Extension of Oxleys Analysis

of Machining to Use Different

Material Models
Amir H. Adibi-Sedeh
Vis Madhavan
e-mail: vis.madhavan@Wichita.edu

Behnam Bahr
College of Engineering, Wichita State University,
Wichita, KS 67260-0035

The aim of the present work is to extend the applicability of Oxleys analysis of machining
to a broader class of materials beyond the carbon steels used by Oxley and co-workers.
The Johnson-Cook material model, history dependent power law material model and the
Mechanical Threshold Stress (MTS) model are used to represent the mechanical properties of the material being machined as a function of strain, strain rate and temperature. A
few changes are introduced into Oxleys analysis to improve the consistency between the
various assumptions. A new approach has been introduced to calculate the pressure
variation along the alpha slip lines in the primary shear zone including the effects of both
the strain gradient and the thermal gradient along the beta lines. This approach also has
the added advantage of ensuring force equilibrium of the primary shear zone in a macroscopic sense. The temperature at the middle of the primary shear zone is calculated by
integrating the plastic work thereby eliminating the unknown constant . Rather than
calculating the shear force from the material properties corresponding to the strain, strain
rate and temperature of the material at the middle of the shear zone, the shear force is
calculated in a consistent manner using the energy dissipated in the primary shear zone.
The thickness of the primary and secondary shear zones, the heat partition at the primary
shear zone, the temperature distribution along the tool-chip interface and the shear plane
angle are all calculated using Oxleys original approach. The only constant used to fine
tune the model is the ratio of the average temperature to the maximum temperature at the
tool-chip interface (). The performance of the model has been studied by comparing its
predictions with experimental data for 1020 and 1045 steels, for aluminum alloys 2024T3, 6061-T6 and 6082-T6, and for copper. It is found that the model accurately reproduces the dependence of the cutting forces and chip thickness as a function of undeformed
chip thickness and cutting speed and accurately estimates the temperature in the primary
and secondary shear zones. DOI: 10.1115/1.1617287

Introduction

The physics of metal cutting involves complex interactions

among a multitude of phenomena such as plasticity, friction, heat
generation, heat flow, material damage and phase changes that are
dealt with in separate disciplines such as mechanics, heat transfer,
tribology, materials science and mathematics. Since the time of
Trescas studies on the planing of metals, there have been significant improvements in our understanding of machining resulting
from advances in these disciplines. However, machining still
poses ample challenges to researchers.
Machining process development involves determination of inputs such as machine tool operating parameters and the cutting
tool geometry, material and coating to be adopted in order to
obtain desired values of outputs such as cutting force, chip morphology, machined surface characteristics and tool life. Design
and development of machining processes and cutting tools is still
primarily based on empirical knowledge, with additional experimentation carried out as needed. This approach is costly, time
consuming and often leads to non-optimal manufacturing 1.
Continuous enhancements to the accuracy of machining models
and determination of accurate values for critical inputs needed for
these models are prerequisites for scientific design of machining
processes.
Over the past century and a half, many researchers have tried to
understand the machining process under the framework of plasticity theory. Piispanen 2 and Ernst 3 proposed the card model of
chip formation in which the shear plane angle is unknown.
Contributed by the Manufacturing Engineering Division for publication in the
JOURNAL OF MANUFACTURING SCIENCE AND ENGINEERING. Manuscript received
March 2002; Revised June 2003. Associate Editor: Y. Shin.

656 Vol. 125, NOVEMBER 2003

Merchant 4, using energy minimization, found that this angle

could be determined if the average friction angle at the tool-chip
interface were known. Various other researchers have developed
2D models of machining by the use of slip line field theory 57
used in conjunction with Hills theory of permissible intensity of
stress singularities as applicable to machining 8. The most significant contribution towards modeling of practical machining has
been the parallel-sided shear zone theory of Oxley and co-workers
912. The unique feature of Oxleys machining theory 12 is
that the dependence of material flow stress upon strain, strain rate
and temperature is considered to obtain the shear angle and
other outputs of interest. This theory has been modified to analyze
a range of machining operations 1315 and extensive experimental validation of the theory has been carried out 16. One
significant limitation of the model is the fact that it has been
almost exclusively applied to carbon steels. There is a need to
extend the applicability of this theory to other materials commonly machined.

Description of Oxleys Machining Theory

Oxleys model of machining was developed based on experimental observations 17 of the material deformation. As shown in
Fig. 1, the deformation in metal cutting is concentrated in two
zonesa primary shear zone centered about AB the nominal
shear plane of length L and a secondary shear zone along the
tool-chip interface. Though the actual shapes of the two zones are
approximately as depicted in Fig. 1, the primary shear zone is
assumed to be parallel-sided and the secondary shear zone is assumed to be of constant thickness, to simplify analysis. The parameter c is used to represent the relative length of the primary

Copyright 2003 by ASME

Transactions of the ASME

minimizes the overall cutting force 19. This is taken to be the

correct value of and the corresponding shear plane angle, forces,
strain rates and temperatures are taken to be the actual values of
these output variables.
Oxley 12 and co-workers developed a computer program to
carry out the above analysis. It is found that the analysis yields
results in good agreement with experiments when parameters
and are used to tune the model. Typical values of these parameters used by Oxley and co-workers range from 0.5 to 1.0 for
carbon steels.

