ARTICLE IN PRESS

International Journal of Machine Tools & Manufacture 47 (2007) 401–411

www.elsevier.com/locate/ijmactool

Modelling the effects of tool-edge radius on residual stresses when

orthogonal cutting AISI 316L
Mohamed N.A. Nasra,b,, E.-G. Nga, M.A. Elbestawia,b
a
McMaster Manufacturing Research Institute (MMRI), McMaster University, 1280 Main St. W., Hamilton, Ontario, Canada L8S 4L7
b
Department of Mechanical Engineering, McMaster University, 1280 Main St. W., Hamilton, Ontario, Canada L8S 4L7
Received 27 January 2006; received in revised form 2 March 2006; accepted 6 March 2006
Available online 22 May 2006

Abstract

Tool-edge geometry has significant effects on the cutting process, as it affects cutting forces, stresses, temperatures, deformation zone,
and surface integrity. An Arbitrary-Lagrangian–Eulerian (A.L.E.) finite element model is presented here to simulate the effects of
cutting-edge radius on residual stresses (R.S.) when orthogonal dry cutting austenitic stainless steel AISI 316L with continuous chip
formation. Four radii were simulated starting with a sharp edge, with a finite radius, and up to a value equal to the uncut chip thickness.
Residual stress profiles started with surface tensile stresses then turned to be compressive at about 140 mm from the surface; the same
trend was found experimentally. Larger edge radius induced higher R.S. in both the tensile and compressive regions, while it had almost
no effect on the thickness of tensile layer and pushed the maximum compressive stresses deeper into the workpiece. A stagnation zone
was clearly observed when using non-sharp tools and its size increased with edge radius. The distance between the stagnation-zone tip
and the machined surface increased with edge radius, which explained the increase in material plastic deformation, and compressive R.S.
when using larger edge radius. Workpiece temperatures increased with edge radius; this is attributed to the increase in friction heat
generation as the contact area between the tool edge and workpiece increases. Consequently, higher tensile R.S. were induced in the near-
surface layer. The low thermal conductivity of AISI 316L restricted the effect of friction heat to the near-surface layer; therefore, the
thickness of tensile layer was not affected.
r 2006 Elsevier Ltd. All rights reserved.

Keywords: Finite element modelling; A.L.E.; Tool-edge radius; Residual stresses

1. Introduction eventually wear out or break off and cutting will be

Tool-edge geometry has a significant effect on almost all
cutting parameters such as cutting forces, tool life, surface
integrity (surface finish and residual stresses (R.S.)), and
temperatures. Therefore, understanding the effects of
different edge preparations and predicting them, if
possible, is of a crucial importance; especially that using
such tools is mandatory in certain cases. Furthermore, even
if cutting starts with an up-sharp tool, the tool tip will
Corresponding author. McMaster Manufacturing Research Institute
(MMRI), Department of Mechanical Engineering, 1280 Main St. W.,
Hamilton, Ontario, Canada L8S 4L7. Tel.: +1 905 525 9140x24544;
fax: +1 905 521 9742.
E-mail address: ahmedmn@mcmaster.ca (M.N.A. Nasr).
continued with a non-sharp edge. The effects of different
edge preparations on induced R.S. are critically important,
as nearly all machining operations induce R.S. Further-
more, R.S. play a crucial role in controlling part
performance as they affect its properties in many different
ways [1]. The current work is concerned with rounded-edge
tools and the effects of edge radius on induced R.S.
Experimental investigations have been always an im-
portant method for studying different phenomena in metal
cutting. Some experimental studies on the effects of tool-
edge radius on R.S. could be found in [2–5]. Thiele and
Melkote [2] have found that tool-edge radius has a
significant effect on surface R.S. and microstructure when
finish hard turning AISI 52100 steel. R.S. in the axial and
circumferential directions were more compressive with

