). Finite element simulation of the
orthogonal metal cutting process for qualitative understanding of the effects of crater
wear on the chip formation process. Journal of Materials Processing Technology,
127(3), 309-324.
[Lin, 1992] Lin, Z., & Lin, S. (1992). A coupled finite element model of thermo-elasticplastic large deformation for orthogonal cutting. Journal of Engineering Materials and
Technology, 114(2), 218-226.
[Liu, 2000] Liu, C., & Guo, Y. (2000). Finite element analysis of the effect of sequential
cuts and toolchip friction on residual stresses in a machined layer. International
Journal of Mechanical Sciences, 42(6), 1069-1086.
100
[Liu, 1968] Liu, C., & Gurland, J. (1968). The fracture behavior of spheroidized carbon
steels. Transactions of American Society for Metals, 61, 156-167.
[Liu, 2014] Liu, J., Bai, Y., & Xu, C. (2014). Evaluation of ductile fracture models in
finite element simulation of metal cutting processes. Journal of Manufacturing Science
and Engineering, 136(1), 011010.
[Majzoobi, 2011] Majzoobi, G., & Dehgolan, F. R. (2011). Determination of the
constants of damage models. Procedia Engineering, 10, 764-773.
[Mamalis, 2001] Mamalis, A., Horvath, M., Branis, A., & Manolakos, D. (2001). Finite
element simulation of chip formation in orthogonal metal cutting. Journal of Materials
Processing Technology, 110(1), 19-27.
[McClintock, 1968] McClintock, F. A. (1968). A criterion for ductile fracture by the
growth of holes. Journal of Applied Mechanics, 35(2), 363-371.
[Merchant, 1945a] Merchant, M. E. (1945a). Mechanics of the Metal Cutting Process.
I. Orthogonal Cutting and a Type 2 Chip. Journal of Applied Physics, 16(5), 267-275.
doi:doi:http://dx.doi.org/10.1063/1.1707586
[Merchant, 1945b] Merchant, M. E. (1945b). Mechanics of the Metal Cutting Process.
II. Plasticity Conditions in Orthogonal Cutting. Journal of Applied Physics, 16(6), 318324. doi:doi:http://dx.doi.org/10.1063/1.1707596
[Ralls, 1976] Ralls, K., Courtney, T. H., & Wulff, J. (1976). Introduction to materials
science and engineering: John Wiley & Sons, Inc. ISBN: 978-0-471-70665-6
[Reddy, 2011] Reddy M.R., Ravikumar, P., Rao, G.K.M (2011). Effect of feed rate on
the generation of surface roughness in turning. International Journal of Engineering
Science and Technology, 3 (11), 8099-8105
[Rice, 1969] Rice, J. R., & Tracey, D. M. (1969). On the ductile enlargement of voids
in triaxial stress fields. Journal of the Mechanics and Physics of Solids, 17(3), 201-217.
[Shet, 2000] Shet, C., & Deng, X. (2000). Finite element analysis of the orthogonal
metal cutting process. Journal of Materials Processing Technology, 105(1), 95-109.
[Soo, 2007] Soo, S., & Aspinwall, D. (2007). Developments in modelling of metal
cutting processes. Proceedings of the Institution of Mechanical Engineers, Part L:
Journal of Materials Design and Applications, 221(4), 197-211.
[Strenkowski, 1985] Strenkowski, J. S., & Carroll, J. (1985). A finite element model of
101
orthogonal metal cutting. Journal of Manufacturing Science and Engineering, 107(4),