3 Extension of Oxleys Machining Analysis to Use

Other Material Models

Fig. 1 Model of chip formation used in Oxleys analysis for

orthogonal machining 12

shear zone with respect to the thickness of the primary shear zone
and is the ratio of the thickness of the secondary shear zone to
the chip thickness.
In the primary shear zone the slip line along the direction AB
Fig. 1 is an alpha slip line and in the secondary shear zone the
slip line along the chip face is a beta slip line. Assuming that the
strain at AB is uniform, equal to one half the strain in the primary
shear zone, and further assuming that the temperature and strain
rate are uniform along AB, the shear stress along AB is calculated
for a given c. A parameter is introduced to obtain the temperature of the middle of the shear zone as a fraction of the temperature rise through the primary shear zone. The gradient of shear
stress along the beta slip lines is obtained from constitutive equation 12 assuming the strain rate is a maximum along AB 10
and that the component due to the temperature gradient is negligible. Using the force equilibrium of an element on the free surface of work material close to point A the hydrostatic pressure at
point A, p A , is obtained. Applying force equilibrium along the
shear plane direction and using the gradient of strength perpendicular to the nominal shear plane the normal pressure variation
along AB can be determined. Knowing the pressure and the shear
stress along AB, the resultant force, its direction of action and its
moment about B the tool tip are calculated.
Assuming uniform distribution of the normal stress along the
rake face, the tool-chip contact length is obtained so that the moment of the normal force about the point B equals the moment of
the resultant force along the shear plane. The normal stress n and
the shear stress int on the rake face are obtained from the corresponding forces and the contact length. The normal stress on the
rake face can be obtained in a different manner by assuming that
the alpha slip line in the primary shear zone turns to meet the tool
perpendicular to the rake face. The parameter c is fixed at the
value that causes n calculated by both ways to be the same 11.
Assuming that there is sticking friction at the tool-chip interface, the shear strength of the chip (k chip ) at the average temperature of the tool-chip interface should be equal to int . The maximum temperature at the tool-chip interface is calculated using
equations derived from Boothroyds 18 results. A parameter is
used to obtain the average temperature at the tool-chip interface,
T a v e , from the maximum temperature rise, m . For each assumed shear plane angle, all of these calculations are carried out
to find int , and the highest value of for which int k chi p is
taken to be the shear plane angle 9.
The calculations above are repeated for various values of , the
relative thickness of the secondary shear zone, to obtain the that
Journal of Manufacturing Science and Engineering

We have extended Oxleys model to use any of a range of

material models, namely, the Oxleys material model for carbon
steels, the Johnson-Cook material model, the history-dependent
power law model used by Maekawa and co-workers, and the MTS
model, so that machining of a wide range of materials can be
analyzed. In the following description of the extension it should
be noted that all assumptions and techniques used, other than
those explicitly discussed are the same as in Oxleys model.
3.1 Description of the Material Models Used. Oxley and
co-workers used the velocity modified temperature concept to describe material properties as a function of strain rate and temperature. The velocity modified temperature, defined as
T mod 1 log10 / 0 T

(1)

increases as the temperature increases and decreases as the strain

rate increases. and 0 are constant for a given material. The flow
stress is related to the strain through the power law
1 (T mod ) n ( T mod ) , where both the strength coefficient and the
strain hardening exponent are functions of the velocity modified
temperature. This leads to a complex relationship between the
deformation and the accompanying increase in temperature. In
Oxleys analysis, calculation of the total deformation work is
avoided and pointwise calculations of the stresses, depending
upon the local strain, strain rate and temperature are used to evaluate forces. Though the velocity modified temperature concept was
originally restricted to steels, it has been extended to aluminum
alloys also 20,21.
In recent years, an extensive amount of characterization of material properties at high strain rate and temperature has been carried out for use in simulation of high velocity impacts. The
Johnson-Cook model and the MTS model are widely used constitutive models for which the coefficients are available for a variety
of materials 22,23. Another model used widely by Maekawa and
co-workers 24 is a history dependent power law material model.
The equation below shows the structure of the Johnson-Cook
material model, and lists the coefficients to be used for representing the mechanical properties of materials 22.

AB n 1C ln

TT r
T m T r

(2)

In the above equation, is the true stress, is the true strain, is

the strain rate, 0 is the reference strain rate, T is the temperature
of the material, T r is the reference temperature and T m is the
melting temperature. Brief descriptions of the MTS material
model and the power law material model used by Maekawa and
co-workers are given in the Appendix. It is apparent that Oxleys
model can be extended to handle other materials by incorporating
these material models, especially since it is already complex
enough that a computer is required to carry out the calculations.
3.2 Calculation of the Shear Plane Temperature. In Oxleys analysis the heat partition coefficient at the primary shear
zone, , is calculated using the nondimensional number R T tan
evaluated at the average temperature along the shear plane, which
depends on an arbitrary parameter . The overall temperature rise
NOVEMBER 2003, Vol. 125 657

of the material in moving through the primary shear zone is found

using the total plastic work in the primary shear zone, accounting
for the heat conducted to the workpiece through the parameter .
It should be noted that the total power is calculated as F s V s , with
F s obtained from the material model for the temperature, strain
and strain rate corresponding to the midplane, not by the integration of the power required for deformation. The midplane temperature is obtained by scaling the total temperature increase in
the primary shear zone with a parameter . Iteration through the
calculations is used to converge to the correct temperature at the
middle of the shear zone.
Using a slightly different approach, we calculate the temperature of the midpoint of the shear plane by equating the fraction
of the total plastic work done up to the midplane of the primary
shear zone to the energy expended to increase the temperature of
the material. Point-wise evaluation of the properties of each of the
layers in the primary shear zone is used for the calculation. For
the Johnson-Cook material model the temperature of the shear
plane, T AB , is found from

T AB

TW

C p T
TT r
1
T m T r

dT 1 A AB

1C ln

s
0

B
n1
n1 AB

(3)

where T w , the temperature of the lower part of the shear zone, is

assumed equal to the incoming workpiece material temperature,
the partition coefficient is assumed to be constant and evaluated
at T AB , AB cos /23 cos()sin is the strain in the deformed material in the middle of the shear plane and s cV s /L is
the strain rate in the primary shear zone assumed to be constant
throughout the primary shear zone. The above equation can be
solved iteratively or explicitly for T AB . The temperature of the
upper part of the shear zone, T EF , can be obtained in a similar
manner as

T EF

TW

C p T
TT r
1
T m T r

dT 1 A EF

1C ln

s
0

Fig. 2 Temperature variation along the thickness of the primary shear zone predicted by the modified model for Al 2024T3. Cutting conditions: 8, t 1 160 m, V 1.31 ms and b 1
4.7 mm. 0.38, AB 0.63. Average temperature110.7C,
temperature of the middle of the shear zone117.7C and
0.54.