0890-6955/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijmachtools.2006.03.004
 ARTICLE IN PRESS
402 M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411
larger edge radius. Microstructural analysis showed ther-
mally induced phase transformation, white layer, at all
feeds when using edge radius of 122 mm and only at high
feeds when using sharp tools (radius of 23 mm). Jang et al.
[3] studied the effects of different machining parameters
(cutting speed, feed rate, depth of cut, and tool-edge
radius) on surface R.S. when turning AISI 304 stainless
steel. Tool-edge radius was found to have the most
significant effect on R.S., where larger radius resulted in
higher surface tensile R.S. in cutting direction. The authors
reported that the effect of edge radius on inducing tensile
R.S. in stainless steels is more noticeable than in other
steels, and attributed this to the low thermal conductivity
of stainless steels. Liu and Barash [4] found that larger edge
radius results in less near surface tensile R.S. in the cutting
direction when orthogonal and oblique cutting AISI 1008
steel. The same result was found by Arunachalam et al. [5]
when machining (facing) Inconel 718.
Analytical modelling has been used to understand the
effects of tool-edge geometry on different cutting para-
meters in orthogonal cutting, such as in [6,7]. Manju-
nathaiah and Endres [6] presented an analytical model for
tools with edge radius and verified it by orthogonal cutting
70–30 brass and mild steel. Both cutting and ploughing
force components increased with edge radius. Higher
strain and strain rate were obtained, which also resulted
in higher specific cutting energy. It is important to note
that the proposed model did not consider the stagnation
zone or build-up edge phenomenon. Ren and Altintas [7]
studied the effects of chamfer angle and validated their
model by high-speed orthogonal cutting P20 mold steel.
Cutting force components increased with chamfer angle,
with more significant increase in the thrust component.
Chip thickness was almost constant for different chamfer
angles, i.e. constant shear angle, which was explained in
terms of having a dead-metal zone that acted as the
effective cutting edge making different edges almost the
same. The main disadvantage of analytical modelling is its
questionable and limited applicability as it is based on
simplifications and assumptions that hardly agree with
reality.
Finite element (F.E.) analysis has played an important
role in simulating and understanding the metal cutting
process by having an insight looking at what is going on
during cutting, which cannot be achieved by experimental
or analytical methods. Both the Lagrangian and Eulerian
F.E. techniques have been used in studying the effects of
tool-edge preparation in orthogonal cutting. Kim et al. [8]
used the Eulerian approach to study the effects of tool-edge
radius when orthogonal cutting 0.2% carbon steel. Cutting
forces followed the same trend as in [6,7], the deformation
zone extended deeper and wider into the workpiece, and
the effective strain rate decreased with edge radius. Cutting
temperatures increased with edge radius, which was
attributed to more heat generation due to friction between
the tool tip and workpiece as the contact area increased.
No apparent change was noticed in chip thickness. Yen
et al. [9] examined different edge preparations (honed and
chamfered edges) while modelling orthogonal cutting of
0.2% carbon steel using Lagrangian formulation with
continuous remeshing. More chip curling and larger
sticking region were noticed with larger edge radius. Chip
thickness increased while the shear angle decreased with
edge radius. Tool-tip temperature decreased and then
increased with edge radius, having a minimum at a certain
edge radius. This was because larger edge radius results in
larger contact area between the tool and workpiece, which
increases heat generation due to friction but at the same
time increases heat dissipation into the tool. These two
contradicting effects balance out, resulting in an optimum
radius for minimum tool temperature.
Each of the Lagrangian and Eulerian techniques has its
own advantages and drawbacks that make it suitable for
modelling certain cases and unsuitable for others, where
the advantages of one of them are the disadvantages of the
other and vice versa. The Eulerian approach handles
material flow around tool tip in a perfect way without the
need to define a failure criterion, which is mandatory for
the Lagrangian approach and adds to its modelling errors.
However, the chip shape has to be known a priori, which
represents a huge drawback to the Eulerian technique.
Furthermore, R.S. could not be estimated because the
material elastic behaviour is not considered [10]. On the
other hand, the chip is generated automatically in the
Lagrangian approach as the mesh totally represents the
underlying material but high element distortion may
terminate the analysis. In Lagrangian formulation, a
parting line, along which chip separation takes place,
needs to be defined. This makes it unsuitable for modelling
tools with blunt edges because such line cannot be
predicted. Furthermore, remeshing has to be used with a
very fine mesh around the tool edge, which makes the
process very expensive numerically.
The Arbitrary-Lagrangian–Eulerian (A.L.E.) technique
is a relatively new modelling technique that represents a
combination of the Lagrangian and Eulerian techniques
without having their drawbacks. It was first introduced in
modelling the cutting process by the end of the last decade,
and some of the recent A.L.E. cutting models were
presented in [10–13]. Movahhedy et al. [11] presented the
first A.L.E. model to study the effects of tool-edge
preparation. They studied the effects of chamfer angle on
orthogonal cutting of P20 mold steel. The model succeeded
in simulating the cutting process showing a dead-metal
zone trapped under the chamfer and filling almost all of the
missing edge, which was also noticed experimentally.
Cutting forces increased with chamfer angle, especially
the thrust component. The size and shape of the shear zone
were not affected greatly, which was attributed to the
presence of the dead-metal zone that acted as the effective
cutting edge and made the chip formation process almost
the same for different cases.
Austenitic stainless steels are widely used in many
applications, such as chemical industries and nuclear
 ARTICLE IN PRESS
M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411 403