349-354.
[Sun, 1991] Sun, J., Deng, Z.-J., & Tu, M.-J. (1991). Effect of stress triaxiality levels
in crack tip regions on the characteristics of void growth and fracture criteria.
Engineering Fracture Mechanics, 39(6), 1051-1060.
[Teng, 2006] Teng, X., & Wierzbicki, T. (2006). Evaluation of six fracture models in
high velocity perforation. Engineering Fracture Mechanics, 73(12), 1653-1678.
[Thomason, 1998] Thomason, P. (1998). A view of ductile-fracture modelling. Fatigue
& Fracture of Engineering Materials & Structures, 21(9), 1105-1122.
[Usta, 1999] Usta, M. (1999). Finite element analysis of orthogonal metal cutting
operations. Ph.D. Thesis, Mechanical Engineering Department of Middle East
Technical University, Ankara, Turkey.
[Usui, 1982] Usui, E., & Shirakashi, T. (1982). Mechanics of machiningfrom
descriptive to predictive theory. On the Art of Cutting Metals75 Years Later, 7, 13-35.
[Vaziri, 2010] Vaziri, M., Salimi, M., & Mashayekhi, M. (2010). A new calibration
method for ductile fracture models as chip separation criteria in machining. Simulation
Modelling Practice and Theory, 18(9), 1286-1296.
[Wierzbicki, 2005] Wierzbicki, T., Bao, Y., Lee, Y.-W., & Bai, Y. (2005). Calibration
and evaluation of seven fracture models. International Journal of Mechanical Sciences,
47(45), 719-743. doi:http://dx.doi.org/10.1016/j.ijmecsci.2005.03.003
[Wierzbicki, 1998] Wierzbicki, T., & Werner, H. (1998). Cockroft and Latham
Revisited. Impact & Crashworthiness Laboratory Report(16).
[Wilkins, 1978] Wilkins, M. L. (1978). Mechanics of penetration and perforation.
International Journal of Engineering Science, 16(11), 793-807.
[Xie, 1998] Xie, J., Bayoumi, A., & Zbib, H. (1998). FEA modelling and simulation of
shear localized chip formation in metal cutting. International Journal of Machine Tools
and Manufacture, 38(9), 1067-1087
[Zhang, 1999] Zhang, L. (1999). On the separation criteria in the simulation of
orthogonal metal cutting using the finite element method. Journal of Materials
Processing Technology, 89, 273-278.
102
APPENDIX A
The grips of the tensile tests used as the fixture were designed at McMaster
University by Dr. Mike Bruhis, as shown in Figure A1. The stainless steel was chosen
as the materials of the grips to reduce the wear during the high temperature tests.
(a)
(b)
Figure A1: (a) Dimensions of the upper grip; (b) Dimensions of the bottom grip.
103

Johnson-Cook Failure Parameters

Uploaded by

Copyright:

Available Formats

Johnson-Cook Failure Parameters

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Johnson-Cook Failure Parameters

Uploaded by

Copyright:

Available Formats

What is the thesis about?

What is the thesis about?

What method is used to calibrate the fracture parameters?

What method is used to calibrate the fracture parameters?

JOHNSON-COOK FAILURE PARAMETERS

CALIBRATION OF THE JOHNSON-COOK FAILURE PARAMETERS AS

By KEYAN WANG, B. ENG.

A Thesis Submitted to the School of Graduate Studies in Partial Fulfilment of

McMaster University Copyright by Keyan Wang, March 2016

MASTERS OF APPLIED SCIENCE (2016)

TITLE: Calibration of the Johnson-Cook Failure Parameters as the Chip Separation

AUTHOR: Keyan Wang, B. Eng. (McMaster University)

SUPERVISORS: Dr. Eu-Gene Ng

NUMBER OF PAGES: xx,103

2.2.3 Bao-Wierzbicki Fracture Model .................................................................. 23

4.2.2 Material Properties ..................................................................................... 50

5.4.3 Calibration and Verification of the Johnson-Cook Fracture Constant D4 . 79

JOHNSON-COOK FRACTURE CONSTANTS

Figure 2.14: Dependences of averaged failure strain , = 2 (0 ) and

normalized failure strain for the tensile specimens at different temperatures.

LIST OF TABLE CAPTIONS

3 , 1 2 3 are the rate of principle strains. [Bao,

NOMENCLATURE AND LIST OF SYMBOLS

: Equivalent plastic displacement at fracture (m)

: Equivalent plastic displacement (m)

Material constant (-)

: True fracture strain measured from a smooth specimen test (-)

: Void nucleation strain (-)

: Stress tensor (MPa)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

Ductile fracture plays an important role in a successful finite element simulation

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

2.1.2 Effect of Stress State on Ductile Fracture

1 2 3 [Bai, 2009]. The stress triaxiality parameter , as shown in Equation (2.5),

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

spherical coordinate system ( f , , ) [Bai, 2009]. The relationship between

Bridgman [1952] observed that the ductility of material is highly sensitive to

an analysis of the void growth. The function of f ( ) is shown in Equation (2.9).

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

explore a formula of f ( ) as listed in Equation (2.10).

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

The experimental results dominated by different fracture mechanisms are shown

f 1.9 2 0.18 0.21

for 0.4 0.95

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

Figure 2.9: A newly postulated 3D asymmetric fracture locus. [Bai, 2008]

f ( , ) [ ( D1e D D5e D ) D3e D ] ( D1e D D5e D ) D3e D (2.15)

2.1.3 Effects of Temperature and Strain Rate on Ductile Fracture

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)

Figure 2.11: Split Hopkinson device. [Autenrieth, 2009]

M. A. Sc Thesis - K. Wang - McMaster University - Mechanical Engineering (2016)