B
n1
n1 EF

(4)

tively. The average shear stress along the shear plane s is calculated as the shear force divided by the shear plane area. In Oxleys
analysis, the shear strength of the material at the strain and temperature of the midpoint of the shear zone is taken to be the shear
stress at the shear plane, from which the shear force is then calculated, without any guarantee that energy balance Eq. 6 will
be satisfied. Figure 3 shows the variation of the shear strength
through the thickness of the primary shear zone for Al 2024-T3.
Because of the competition between the temperature and strain
there is a maximum in the shear strength along the shear zone.
Also, the value of shear strength at the midplane is larger than the
value of s obtained from the above energy balance.
3.4 Calculation of the Hydrostatic Pressure at the Cutting
Edge and the Normal Stress on the Tool-Chip Interface.
From the equilibrium of the free surface near point A the hydrostatic pressure, p A , can be obtained as

in which EF 2 AB and is evaluated at T AB . Figure 2 shows

the distribution of the temperature in the primary shear zone as a
function of the distance through the thickness of the zone for Al
2024-T3 obtained from the above relations. Although the average
strain is considered to be at the middle of the shear zone, the
average temperature of the primary shear zone is different from
the temperature of the midplane AB due to the nonlinear dependence of the strength on the strain and temperature of the material.
3.3 Calculation of the Shear Force Along the Primary
Shear Zone. Since the deformation work per unit volume w
can be obtained as
w

EF

T EF C p T dT
w

(5)

we can obtain the shear force, F s , in a consistent manner by

setting
F s V s wVt 1 b 1

(6)

F s wVt 1
s
As
V sL
where V s is the shear velocity and V, t 1 and b 1 are the cutting
velocity, undeformed chip thickness and workpiece width, respec658 Vol. 125, NOVEMBER 2003

Fig. 3 Variation of shear strength along the thickness of the

primary shear zone predicted by the model for Al 2024-T3. Cutting conditions: 8, t 1 160 m, V 1.31 ms and b 1
4.7 mm. Note that the shear strength at the midplane of the
shear zone, k AB , is different from the energy equivalent shear
strength s .

Transactions of the ASME

p A s 12

(7)

This is similar to Oxleys expression for p A , but rather than using

the shear strength at the midplane of shear zone, energy equivalent
shear strength has been used. The requirement for equilibrium of
the shear zone dictates that the hydrostatic pressure at point B in
Fig. 1 be related to the hydrostatic stress at point A by
p A p B k u k l c

(8)

where k u and k l are the shear strengths at the upper and lower
boundaries of the primary shear zone. k u and k l can be obtained
knowing the temperature, strain and strain rate corresponding to
the upper and lower boundaries of the primary shear zone. Calculating p B from the above equation automatically takes into consideration the effects of the gradient in strain as well as the gradient in temperature along the alpha slip line, the latter being
neglected in Oxleys analysis. The angle between the resultant
force and the direction of the primary shear zone can be obtained
using the known pressure distribution and shear stress along the
shear plane.

tan1 12

c k u k l

4
2s

Fs
cos
cos

(10)

Fs
F
sin
cos
The slip line field equation describing the relation between the
pressure at point B and the normal stress on the tool-chip interface
can be written in terms of s as

n p B 2 s

(11)

3.5 Calculations Pertaining to the Secondary Shear Zone.

The length of contact H along the tool-chip interface can be
obtained from moment equilibrium of the chip, using the same
approach adopted by Oxley 12, as
H

b 1 L 2 2p A p B
N
3

(12)

where N is the normal force on the tool rake face.

Assuming uniform distribution of the normal force on the tool
rake face, the normal stress can be obtained as N/Hb 1 . Equating
this and n calculated from Eq. 10, the parameter c can be
obtained as
c

1
N

1 2
k l k u Hb 1 s
2

(13)

Assuming sticking friction over the tool-chip interface, the average value of the strain in the chip material can be considered to
be
int EF

int H

3V c

(14)

in which int V c / t 2 is the average strain rate in the secondary

shear zone and V c is the chip velocity.
Stevenson and Oxley 10 obtained Eq. 15 below to represent
the results of Boothroyds numerical calculations for the ratio of
the mean temperature increase of the chip ( c ) to the maximum
Journal of Manufacturing Science and Engineering

m
R Tt 2
0.060.195
c
H

0.5

0.5 log10

R Tt 2
H

(15)

The same approach is adopted here to determine the maximum

temperature increase along the tool-chip interface after obtaining
c from the friction force at the tool-chip interface and the chip
velocity. A constant is used to relate the mean temperature along
the tool-chip interface to the maximum temperature increase along
the tool-chip interface as T a v e T EF m .
Knowing the average temperature, strain and strain rate along
the tool-chip interface, the shear strength of the material (k chi p )
can be evaluated. The largest value of that satisfies the equality
of the shear strength obtained from the friction force ( int
F/Hb 1 ) and the one from the constitutive equation (k chi p ) is
chosen to be the correct value of .