power stations, because of their high mechanical properties Table 1

and high corrosion and oxidation resistances. At the same Physical properties of AISI 316L and Kennametal K313 [18]
time, they are considered difficult-to-machine materials Property AISI 316L K313
because they experience high strain hardening effects as
well as low thermal conductivity [14,15]. The commercial Modulus of Elasticity (GPa) 193 615
F.E. software package ABAQUS/Explicit 6.4-1 was used in Poisson’s ratio 0.27–0.3 (0.285 used) 0.22
Density (kg/m3) 8000 14,900
the current work to simulate the effects of tool-edge radius
Specific heat capacity 500 138
on machining-induced R.S. while orthogonal dry cutting (J/kg 1C)
stainless steel AISI 316L (160–170 HV). A.L.E. was used Thermal conductivity 16.2 at 100 1C 79
because of its superior advantages in modelling such a case. (W/m 1C)
One of the main motivations for the current work is that no 21.4 at 500 1C
Melting temperature (1C) 1375–1400 (1385 used) —
clear physical explanation was given in any of the previous Linear coefficient of thermal 19.9  10
6 —
works. The current work explains the effects of tool-edge expansion (mm/mm 1C)
radius on induced R.S. based on the phenomenon of
stagnated metal zone, which gives a reasonable acceptable
physical explanation for the presented results and the
results available in the literature.
Table 2
J–C parameters for AISI 316L [20]
2. Finite element modelling
:
A (MPa) B (MPa) n C eo m
A plane strain A.L.E. F.E. model was built using
305 1161 0.61 0.01 1 0.517
ABAQUS/Explicit 6.4-1 to simulate orthogonal dry cutting
of stainless steel AISI 316L with continuous chips and
studying the effects of tool-edge radius on induced R.S.
The analysis was done in two steps, the first step simulated
J–C parameters for AISI 316L are given in Table 2:
the cutting process while the second one simulated the
: !!   m 
stress-relaxation process to obtain R.S. where the work- 
pl n  epl y
yref
piece was left to cool down to room temperature. Coupled s¼ AþB e 1 þ C ln : 1
,
eo ymelt
yref
temperature–displacement analysis was used in both steps
to allow for temperature-dependent properties and heat (1)
transfer. The current model was proved to be valid by where s is the material
:
pl
flow stress (MPa), e the equivalent
:
estimating R.S. profiles that followed the same trend found plastic strain, epl the equivalent plastic strain rate, eo the
experimentally in the literature, in [2,14,16,17], when reference plastic strain rate, at yref , A, B, n, m the material
cutting different workpiece materials. Furthermore, the constants, y the current temperature, ymelt the melting
in-depth R.S. profile for sharp tool was compared to that temperature, and yref the reference temperature at which
obtained experimentally in [15] under the same cutting material constants are determined.
conditions; both profiles matched in an acceptable way.
The model has also captured the phenomenon of dead-
2.2. Cutting conditions
metal zone, where its size and shape were found to follow
the same trend found in the literature [6,9,11].
The current work is based on simulating dry orthogonal
cutting with four different edge radii (R): 0.02, 0.05, 0.075,
2.1. Material properties and 0.1 mm to study the effects of tool-edge radius on
induced R.S. Throughout this work, these tools are
The modelled workpiece material is stainless steel AISI referred to as R20, R50, R75, and R100, respectively.
316L (160–170 HV) with an approximate grain size of The used cutting conditions were as follows:
50 mm [15], while the cutting tool is an uncoated
Cutting speed ¼ 125 m/minClearance angle ¼ 51
Kennametal K313 carbide insert. The physical properties
Rake angle ¼ 01 Uncut chip thickness ¼ 0.1 mm
of used materials are given in Table 1.
The Johnson–Cook (J–C) plasticity model, Eq. (1),
available in ABAQUS/Explicit, was used to model the R20 represents the sharp tool in the current simulations
workpiece material. This model is suitable for modelling for model validation; this is supported by having a good
cases with high strain, strain rate, strain hardening, and matching between its induced R.S. profile and the
non-linear material properties, which represent the main experimental profile for sharp tool presented by M’Saoubi
numerical challenges when modelling metal cutting [19]. et al. [15] (see Section 3.1). Besides, 0.02 mm fits well in the
Furthermore, it has been widely used in modelling the commercial range for tool-edge radius of the simulated
cutting process and has proved its suitability [12,20]. The tool.
 ARTICLE IN PRESS
404 M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411

edge as if it is a fluid. This eliminates one of the sources

of numerical errors in the modelling process;
2. there is no need to define a parting line along which chip
separation occurs. This is very crucial for the current
case since the parting line cannot be predicted when
having a non-sharp tool. Furthermore, even if a parting
line could be predicted it still has its own problems
whether node separation or element deletion will be
used;
3. no crack is formed ahead of the cutting edge which helps
modelling the cutting process in a more realistic way;
4. although a fine mesh is required around the tool tip to
grasp high stress gradients, it is still much coarser than
that required when using Lagrangian formulation with
Fig. 1. Boundary conditions, partitioning scheme, and material flow. continuous remeshing.