4
(9)

Knowing F s obtained from Eq. 6, the normal and tangential

components, N and F, of the resultant cutting force on the tool
rake face can be obtained following Oxley as
N

temperature increase at the tool-chip interface ( m ). For a rectangular secondary shear zone, this ratio is expressed as a function
of the relative thickness of the secondary shear zone , chip
thickness (t 2 ), length of contact H and thermal number (R T ) as

Results and Discussion

4.1 Comparison of Extended Model With Oxleys Original

Model for 1020 Carbon Steel. We have developed a program
using the Maple symbolic mathematics package to carry out this
analysis. To verify the implementation of the model and to study
the effect of introducing the above changes into Oxleys original
model for plain carbon steels we also implemented Oxleys original model following the algorithm given by Oxley 12 in all
details. Figure 4 shows a comparison of the results of this implementation with the results published by Hastings et al. 16 for
0.2% carbon steel under one particular set of cutting conditions. In
all the graphs presented here experimental data points are marked
by symbols and lines represent the model predictions. It is observed that the correlation is very good, but not exact, especially
at low values of cutting speed. This may be attributed to uncertainties in the exact values of particular parameters such as ,
and c used by Hastings et al. to tune the model.
Figure 4 also compares the performance of the extended model
with that of Oxleys original model. It can be seen that even
though we have eliminated the degree of freedom provided by the
parameter , the predictions are slightly improved for cutting
force and chip thickness. This is one of the main contributions of
the present work and may be attributed to the self-consistency of
the model and the imposition of energy balance in the primary
shear zone for calculation of the shear force.
4.2 Comparison of Extended Model With Experimental
Data for Other Materials. We have used the extension of Oxleys analysis with the modifications described above to analyze
the machining of aluminum alloys 2024-T3, 6061-T6 and 6082T6, 1045 steel and copper. The constants of the Johnson-Cook
material model for 1045 steel and Al 6082-T6 were obtained from
Jaspers 25, for Al 2024-T3 from Johnson and Cook 22 and for
6061-T6 from Rule 26. The coefficients A, B and n in the
Johnson-Cook model were modified such that the model better
represents the data obtained from compression of those materials
to high strains 27,25. The modified constants of the JohnsonCook material model for each of the materials used in this analysis are listed in Table 1. The constants for the Maekawas material
model for 1045 steel and the MTS material model for copper were
obtained from Childs 24 and Follansbee and Kocks 28, respectively. These constants are listed in the appendix.
The experimental data used in the comparison for Al 2024-T3
and 6061-T6 were obtained by tube turning 29. The Al 2024-T3
tubes had a diameter of 70 mm and a wall thickness of 4.7 mm.
NOVEMBER 2003, Vol. 125 659

Table 1 Constants for the Johnson-Cook material model for

the materials used in this study after adjusting to fit stress
strain curve obtained from compression test
Material

T m (C)

A (MPa)

B (MPa)

Al 2024-T3*
Al 6061-T6**
Al 6082-T6***
AISI 1045****

502
582
582
1460

325
293.4
250
553.1

414
121.26
243.6
600.8

0.015
0.002
0.00747
0.0134

0.2
0.23
0.17
0.234

1
1.34
1.31
1

*Constants C, n and m adapted from 22

**Constants C, n and m adapted from 26
***Constants C, n and m adapted from 25
****Adapted from Jaspers 25

Fig. 4 Effect of the new modifications to Oxleys model on a

Cutting force, b Thrust force, and c Chip thickness. Cutting
conditions: 5, t 1 0.5 mm, b 1 4 mm and 0.2% carbon
steel. It can be seen that even though we have eliminated the
degree of freedom provided by the parameter , the predictions
are slightly improved for cutting force and chip thickness.

The diameter and the wall thickness for 6061-T6 tubes were 57
mm and 3.3 mm respectively. The inserts used were Kennametal
grade K68 with an edge radius of approximately 10 m as speci660 Vol. 125, NOVEMBER 2003

fied by the manufacturer. The experimental data for 6082-T6 was

obtained from Jaspers 25.
In all the analyses carried out for obtaining the results reported
below, , the ratio of the average temperature increase along the
tool-chip interface to the maximum temperature increase along
this interface, is set to 0.6. Figure 5 shows the results for Al
2024-T3 as a function of the cutting speed for two different undeformed chip thicknesses of 80 m and 160 m and rake angle
of 0. The predictions show good agreement with the experimental data. As expected, increasing the cutting speed decreases the
cutting forces and the chip thickness. It is found that increasing
the undeformed chip thickness increases the thickness of the primary shear zone and decreases the strain rate in the primary shear
zone. For instance, for the undeformed chip thickness and cutting
speed equal to 80 m and 1.31 m/s respectively, doubling the
undeformed chip thickness increases the thickness of the primary
shear zone by 66% and reduces the strain rate in the primary shear
zone by 37%. It is also found that increasing the cutting speed
reduces the thickness of the primary shear zone and increases the
strain rate in the primary shear zone. For instance, for the undeformed chip thickness and cutting speed equal to 80 m and 1.31
m/s respectively, doubling the cutting speed decreases the thickness of the primary shear zone by 19% and increases the strain
rate in the primary shear zone by 1.6 times. It can also be seen that
c, the ratio of the length of the shear zone to its thickness, increases with increase in the undeformed chip thickness and cutting speed.
Figure 6 shows that increase in the rake angle decreases the
cutting forces. The model overpredicts the cutting forces for negative rake angles. It is found that increasing the rake angle increases the thickness of the primary shear zone and decreases the
strain rate in the primary shear zone. For instance, beginning with
rake angle and undeformed chip thickness equal to 0 and 80 m
respectively, increasing the rake angle by 10 increases the thickness of the primary shear zone by 17% and decreases the strain
rate in the primary shear zone by 43%. This effect of the rake
angle seems to be more pronounced for higher undeformed chip
thicknesses. For instance for 160 m undeformed chip thickness
and 1.31 m/s cutting speed, the strain rate is reduced in half when
the rake angle is changed from 0 to 8. It can also be noted that
increasing the rake angle greatly decreases the c value. Figure 7
shows the capability of the model to capture the size effect. As
shown in Fig. 7a, the specific cutting energy increases as the
undeformed chip thickness decreases. Figure 7b shows the
variation in k chi p and int as a function of the shear plane angle .
It can be seen that the shear plane angle, which is the angle where
k chi p int , increases as the undeformed chip thickness increases.
It can also be seen that while the shear stress required for force
balance ( int ) at any given is just a little lower for the higher
undeformed chip thickness due to variation in int ), the shear
strength (k chi p ) of the chip material corresponding to the strain,
strain rate and temperature at the secondary shear zone is much
smaller at any given for larger undeformed chip thicknesses due
to the much higher temperature at the secondary shear zone.
Transactions of the ASME