2.3. Model description Region D gains its importance by being responsible for
chip formation, which makes its mesh design very critical.
2.3.1. Geometry and boundary conditions An initial chip shape has to be assumed to start the
Fig. 1 shows the basic geometry of the used F.E. model analysis, but it has to be noted that this initial shape does
with the applied boundary conditions, where cutting takes not affect the final chip shape or the mechanics of cutting.
place in plane 1–2 under plane strain condition. It also It only represents the shape of the chip tip as can be seen
shows the material flow between the four sections (A, B, C, clearly in Fig. 2, where the initial and final chip shapes are
and D) of the workpiece, which are essential to applying shown. Assuming an initial chip shape is not so simple and
A.L.E. as will be explained later. It has to be noted that the it may need some iterations. It has two critical dimensions,
tool cannot be moved when using A.L.E. and the cutting the corner radius (r) and the deformed chip thickness (tc),
velocity has to be applied to the workpiece instead. shown in Fig. 1, which may cause the analysis to be
terminated due to excessive element distortion. These two
2.3.2. Mesh motion and chip formation dimensions are assumed based upon the uncut chip
A.L.E. has been introduced in modelling the cutting thickness (t) and material properties; then an initial
process because it combines the advantages of both analysis is run. According to the resulting chip shape and
Lagrangian and Eulerian techniques without having their mesh distortion, the trend for changing (r) and (tc) will be
drawbacks. This is done by using a combination of known based upon experience. It is important to note that
Lagrangian and Eulerian regions in the same model the number of required iterations to reach a suitable initial
[10,11], where the decision between having a Lagrangian chip shape is far away smaller than that required in a pure
or an Eulerian region depends on the required modelling Eulerian model. This is because the chip still has the ability
features. In the current model, regions A, C, and D were
modelled as Lagrangian regions with adaptive meshing,
where the mesh is attached to the underlying material while
continuous remeshing is being applied. Since the mesh
deforms with the underlying material, free surfaces can be
modelled properly and boundary conditions can be applied
in a simple way. Consequently, automatic chip formation
takes place without the need to know the correct chip shape
a priori as in Eulerian formulation. Furthermore, contin-
uous remeshing helps avoiding excessive element distor-
tion, which represents one of the main drawbacks when
using Lagrangian modelling [10].
Region B was modelled as an Eulerian region, where the
mesh is totally fixed in space and the material flows
through it. This solves the problems encountered when
using Lagrangian formulation around the tool tip where
chip separation starts. This is because when the Eulerian
formulation is used:
1. there is no need to define a failure criterion for chip
formation because the material flows around the tool Fig. 2. Automatic chip formation (F.E. mesh).
 ARTICLE IN PRESS
M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411
to reshape itself and grow automatically, where the only
limitation here is element distortion that takes place
between the initial and final chip shapes. Therefore, if the
initial shape is far away from the correct final shape,
excessive element distortion may take place and may
terminate the analysis.
2.3.3. Analysis
An explicit, dynamic, thermal–mechanical coupled
analysis was used in both simulation steps: the cutting
step and the stress-relaxation step. Metal cutting simula-
tions are usually solved using explicit integration methods,
although some few works are available in the literature that
used implicit methods, for example [21–23]. In explicit
integration, a system of de-coupled differential equations is
solved based on an element-by-element basis, in which only
the element stiffness matrix is formulated and saved
without the need to the global stiffness matrix. On the
other hand, the global stiffness matrix has to be formulated
and saved in implicit integration, and the whole system of
differential equations has to be solved simultaneously.
Therefore, explicit methods are computationally more
efficient especially when non-linearity is encountered. This
becomes more evident in thermally coupled analysis, as in
metal cutting, because structural and thermal variables are
solved simultaneously. On the other hand, explicit integra-
tion is conditionally stable because the critical time step
depends on the minimum element size and the speed of
wave propagation, while implicit integration is uncondi-
tionally stable [24,25].
2.3.3.1. Element type. Plane stain, quadrilateral, linearly
interpolated, and thermally coupled elements with auto-
matic hourglass control and reduced integration were used
in the current analysis. For A.L.E. formulation, ABAQUS/
Explicit only supports linearly interpolated elements with
reduced integration, which are suitable for modelling the
cutting process when a reasonable number of elements is
used. At the same time, thermally coupled elements were
used to account for temperature-dependent material
properties, and hourglass control was mandatory due to
high element deformation.
2.3.3.2. Tool–workpiece contact and heat genera-
tion. There are two sources for heat generation in metal
cutting, material plastic deformation and friction. Most of
the plastic deformation energy is converted into heat,
which is usually between 85% and 95% [26]. This
percentage was taken as 90% in the current work, which
has been widely used in the literature [21,26–28]. An initial
temperature of 20 1C was assumed; heat radiation and
convection were neglected in the cutting step as they are
negligible when compared to conduction. In the stress-
relaxation step, where the workpiece was left to cool down
to room temperature, convection to air was considered
with a coefficient of heat convection of 10 W/m2 1C and
405
sink temperature of 20 1C. No gap conductance was
considered in the current model.
Friction is one of the most complicated issues in metal
cutting that is still not completely understood. Different
friction models have been used in the literature, some
examples could be found in [12,21,27]. The simple
Coulomb friction model, which assumes constant coeffi-
cient of friction along the tool–workpiece contact length,
was used in the current work. Although it is a simple
unrealistic model, it has been widely used in metal cutting
simulations [12,26,29] because the real coefficient of
friction cannot be measured any way; besides, it gives
reasonable results. A coefficient of friction of 0.2 was
assumed in the current work, which is supported by having
good agreement between the estimated F.E. residual stress
profile and that obtained experimentally by M’Saoubi et al.
[15]. Based on material properties and using Eq. (2), 40.3%
of the friction heat goes into the workpiece. It has to be
noted that contact definitions were deactivated in the
stress-relaxation step and the workpiece was left to cool
down to room temperature.
% of friction heat into workpiece ¼ H w:p: =ðH w:p: þ H chip Þ,
(2)
where H ¼ ðKrCÞ0:5 , K is the thermal conductivity (W/m 1C),
r the density (kg/m3), and C the specific heat capacity
(J/kg 1C)
3. Results and discussion
3.1. Model validation
The presented F.E. model, as any numerical model, was
firstly validated. Current validation was based on compar-
ing the R.S. F.E. results to experimental results available in
the literature in [2,14–16]. Although different workpiece
materials (AISI 316L, AISI 304, and AISI 52100) and
different cutting conditions were used, they all supported
the current results by showing the same trend. M’Saoubi
et al. [15] used exactly the same materials and cutting
conditions as those used here. Besides, they were the only
work that presented an in-depth R.S. profile for AISI 316L
(170 HV), while studying the effects of different cutting
parameters on R.S. when orthogonal dry cutting AISI
316L using sharp-edged tools. Therefore, their results were
used to validate the current in-depth R.S. profile, which is
thought to be a strong evidence of validation for the
current model. X-ray diffraction was used for R.S.
measurements and the surface was electrolytically polished
to determine in-depth stresses. Furthermore, the current
model has correctly modelled the stagnation-zone phenom-
enon in a way that matched the experimental results
found in the literature, in [6,9,11], as will be explained in
Section 3.3.
Fig. 3 shows a schematic diagram for the current F.E.
model with the corresponding experimental set-up used by
M’Saoubi et al. [15]. R.S. parallel to the cutting direction

406
tool
cutting direction
3 1
workpiece
(a) F.E. model
Fig. 3. F.E. model and corresponding experimental set-up. (a) F.E.
model, (b) experimental set-up.
Residual stress in cutting direction "RS11" (MPa)
1000.00
800.00
600.00
400.00
200.00
0.00
-200.00
0.00
Depth from machined surface (microns)
Fig. 4. F.E. and experimental RS11 profiles for sharp tool (model
validation).
(RS11) and perpendicular to the cutting plane (RS33) were
found to be the only significant components, while the
shear stress component (RS12) and the feed direction
component (RS22) were negligible. This matches the
experimental results presented in [15]; besides, the same
trend was found by Outeiro et al. [14] and Jang et al. [3]
after turning AISI 316L and AISI 304, respectively.
Fig. 4 shows the R.S. profile parallel to the cutting
direction (RS11) for the simulated sharp tool (R20) and the
corresponding experimental profile obtained by M’Saoubi
et al. [15]. The F.E. profile was based on taking the average
of 10 elements at each depth; these elements were at the
mid-length of the machined workpiece to avoid any
transient effects. Although both curves do not match
exactly, which was expected a priori, the F.E. model
correctly estimated the in-depth R.S. profile showing the
same trend and starting with almost the same surface value
as the experimental one. It has to be noted that similar
profiles were obtained in [3,14,27] when cutting AISI 316L
and AISI 304 under different cutting conditions. In
general, an exact match between F.E. and experimental
results could not be expected because of the different
sources of errors in each one of them. The main sources of
errors encountered in F.E. modelling could be summarized
ARTICLE IN PRESS
M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411
as: material modelling, numerical integration and inter-
workpiece cutting (1)
2 polation, assumed friction condition, and remeshing errors.
Furthermore, simulations were run with a constant tool-
axial (3)
edge radius, while practically the tool edge may wear out or
radial (2) break down during cutting. On the other side, the main
sources of errors in experimental results are those
tool
encountered in R.S. measurements, especially in measuring
(b) Experimental setup
the etched depth. Besides, the actual workpiece material is
not 100% homogeneous, while it was considered to be a
pure homogeneous material in the F.E. model. The
predicted R.S. profile had higher tensile and compressive
magnitudes than the experimental profile throughout the
workpiece depth. It is important to note that both profiles
Experimental [15] (F.E. and experimental) show a state of equilibrium
A.L.E. (tool R20)
between tensile and compressive R.S, as it should be. The
error bars in Fig. 4 represent the error in measuring R.S
using X-ray diffraction method for the experimental
profile, while they represent the difference between the
maximum and minimum stress values in the averaged
elements for the F.E. profile.
The R.S. profile in the axial direction during experi-
mental cutting, which corresponds to RS33 in the F.E.
model, was not presented in [15]. Only its maximum and
minimum values were stated as +700 and
120 MPa,
respectively. The corresponding values arising from the
current simulations are +690 and
159 MPa, respectively,
which supports the validity and reliability of the presented
200.00 400.00 600.00 800.00 model.
3.2. Effects of tool-edge radius
Figs. 5a and b show the estimated in-depth R.S. profiles
for RS11 and RS33, respectively, for different edge radii.
The magnitude of R.S., in both profiles, increased with
tool-edge radius in both the tensile and compressive
regions. Thiele et al. [2] and Shivpuri et al. [16] have also
reported an increase in the magnitude of R.S. when using
larger edge radius while turning steel AISI 52100. In the
current case, RS11 had slightly higher magnitudes than
RS33 in both tensile and compressive layers in all
simulated cases. The thickness of the tensile R.S. layer
was almost the same for both components (RS11 and
RS33) and independent of the edge radius with a value of
14075 mm. The location of maximum compressive R.S. for
both components went deeper into the workpiece with
larger edge radius, which has been also reported in [2,16].
It is believed that RS33 has aroused as a direct result to
the existence of RS11 because in orthogonal cutting,
deformation takes place only in the cutting plane (plane
1–2, Fig. 1). No previous attempts are found in the
literature discussing the reasons for having such stresses; all
the available work focussed on RS11 as it could be directly
related to and explained in terms of cutting parameters.
According to [14,15], RS33 usually shows the same trend as
RS11when cutting AISI 316L, but this is not guaranteed.
Because of the aforementioned reasons, the current work
will focus on RS11 in the detailed discussion.
 ARTICLE IN PRESS
M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411 407

1600.00 800.00
Residual stress in cutting direction "RS11" (MPa)

Depth from machined surface

Tool edge: zero (at surface)
R20 (sharp) 25 microns
R50
50 microns
1200.00 R75
75 microns
R100
600.00

Workpiece temperature (C)

800.00

400.00
400.00

0.00
200.00
tensile
layer
-400.00
0.00 200.00 400.00 600.00 800.00
(a) Depth from machined surface (microns) 0.00
20.00 40.00 60.00 80.00 100.00
1200.00
(a) Tool-edge radius (microns)
Tool edge:
R20 (sharp) 1600.00
Residual stress perpendicular to

Depth from machined surface

R50
cutting plane "RS33" (MPa)

800.00 zero (at surface)

R75
R100 25 microns
50 microns
1200.00 75 microns
Tensile RS11 (MPa)

400.00

800.00
0.00

tensile
layer 400.00
-400.00
0.00 200.00 400.00 600.00 800.00
(b) Depth from machined surface (microns)

Fig. 5. (a) Effect of tool-edge radius on in-depth RS11 profile. (b) Effect 0.00
of tool-edge radius on in-depth RS33 profile. 20.00 40.00 60.00 80.00 100.00
(b) Tool-edge radius (microns)

Fig. 6. Effect of tool-edge radius on workpiece temperature and tensile

3.2.1. R.S. tensile layer RS11 at different depths. (a) Workpiece temperature, (b) tensile RS11.
Fig. 5a shows that the magnitude of surface tensile RS11
is significantly affected by tool-edge radius, where it current results as no temperature rise was noticed deeper
increased from a minimum of +763 MPa for sharp tool than 130–140 mm from the surface. Therefore, the tensile
(R20) to a maximum of +1337 MPa for tool R100 (75% layer was found to have almost the same thickness for all
increase). The same behaviour was found when turning edge radii, although the magnitude of surface R.S. has
stainless steel AISI 304 [3]. This is likely due to the increase experienced a significant increase with edge radius, as
in cutting temperatures with edge radius as shown in shown in Fig. 5. Larger edge radius results in larger contact
Fig. 6a, because it is well known that surface tensile R.S. area between the tool tip and the workpiece material,
result from thermal gradients rather than mechanical which in turns results in two contradicting phenomena:
deformations. Stainless steels, in general, have low thermal first, an increase in the amount of heat generated due to
conductivity when compared to carbon or alloy steels. The friction which has the tendency to increase workpiece
low thermal conductivity of AISI 316L (16.2–21.4 W/m 1C) temperature; second, an increase in the amount of heat
helped entrapping the heat energy in the near-surface layer; dissipated from the workpiece into the tool which has the
consequently, the effect of edge radius on workpiece tendency to decrease workpiece temperature. The overall
temperature was only significant in the near-surface layer result of these two contradicting phenomena could not be
and it faded away quickly with depth. This was clear in the predicted a priori.
 ARTICLE IN PRESS
408 M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411