Fig. 6 Comparison of predicted a Cutting force, b Thrust

force with experimental data for Al 2024-T3 29. Cutting conditions: V 1.31 ms and b 1 4.1 mm, different rake angles and
undeformed chip thicknesses.

Fig. 5 Comparison of predicted a Cutting force, b Thrust

force, and c Chip thickness with experimental data for Al
2024-T3 29. Cutting conditions: 0 and b 1 4.7 mm, different cutting speeds and undeformed chip thicknesses.

Figures 8 and 9 show the variation of the cutting forces and

chip thickness with cutting speed while cutting Al 6061-T6 and Al
6082-T6 respectively. The model predictions are in good agreement with experimental data. It was found that under the same
cutting conditions the values of c obtained for Al 6061-T6 and Al
2024-T3 are the greatest and the least respectively, and the values
Journal of Manufacturing Science and Engineering

of c obtained for Al 6082-T6 are between those for Al 6061-T6

and Al 2024-T3. For instance for 8, undeformed chip thickness equal to 160 m and cutting speed of 1.31 m/s, the values of
c obtained from the model are 9.22, 3.68 and 2.98 for Al 6061-T6,
Al 6082-T6 and Al 2024-T3 respectively. For Al 6061-T6 and Al
6082-T6, increase in the cutting speed and undeformed chip thickness both lead to decrease in the c value, while a similar change
causes an increase in the c value for Al 2024-T3. With respect to
the thickness of the primary shear zone the situation is little different; the greatest and the least values are found for Al 6082-T6
and Al 6061-T6 respectively. The decrease in the thickness of the
primary shear zone (L/c) and increase in the strain rate with
increase in the cutting speed and decrease in the undeformed chip
thickness obtained in the case of Al 2024-T3 are also found to be
the case for Al 6061-T6 and Al 6082-T6. While the ratio of the
length to the thickness of the primary shear zone c tends to
highlight changes in material properties, the trends in the thickness of the primary shear zone and the average strain rate through
this zone are dependent only upon the cutting conditions.
Figure 10 shows the comparison between the shear plane temperature predicted using the model and the experimental values
25 obtained using an infrared camera to measure the temperature
of point A in Fig. 1. The experimental value is found to be between T AB and T EF predicted by the model, but closer to T EF .
Figure 11 shows predictions of the specific cutting energy and
NOVEMBER 2003, Vol. 125 661

Fig. 7 a Comparison of the size effect predicted by the

model with the experimental data for Al 2024-T3 29, and b
The decrease in specific cutting energy with increase in undeformed chip thickness can be attributed to the increase in
shear angle, caused by the increase in temperature at the toolchip interface and the consequent decrease in k chip . Cutting
conditions: 8, V 2.62 ms and b 1 4.1 mm.

shear plane angle for cutting 1045 steel with different undeformed
chip thicknesses using different material models. The size effect
can again be seen in Fig. 11a. It can also be seen that the
Johnson-Cook material model predicts the variation of the specific
cutting force more accurately compared to the Oxley material
model and the Maekawa material model, especially for the larger
undeformed chip thicknesses. It can be seen from Fig. 11b that
while all the models tend to underpredict the shear plane angle,
the Johnson-Cook material model leads to predictions closest to
the experimental data. Note that in this case Oxleys original
model predicts specific cutting force and the shear plane angle
better compared to the modified model with different material
models but the difference is of the same order of magnitude as the
variation in the experimental data.
Figure 12 shows predictions of the cutting force and temperature at the tool-chip interface for cutting 1045 steel with different
cutting speeds using different material models. As can be seen in
Fig. 12a, the Johnson-Cook material model predicts the cutting
force most accurately. It can be noted that the cutting force predicted by the Maekawa material model is almost constant for different cutting speeds. Also, Oxleys material model gives good
agreement with experimental data only for low values of cutting
662 Vol. 125, NOVEMBER 2003

Fig. 8 Comparison of predicted a Cutting force, b Thrust

force, and c Chip thickness with experimental data for Al
6061-T6 29. Cutting conditions: 8, b 1 3.3 mm, different
cutting speeds and undeformed chip thicknesses.

speed. Figure 12b shows the effect of cutting speed on the toolchip interface temperature. The temperature was measured experimentally 30 using the tool-work thermocouple technique. The
measured values of temperature are between the maximum temperature and the average temperature at the tool-chip interface.
The experimental data is closer to the average temperature for the
Oxley and Maekawa material models, whereas it is closer to the
maximum temperature for the Johnson-Cook material model for
Transactions of the ASME

Fig. 10 Comparison of the temperature of the shear plane with

experimental data 25 for Al 6082-T6 for different undeformed
chip thicknesses. Cutting conditions: 6, V 6 ms and different undeformed chip thicknesses. Note that the measured
temperature seems to be close to the temperature at the end of
the primary shear zone.