Since the properties of the near-surface layer in any 170.00

component play a crucial role in controlling its perfor-

Thickness of plastically deformed layer (microns)

mance, especially its R.S. distribution; therefore, more
attention was given to such region trying to understand the 160.00
reason behind the increase in tensile R.S with edge radius.
Fig. 6 shows the effect of tool-edge radius on workpiece
temperature and the magnitude of tensile R.S. at different
150.00
depths. Both parameters were found to vary linearly with
tool-edge radius supported by having high correlation
factors ranging from 0.96 to 0.99.
140.00
3.2.2. R.S. compressive layer
The magnitude of tensile R.S. was found to increase with
edge radius; therefore, a corresponding increase in the 130.00
magnitude of compressive R.S. is expected. This is because
R.S., by definition, have to be in equilibrium state. Figs. 5a
and b show an increase in the magnitude of compressive
120.00
RS11 and RS33, respectively, with tool-edge radius.
Furthermore, the location of their maximum value went 20.00 40.00 60.00 80.00 100.00
deeper into the workpiece with larger edge radius, as was Tool-edge radius (microns)
also reported in [2,16]. This is attributed to material plastic Fig. 8. Thickness of plastically deformed layer.
deformation, which was found to increase with tool-edge
radius. Fig. 7 shows the distribution of equivalent plastic
strain (PEEQ) through workpiece thickness just under the It is important to note that compressive R.S. went deeper
tool tip during cutting for different edge radii. It is clear into the workpiece than the plastically deformed region, i.e.
that the tool-edge radius had a significant effect on the R.S. were induced in an un-plastically deformed part. This
surface/maximum value of PEEQ bringing it up from might seem a little bit confusing because R.S. are known to
about 5.6 for tool R20 to about 8.6 for tool R100. Besides, be induced by either inhomogeneous plastic deformation,
the thickness of plastically deformed layer increased with whether being mechanical or thermal, or phase transfor-
edge radius as shown in Fig. 8. The increase in plastic mation which was not considered in the current model.
deformation with tool-edge radius is well understood and Since the plastically deformed and un-deformed regions are
explained in terms of the formed dead-metal (stagnation) next to each other, the elastically deformed region will
zone as will be explained in the next section. always try to restrain the plastic deformation of the above
region, and vice versa, till they reach a balanced state.
Therefore, R.S. could be induced in the plastically un-
10.00
deformed region as was found in the current work.
Furthermore, the process of inducing R.S. is not a direct
Tool edge:
result to either thermal gradients or mechanical deforma-
R20 (sharp)
R50
tion, but it is a combination of both of them, which makes
8.00 it a complicated process.
R75
Equivalent plastic strain "PEEQ"

R100

3.3. Dead/stagnated metal zone

6.00
Fig. 9 shows the distribution of total velocity (V) in the
workpiece around the tool tip during steady state cutting,
where the tool is not displayed. When the total velocity
4.00
tends to be almost zero, the material will not be moving
any more and will stagnate forming what is known as
‘‘dead metal’’ or ‘‘stagnated metal’’ zone. The stagnation
2.00 zone is simply formed due to the entrapment of some
workpiece material underneath the rounded tip of the tool.
This zone is shown in dark colour in Fig. 9 where it is
clearly formed for non-sharp tools, R50, R75, and R100,
0.00
and its size increased with tool-edge radius. For sharp tool
0.00 40.00 80.00 120.00 160.00
(R20), a small unstable stagnated region was noticed that
Depth from surface (microns)
changed in shape and size during cutting. In all cases, the
Fig. 7. Equivalent plastic strain distribution. stagnation zone started to form in the early beginning of
 ARTICLE IN PRESS
M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411
V (m/s)
0.00 0.08 0.17 0.25 0.33 0.42 0.50 0.58 0.67 0.75 0.83 0.92 1.00 2.20
R20
Unstable stagnated material
0.05 mm
R75
T
Stagnation zone
∆h
h75
0.05 mm
Fig. 9. Total velocity (V) distribution during steady state cutting.
cutting, transient state; then its shape and size stabilized
only when steady state cutting was reached. In general,
non-sharp tools usually develop such phenomenon which
has been reported earlier for blunt and chamfered tools in
[6,9,11], and its size was found to increase with chamfer
angle where it almost filled the missing sharp edge. It is
important to keep in mind that the dead-metal zone acts as
the first effective cutting edge during cutting [11], which
adds to the complexity of the cutting process especially that
its size and shape cannot be predicted a priori.
Since the stagnation zone acts as the first effective cutting
edge, it is believed to play an important role in controlling
and explaining the current results. Consequently, it was
important to find more about its details. The stagnated
material was almost triangular in shape, as shown in Fig. 9,
and its size (area of triangle) increased with edge radius.
Fig. 10 shows a clear linear relation between the size of
stagnation zone and tool-edge radius. The stagnation-zone
tip, circled in white in Fig. 9, showed a very interesting
behaviour. Its height from the machined surface, denoted
409
R50
Stagnation zone
h50
0.05 mm
T R100
Stagnation zone
h100
0.05 mm
by ‘‘h’’ in Fig. 9, increased with edge radius, while it was
always at the same location ahead of the tool rake face,
denoted by ‘‘T’’ in Fig. 9, which was about 5 mm. Table 3
shows the height (h) for different cases that experienced
stagnation-zone formation. The stagnation-zone tip is
where the workpiece material starts to split into two parts,
one forming the chip and the other forming the workpiece
new surface; therefore, when it moves upwards (h
increases) more material is ploughed into the new
machined surface which tends to increase the thickness of
the plastically deformed region with edge radius as was
shown in Fig. 8. Fig. 11 shows schematically the flow of
plastically deformed material on its way to form the surface
layer in the machined surface for two cases, case (a) and
case (b). The stagnation-zone tip for case ‘‘a’’ is higher than
case ‘‘b’’, which means that the used edge radius in case
‘‘a’’ was larger than in case ‘‘b’’. The flowing material is
restricted from both sides, one side by the workpiece (w.p.)
free surface and the other side by the elastic material
underneath; therefore, case ‘‘a’’ will experience more
 ARTICLE IN PRESS
410 M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411

500.00 temperature. Away from the workpiece surface higher

compressive R.S. were noticed with larger edge radius. This
means that the thermal effect was more significant than the
mechanical one in the near-surface layer, while the opposite
Size of stagnation zone (micron2)

400.00 took place away from the surface. One of the key issues in
explaining such phenomenon is the low thermal conduc-
tivity of stainless steel AISI 316L as stated earlier. There
are some analytical works available in the literature, such
300.00 as [30,31], that studied the location of stagnation point
when using blunt tools. These works did not consider the
existence of stagnated metal zone and considered the
stagnation point to exist on the tool cutting edge.
200.00
4. Conclusions.

 The proposed A.L.E. model succeeded in modelling the

100.00 process of orthogonal cutting stainless steel AISI 316L
50.00 60.00 70.00 80.00 90.00 100.00 and predicted the residual stress profile in a way that
Tool-edge radius (microns) matched what was found experimentally.
 Larger tool-edge radius induced higher tensile R.S. in
Fig. 10. Stagnation zone size vs. edge radius. the near-surface layer while the thickness of tensile layer
was unaffected. Higher tensile stresses are attributed to
Table 3 the increase in workpiece temperature with edge radius,
Height of stagnation-zone tip from machined surface
as more heat is generated when the friction contact area
Tool R50 R75 R100 between the tool tip and workpiece increases. The
thickness of tensile layer was not affected because of the
Height ‘‘h’’ mm 20 25 35
low thermal conductivity of AISI 316L that restricted
% of uncut chip thickness 20 25 35
the effects of temperature rise to the near-surface layer.
 Larger tool-edge radius induced higher compressive
R.S., far from the surface, and moved the location of
location of stagnation tip their maximum magnitude deeper into the workpiece.
a This is attributed to higher material plastic deformation
b and more material being ploughed into the new
w.p. free surface
machined surface.
 A stagnation zone is developed for non-sharp tools
elastic boundary underneath the tool edge, which acts as the first effective
cutting edge. The location of the stagnation-zone tip
Fig. 11. Flow of plastically deformed material. moved away from the machined surface as the tool-edge
radius increased. This explained the increase in the
amount of ploughed material and material plastic
mechanical plastic deformation as it goes through larger deformation (PEEQ), which consequently explained
geometrical deformation. This explains the increase in the the increase in compressive R.S. and ploughing force
surface value of PEEQ with edge radius as was shown in component with edge radius.
Fig. 7. Also the increase in workpiece temperature with
edge radius helped increasing such component, as it helps
in plastically deforming the material. Acknowledgement
Mechanical plastic deformation tends to induce higher
compressive R.S.; therefore, when more material is The authors would like to thank the AUTO21—Theme
ploughed into the workpiece surface, higher compressive C203 and Natural Sciences and Engineering Research
R.S. will be induced as was shown in Fig. 5. It is important Council (NSERC) of Canada—Discovery Grant for the
to note that the formation of either tensile or compressive financial funding for this research.
R.S. could not be predicted based on thermal gradients or
mechanical deformation independently because they both
References
happen simultaneously. Therefore, although the mechan-
ical portion of plastic deformation increased with tool-edge [1] E. Brinksmeier, J.J. Cammett, W. Konig, P. Leskovar, J. Peters, H.K.
radius, the surface R.S. were found to become more tensile, Tonshoff, Residual stresses—measurements and causes in machining
which was attributed to the increase in workpiece surface processes, Annals of the CIRP 31 (1982) 491–510.
 ARTICLE IN PRESS
M.N.A. Nasr et al. / International Journal of Machine Tools & Manufacture 47 (2007) 401–411
[2] J.D. Thiele, S.N. Melkote, Effect of cutting edge geometry and
workpiece hardness on surface residual stresses in finish hard turning
of AISI 52100 steel, Transactions of the ASME, Journal of
Manufacturing Science and Engineering 12 (4) (2000) 642–649.
[3] D.Y. Jang, T.R. Watkins, K.J. Kozaczek, C.R. Hubbard, O.B. Carin,
Surface residual stresses in machined austenitic stainless steel, Wear
194 (1996) 168–173.
[4] C.R. Liu, M.M. Barash, Variables governing patterns of mechanical
residual stress in a machined surface, Journal of Engineering for
Industry 104 (Aug.) (1982) 257–264.
[5] R.M. Arunachalam, M.A. Mannan, A.C. Spowage, Surface integrity
when machining age hardened inconel 718 with coated carbide tools,
International Journal of Machine Tools and Manufacture 44 (2004)
1481–1491.