measured using the tool-work thermocouple is not necessarily the

average temperature at the tool-chip interface and it depends on
the thermo-electrical properties of the tool. Note that if Oxleys
original model were to match the experimentally measured temperature closely, the cutting force would be underestimated. Also
note that the data shown in Fig. 11 are for a cutting speed of 7
m/s, which is beyond the range presented in Fig. 12. Thus it is
expected that if a similar comparison to that shown in Fig. 11
were made at any of the cutting speeds presented in Fig. 12, the
comparisons would be more in favor of the modified
models.
Figure 13 shows the variation of the cutting forces with cutting
speed for commercially pure copper. Since experimental values of
the forces do not reach steady state, representative averages of the
recorded forces are presented. Also the average strain at the toolchip interface was assumed to be a constant factor of the strain of
the deformed material leaving the primary shear zone. This constant value was calibrated for each undeformed chip thickness for
a single cutting speed and the obtained value was used in calculations for the same undeformed chip thickness and different cutting speeds. The average strain at the tool-chip interface for 100
m and 200 m with different cutting speeds obtained this way
were 2.3 EF and 3.1 EF respectively. The predictions of the extended model using MTS material model show good agreement
with the experimental data for different cutting speeds and undeformed chip thicknesses and as expected, increasing the cutting
velocity decreases the cutting forces.

5
Fig. 9 Comparison of predicted a Cutting force, b Thrust
force, and c Chip thickness with experimental data for Al
6082-T6 25. Cutting conditions: 8, different cutting speeds
and undeformed chip thicknesses.

higher cutting velocities. It can also be seen that temperatures

predicted by Maekawas material model are more sensitive with
respect to the cutting speed and follow the trend in experimental
data more closely compared to the other material models. It
should be noted that Stephenson 32 found that the temperature
Journal of Manufacturing Science and Engineering

Conclusions

Oxleys machining theory has been extended to use any of a

range of high strain rate constitutive models commonly used by
the impact physics community and for which the constants are
available for a wide range of materials. In the process, changes
have been made to the theory to improve its self-consistency. The
extended model has been used to analyze the machining of steel,
aluminum and copper. The model predicts all experimentally observed trends well, which is quite remarkable considering the fact
that the model parameters are typically obtained from experiments
at much lower values of strain, strain rate and temperature than
that encountered in machining. Among different material model
used the Johnson-Cook and Maekawas material models perform
best in terms of prediction of cutting forces and temperature reNOVEMBER 2003, Vol. 125 663

Fig. 11 Comparison of predicted a Specific cutting force,

and b Shear plane angle using different material models with
Oxleys original model 12 and experimental data 12 for
0.45% carbon steel. Cutting conditions: 5, V 7 ms and
different undeformed chip thicknesses. Oxleys original model
performs better compared to modified model with different material models. Among different material models used the
Johnson-Cook material model shows the most sensitivity with
respect to changes in undeformed chip thickness.

spectively. This provides a means of support to the suggestion

made by some researchers that machining itself could be used as a
very high strain, strain rate and temperature material test to generate coefficients for material models.
The predictions of the model show that increasing the cutting
speed and decreasing the undeformed chip thickness both decrease the thickness of the primary shear zone and increase the
strain rate. Also, increasing the rake angle increases the thickness
of the primary shear zone and decreases the strain rate. It is
also found that the ratio of the length to the thickness of the
primary shear zone c increases for Al 2024-T3 with increase in
cutting speed and the undeformed chip thickness while a similar
change causes a decrease in the c value for Al 6061-T6 and
Al 6082-T6.
We believe that this model of orthogonal cutting, coupled with
recent progress in the analysis of 3D machining operations with
arbitrary cutting edge geometry 34 may allow the development
of a software package that can provide a quick estimate of the
cutting forces, chip thickness and tool temperature for a variety of
machining processes.
664 Vol. 125, NOVEMBER 2003

Fig. 12 Comparison predicted a Cutting force, and b Temperature of the tool-chip interface using different material models with Oxleys original model 31 and experimental data 30
for 0.45% carbon steel. Cutting conditions: 5, t 1 0.2 mm,
b 1 3 mm and different cutting speeds. Oxleys original model
underpredicts both cutting force and temperature along the
tool-chip interface. The Johnson-Cook and Maekawa material
models perform better in prediction of the cutting force and
temperature respectively.

Though the modified Oxleys model has been shown to successfully predict a wide range of experimental observations for a
range of materials, it is not the case that the model is complete
from a point of view of understanding machining. Questions remain as to the role of energy minimization and the validity of
different assumptions used. A systematic series of studies is currently underway to further refine Oxleys machining theory. This
includes a study of the sensitivity of the model to each of the
assumptions used 35. Another study is aimed at refining the
assumptions based upon results of finite element analysis of machining using the same material and friction models used in the
theory 36. As mentioned by Oxley, better models of the geometry of the secondary shear zone and the normal and frictional
stress distribution at the tool-chip interface will help improve the
accuracy of the analysis and lead to improved understanding of
machining.

Acknowledgment
The authors would like to thank Thomas Samuel and Sivakumar Balasubramanian for providing us the results of experiments
conducted as part of their M.S. theses.
Transactions of the ASME

Fig. 13 Comparison of predicted a Cutting force, and b

Thrust force with experimental data 33 for copper. Cutting
conditions: 8, b 1 1.17 mm, different cutting speeds and
undeformed chip thicknesses.

Nomenclature
A yield strength in the Johnson-Cook material model
A s area of the primary shear zone
B strength coefficient in the Johnson-Cook material
model
b 1 width of the workpiece
C strain rate constant in the Johnson-Cook material
model
c ratio of the length to the thickness of the primary
shear zone
C p specific heat of work material
F friction force at the tool-chip interface
F s shear force along the nominal shear plane
H length of contact between tool and chip
k AB shear strength at the middle of the primary shear
zone
k l shear strength at the lower boundary of the primary
shear zone
k u shear strength at the upper boundary of the primary
shear zone
k chip shear strength at the tool-chip interface obtained from
the material model
L length of the shear plane
Journal of Manufacturing Science and Engineering

m temperature exponent in the Johnson-Cook material

model
N normal force at the tool-chip interface
n strain hardening exponent in Oxleys and the
Johnson-Cook material models
p A hydrostatic pressure at the free surface of the nominal
shear plane
p B hydrostatic pressure at the tool tip
R T thermal number of work material
T temperature
T a v e average temperature at the tool-chip interface
T mod velocity modified temperature
T M melting temperature
T r reference temperature
T w temperature of the uncut work material
t 1 undeformed chip thickness
t 2 chip thickness
V cutting velocity
V s shear velocity
V c chip velocity
w deformation work per unit volume in the primary
shear zone
normal rake angle
heat partition coefficient
the ratio of the secondary shear zone thickness to the
chip thickness
int shear strain rate at the tool-chip interface
strain
AB strain at the midplane of the primary shear zone
EF strain at the upper boundary of the primary shear
zone
strain rate
0 reference strain rate
s strain rate in the primary shear zone
shear plane angle
a constant calibration factor in Oxleys model equal
to the ratio of the work completed till the midplane
of the primary shear zone to the total plastic work
done in the primary shear zone
the strain rate exponent in the velocity modified temperature
the angle of the resultant force on the shear plane
with respect to the shear plane direction
c mean temperature rise of the chip due to the secondary shear
m maximum temperature rise along the tool-chip interface
density
flow stress
int shear strength at the tool-chip interface obtained from
equilibrium
s energy based shear strength of the primary shear
zone
a constant calibration factor equal to the ratio of the
average temperature increase to the maximum temperature rise at the tool-chip interface

Appendix
Mechanical Threshold Stress MTS Model. In this material model the flow stress is expressed as a function of a reference stress , the mechanical threshold stress, or flow stress at the
absolute zero temperature. can be resolved into two components
such that a t . The athermal stress a characterizes the
rate independent interactions of dislocations with long-range barriers such as grain boundaries and t characterizes the rate dependent interactions with short-range obstacles. The general form of
the flow stress can be expressed as a s( ,T) t , in which
NOVEMBER 2003, Vol. 125 665

s is a function of strain rate and temperature. The specific form of

the mechanical threshold model used in this study can be written
as 28

a a 1

kT ln 0 /
g 0 b 3

1/q

1/p

(16)

where the flow stress is expressed as a function of the current

threshold stress, , temperature, T, and strain rate . k is the
Boltzmanns constant, 0 is a constant, g 0 is the activation energy,
is the shear modulus, b is the magnitude of the Burgers vector,
and p and q are constants that represent the energy profile of
obstacles. The strain hardening rate can be obtained as

d
0 1
d

a
s a
tanh 2

tanh 2

(17)

In the case of OFE copper the corresponding values of parameters

used are as follows 28

0 239012 ln 0.034
s so
p2/3,

q1,

so

(19)

a 40 MPa,

so 6.21010 sec1 ,
k/b 0.823 MPa/K,
3

(18)

kT/ b 3 A

0 107 sec1 ,

42 GPa,

A 1 0.31,

g 0 1.6

and

so 900 MPa
Using the above values Eqs. 16 and 17 can be solved simultaneously to obtain the flow stress .
Maekawas Material Model for AISI 1045. Maekawa and
co-workers 24 proposed a history dependent material model for
0.45% carbon steel as follows for which the strength is expressed in MPa.

A 2 e aT

1000

M m 1

e aT/N 1

strain path

1000

m 1 /N 1

N1

(20)

where:
A 2 1350e 0.0011T 167e 0.00006 T275
N 1 0.17e 0.001T 0.09e

0.000015 T340 2

(21)
(22)

M 0.036
a0.00014
m 1 0.0024.

References
1 Armarego, E. J. A., 1998, A Generic Mechanics of Cutting Approach to
Predictive Technological Performance Modeling of the Wide Spectrum of Machining Operations, Mach. Sci. Technol., 22, pp. 191211.
2 Piispaanen, V., 1948, Theory of Formation of Metal Chips, J. Appl. Phys.,
19, pp. 876 881.
3 Ernst, H., 1938, Physics of Metal Cutting, in Machining of Metals, American Society for Metals, Cleveland, Ohio., pp. 134.
4 Merchant, M. E., 1945, Mechanics of the Metal Cutting Processes, ASME J.
Appl. Mech., 11, pp. A168 A175.
5 Lee, E. H., and Shaffer, B. W., 1951, The Theory of Plasticity Applied to a
Problem of Machining, ASME J. Appl. Mech., 73, pp. 405 413.
6 Dewhurst, P., 1978, On the Non-Uniqueness of the Machining Process,
Proc. R. Soc. London, Ser. A, 360, pp. 587 610.