[6] J. Manjunathaiah, W.J. Endres, A new model and analysis of
orthogonal machining with an edge-radiused tool, Transactions of
the ASME, Journal of Manufacturing Science and Engineering 122
(Aug.) (2000) 384–390.
[7] H. Ren, Y. Altintas, Mechanics of machining with chamfered tools,
Transactions of the ASME, Journal of Manufacturing Science and
Engineering 122 (Nov.) (2000) 650–659.
[8] K.W. Kim, W.Y. Lee, H.C. Sin, A finite element analysis of
machining with the tool edge considered, Journal of Materials
Processing Technology 86 (1999) 45–55.
[9] Y.-C. Yen, A. Jain, T. Altan, A finite element analysis of orthogonal
machining using different tool edge geometries, Journal of Materials
Processing Technology 146 (2004) 72–81.
[10] M.R. Movahhedy, M.S. Gadala, Y. Altintas, Simulation of the
orthogonal cutting process using an Arbitrary Lagrangian–Eulerian
finite-element method, Journal of Materials Processing Technology
103 (2000) 267–275.
[11] M.R. Movahhedy, Y. Altintas, M.S. Gadala, Numerical analysis of
metal cutting with chamfered and blunt tools, Transactions of the
ASME, Journal of Manufacturing Science and Engineering 124
(May) (2002) 178–188.
[12] P. Joyot, R. Rakotomalala, O. Pantale, M. Touratier, N. Hakem, A
numerical simulation of steady state metal cutting, Proceedings of the
I MECH E part C, Journal of Mechanical Engineering Science 212
(5) (1998) 331–341.
[13] L. Olovsson, L. Nilsson, K. Simonsson, An A.L.E. formulation for
the solution of two-dimensional metal cutting problems, Computers
and Structures 72 (1999) 497–507.
[14] J.C. Outeiro, A.M. Dias, J.L. Lebrun, V.P. Astakhov, Machining
residual stresses in AISI 316L steel and their correlation with the
cutting parameters, Machining Science and Technology 6 (2) (2002)
251–270.
[15] R. M’Saoubi, J.C. Outeiro, B. Changeux, J.L. Lebrun, A.M. Dias,
Residual stress analysis in orthogonal machining of standard and
resulfurized AISI 316L steels, Journal of Materials Processing
Technology 96 (1999) 225–233.
411
[16] J. Hua, D. Umbrello, R. Shivpuri, Investigation of cutting conditions
and cutting edge preparations for enhanced compressive subsurface
residual stress in the hard turning of bearing steel, Journal of
Materials Processing Technology 171 (2006) 180–187.
[17] C. Wiesner, Residual stresses after orthogonal machining of AISI
304: numerical calculation of the thermal component and comparison
with experiment results, Metallurgical Transactions A 23 (A) (1992)
989–996.
[18] Y.B. Guo, Y.K. Chou, The determination of ploughing force and its
influence on material properties in metal cutting, Journal of Materials
Processing and Technology 148 (2004) 368–375.
[19] Hibbit Karlson and Sorensen Inc., ABAQUS/Explicit User’s Manual
6.4-1, 2003.
[20] H. Chandrasekaran, R. M’Saoubi, H. Chazal, Modelling of material
flow stress in chip formation process from orthogonal milling and
split Hopkinson bar tests, Materials Science and Technology 9 (2005)
131–145.
[21] A.G. Mamalis, M. Horvath, A.S. Branis, D.E. Manolakos, Finite
element simulation of chip Formation in orthogonal metal cutting,
Journal of Materials Processing Technology 110 (1) (2001) 19–27.
[22] A.G. Mamalis, A.S. Branis, D.E. Manolakos, Modelling of precision
hard cutting using implicit finite element methods, Journal of
Materials Processing Technology 123 (3) (2002) 464–475.
[23] M. Baker, J. Rosler, C. Siemers, A finite element model of high speed
metal cutting with adiabatic shearing, Computers and Structures 80
(2002) 495–513.
[24] J.S. Sun, K.H. Lee, H.P. Lee, Comparison of implicit and explicit
finite element methods for dynamic problems, Journal of Materials
Processing Technology 105 (2000) 110–118.
[25] R.D. Cook, D.S. Malkus, M.E. Plesha, Concepts and Applications
of Finite Element Analysis, Wiley, 1989, third ed. (ISBN: 0-471-
84788-7).
[26] C. Shet, X. Deng, Finite element analysis of the orthogonal metal
cutting process, Journal of Materials Processing Technology 105
(2000) 95–109.
[27] C.R. Liu, Y.B. Guo, Finite element analysis of the effect of sequential
cuts and tool-chip friction on residual stresses in a machined
layer, International Journal of Mechanical Sciences 42 (2000)
1069–1086.
[28] A.J. Shih, Finite element simulation of orthogonal metal cutting,
Journal of Engineering for Industry 117 (Feb.) (1995) 84–93.
[29] J.S. Strenkowski, K.-J. Moon, Finite element prediction of chip
geometry and tool/workpiece temperature distributions in orthogonal
metal cutting, Transactions of the ASME, Journal of Engineering for
Industry 112 (Nov.) (1990) 313–318.
[30] P.K. Basuray, B.K. Misra, G.K. Lal, Transition from ploughing to
cutting during machining with blunt tools, Wear 43 (1977) 341–349.
[31] C. Rubenstein, W.S. Lau, P.K. Venuvinod, Flow of workpiece
material in the vicinity of the cutting edge, International Journal of
Machine Tool Design and Research 25 (1) (1985) 91–97.