666 Vol. 125, NOVEMBER 2003

7 Fang, N., Jawahir, I. S., and Oxley, P. L. B., 2001, A Universal Slip-Line
Model With Non-Unique Solutions for Machining With Curled Chip Formation and a Restricted Contact Tools, Int. J. Mech. Sci., 43, pp. 557580.
8 Hill, R., 1954, The Mechanics of Machining: A New Approach, J. Mech.
Phys. Solids, 3, pp. 4753.
9 Fenton, R. G., and Oxley, P. L. B., 1968 69, Mechanics of Orthogonal
Machining: Allowing for the Effects of Strain-Rate and Temperature on ToolChip Friction, Proc. Inst. Mech. Eng., 183, pp. 417 438.
10 Stevenson, M. G., and Oxley, P. L. B., 19691970, An Experimental Investigation of the Influence of Speed and Scale on the Strain-Rate in a Zone of
Intense Plastic Deformation, Proc. Inst. Mech. Eng., 18431, pp. 561576.
11 Oxley, P. L. B., and Hastings, W. F., 1977, Predicting the Strain-Rate in the
Zone of Intense Shear in Which the Chip is Formed in Machining From the
Dynamic Flow Stress Properties of the Work Material and the Cutting Conditions, Proc. R. Soc. London, Ser. A, 356, pp. 395 410.
12 Oxley, P. L. B., 1989, The Mechanics of Machining: An Analytical Approach
to Assessing Machinability, E. Horwood, Chichester, England.
13 Lin, G. C. I., and Oxley, P. L. B., 1972, Mechanics of Oblique Machining:
Predicting Chip Geometry and Cutting Forces From Work Material Properties
and Cutting Conditions, Proc. Inst. Mech. Eng., 186, pp. 813 820.
14 Arsecularatne, J. A., Mathew, P., and Oxley, P. L. B., 1995, Prediction of
Chip Flow Direction and Cutting Forces in Oblique Machining With Nose
Radius Tools, Proc. Inst. Mech. Eng., B209, pp. 305315.
15 Young, H. T., Mathew, P., and Oxley, P. L. B., 1994, Predicting Cutting
Forces in Face Milling, Int. J. Mach. Tools Manuf., 346, pp. 771783.
16 Hastings, W. F., Mathew, P., and Oxley, P. L. B., 1980, Machining Theory for
Predicting Chip Geometry, Cutting Forces, etc., From Work Material Properties and Cutting Conditions, Proc. R. Soc. London, Ser. A, 371, pp. 569587.
17 Palmer, W. B., and Oxley, P. L. B., 1959, Mechanics of Metal Cutting, Proc.
Inst. Mech. Eng., 173, pp. 623 654.
18 Boothroyd, G., 1963, Temperatures in Orthogonal Metal Cutting, Proc. Inst.
Mech. Eng., 177, pp. 789 802.
19 Hastings, W. F., and Oxley, P. L. B., 1976, Minimum Work as a Possible
Criterion for Determining the Frictional Conditions at the Tool/Chip Interface
in Machining, Philos. Trans. R. Soc. London, 282, pp. 565584.
20 Bao, H., and Stevenson, M. G., 1976, A Basic Mechanism for Built-up Edge
Formation in Machining, CIRP Ann., 251, pp. 5357.
21 Kristyanto, B., Mathew, P., and Arsecularatne, J. A., 2000, Determination of
Material Properties of Aluminum From Machining Tests, ICME 2000
Eighth Int. Conf. On Manuf. Eng., Sydney, Australia, August 2730.
22 Johnson, G. J., and Cook, W. H., 1983, A Constitutive Model and Data for
Metals Subjected to Large Strains, High Strain Rates and High Temperatures,
Proceedings of the 7th International Symposium on Ballistics, pp. 541547.
23 Tanner, A. B., McGinty, R. D., and McDowell, D. L., 1999, Modeling Temperature and Strain Rate History Effects in OFHC Cu, Int. J. Plast., 15, pp.
575 603.
24 Private communication between Maekawa and Childs as appeared in Childs, T.
H. C, Material Property Requirements for Modeling Metal Machining,
1997, Colloque C3, Journal de physique, III: XXIXXXIV.
25 Jaspers, S. P. F. C., 1999, Metal Cutting Mechanics and Material Behavior,
PhD thesis, Technische Universiteit Eindhoven.
26 Rule, W. K., 1997, Numerical Scheme for Extracting Strength Model Coefficients From Taylor Test Data, Int. J. Impact Eng., 19, pp. 797 810.
27 Kobayashi, S., and Thomsen, E. G., 1959, Some Observations on the Shearing Process in Metal Cutting, ASME J. Ind., pp. 251261.
28 Follansbee, P. S., and Kocks, U. F., 1988, A Constitutive Description of the
Deformation of Copper Based on the Use of the Mechanical Threshold Stress
as an Internal State Variable, Acta Metall., 36, pp. 8193.
29 Samuel, T., 2000, Investigation of the Effectiveness of Tool Coatings in Dry
Turning of Aluminum Alloys, MS thesis, Wichita State University.
30 OECD-CIRP, 1966, Proceedings of the Seminar on Metal Cutting.
31 Hastings, W. F., Oxley, P. L. B., and Stevenson, M., 1974, Predicting Cutting
Forces, Tool Life etc., Using Work Material Flow Stress Properties Obtained
From High-Speed Compression Tests, Proc. Int. Conf. on Production Engineering, Tokyo, p. 528.
32 Stephenson, D. A., 1991, Tool-Work Thermocouple Temperature Measurements: Theory and Implementation Issues, PED-Vol 55, Sensors and Signal
Processing for Manufacturing, ASME 1992, pp. 8195.
33 Balasubramanian, S., 2001, The Investigation of Tool-Chip Interface Conditions Using Carbide and Sapphire Tools, MS thesis, Wichita State University.
34 Adibi-Sedeh, A. H., Madhavan, V., and Bahr, B., 2002, Upper Bound Analysis of Oblique Cutting With Nose Radius Tools, Int. J. Mach. Tools Manuf.,
429, pp. 10811094.
35 Adibi-Sedeh, A. H., and Madhavan, V., 2002, Effect of Some Modifications
to Oxleys Machining Theory and the Applicability of Different Material Models, Mach. Sci. Technol., 63, pp. 377393.
36 Adibi-Sedeh, A. H., and Madhavan, V., 2003, Understanding of Finite Element Analysis Results Under the Framework of Oxleys Machining Model,
Proceedings of the 6th CIRP International Workshop on Modeling of Machining Operations, Hamilton, Canada, May 20, pp. 115.

Transactions of the ASME