Oblique
Oblique
Oblique
A 3D finite element model (FEM) of the oblique chip formation process was proposed in
Abaqus/Explicit™ (v6.5) using an Arbitrary Lagrangian Eulerian (ALE) formulation. The
sensitivity of the obtained results to variations of tool geometry angles, tool-chip friction, and
cutting conditions was analyzed. Experimental tests were carried out on AISI-4140 steel using
uncoated cemented carbide tools under oblique cutting conditions for validation of the FEM
results, and a good qualitative agreement between them was obtained. The analysis highlighted
the need for a proper identification of the friction on the tool-chip interface for the accurate
reproduction of the chip formation process by means of finite element modeling.
INTRODUCTION
Since the first approaches in the early 1970s (Okushima and Kakino,
1971; Tay et al., 1974), the finite element modeling (FEM) of the cutting
process has evolved toward the implementation of more realistic and
complex phenomena in the cutting models like elasto-viscoplastic material
behavior, variable friction along the rake face, or the effect of tool nose
radius in 3D simulations (Aurich and Hill, 2006; Klocke and Kratz, 2005;
Pantalé et al., 2004). Various works have been reported that incorporate
complex workpiece material microstructure into the simulations (Chuzoy
et al., 2003; Simoneau et al., 2007), being able to introduce microscale
effects into the analysis of the chip formation process.
However, the results obtained by the most commonly employed
techniques in this kind of simulations, namely, material damage-based
element deletion and continuous remeshing of the deformed material,
angle equals the cutting edge inclination angle, rejecting other possible
dependencies.
c = s (1)
Presently, there still appear analytical works that tackle the cumbersome
task of introducing viscoplastic material behavior and friction laws with
complex dependencies into three-dimensional oblique cutting models for
the modeling of such processes (Moufki et al., 2000).
FIGURE 4 Schematic view of the angles analyzed in the oblique cutting geometry.
edge angle (r ) is measured. Ps is the tool cutting edge plane, which is
tangential to the cutting edge and contains the cutting speed direction,
where the inclination angle (s ) is measured. Last, Pf is the working plane,
which is perpendicular to the cutting edge and where the rake angle (n )
is measured (Figure 5).
The adaptive meshing technique employed for the workpiece allows
the modification from the initial shape defined for the model to a
deformed shape given by the process parameters. Moreover, no criterion
is needed to achieve the material separation into chip by the elimination
of elements or debonding duplicated nodes.
Reduced integration and hourglass control eight-node hexahedric
elements with trilinear displacement and temperature calculation
(C3D8RT) having been used in the model, with varying size from 15 m
close to the cutting zone to 0.1 mm in the outer zones.
FIGURE 5 Rake (n ), cutting edge (r ), and inclination (s ) angles on each of the planes for their
measurement (CIRP, 2004).
390 I. Llanos et al.
TABLE 1 Input Parameters for the FE Model. Cells in Grey Show the Input Parameters Varied
in this Study
Material
Plasticity A coefficient (MPa) 598
Johnson–Cook Law B coefficient (MPa) 768
(AISI-4140) n coefficient 0.2092
(Grolleau, 1996) C coefficient 0.0137
m coefficient 0.807
Inelastic heat fraction () 0.9
Density () (Kg · m−3 ) Part (AISI-4140 Steel) 7800
Tool (P10) 10600
Young’s Modulus (E ) (N · m−2 ) Part (AISI-4140 Steel) 21 · 1011
Poisson’s ratio ( ) Part (AISI-4140 Steel) 0.3
Thermal conductivity Part (AISI-4140) 33.1–42.3
() (W · m−1 · K−1 ) Tool (P10) 25
Specific heat Part (AISI-4140) 473–561
(c) (J · Kg−1 · K−1 ) Tool (P10) 200
Coefficient of linear thermal expansion ( ) (K−1 ) (AISI-4140) 0 − 15 · 10−6
Convective heat transfer coefficient (hc )(W · m−2 · K−1 ) 25
Emissivity ( ) 0.35
Mass scaling
Contact
Ther. conductance (Ki )(W · m−2 · K−1 ) 108
Heat partition coefficient ( ) 0.5
Friction coefficient ( ) 0.1/0.23/0.5
Friction energy tran. into heat () 1
Emissivity ( ) 0
Process
Cutting speed (Vc ) (m · min−1 ) 300/200
Uncut chip thickness (h) (mm) 0.2/0.3
Depth of cut (doc) (mm) 1/3
Cutting edge roundness (r ) ( m) 40
Rake angle (n ) (◦ ) +6/ − 6
Clearance angle ( n )(◦ ) 5
Cutting edge angle (r ) (◦ ) 90/75
Cutting edge inclination angle (s )(◦ ) 0/ + 6
Minimum time increment (dt)(s) Not limited/10−8
Finite Element Modeling of Oblique Machining 391
FIGURE 6 Example of chip flow angle estimation in the FE model based on contact pressure
(CPRESS) distribution and the mean velocity direction for the last row of contact elements.
model (2) (Johnson and Cook, 1983) is employed for the latter. Although
there are limitations in the capability of such constitutive equation to
model the coupling between the temperature, strain, and strain rate
(Warnecke and Oh, 2002), this model is selected because of the amount
of data available in the literature. The tool-chip contact on the rake face
is modeled by a Coulomb friction law obtained from orthogonal cutting
tests employing the methodology proposed by Albrecht (Albrecht, 1960;
Arrazola et al., 2008). Values taken for the different input parameters are
given in Table 1.
˙ pl m
pl n ¯ w − amb
¯ = A + B × ¯ × 1 + C × ln × 1− (2)
˙¯ 0 m − amb
FIGURE 7 Image of the experimental setup employed for the oblique machining tests.
FIGURE 8 Measurements of chip flow angle on images taken for a sample test condition.
TEST PLAN
First, three FEM simulations (Table 2) were conducted on the model
with 6◦ rake angle, 75◦ cutting edge angle, 6◦ inclination angle, 300 m ·
min−1 cutting speed, and 0.2 mm · rev−1 undeformed chip thickness (h) to
analyze the effects on the obtained results of:
Once the effects of these input parameters on the results obtained with
the proposed model were analyzed, two experimental validation plans were
carried out on the model with 0.23 friction coefficient, 1 mm depth of cut
TABLE 3 Values Taken for the Analyzed Input Parameters for Comparison of FEM and
Experimental Results
Note: For test plan 1 (Cutting geometry), the cutting conditions were fixed as 300 m · min−1
cutting speed, and 0.2 mm · rev−1 feed. For test plan 2 (Cutting conditions), the cutting geometry
was fixed as 6◦ rake angle, 6◦ inclination angle, and 75◦ cutting edge angle.
(doc), and the application of mass scaling to verify the behavior of the FE
model in comparison to experimental results when modifying the cutting
geometry and the cutting conditions.
For the cutting geometry experimental plan, the values taken for the
rake, inclination, and cutting edge angles were modified to analyze their
effects on the three-dimensional material flow during the cutting process,
whereas the cutting speed and feed were kept constant (300 m · min−1 and
0.2 mm · rev−1 , respectively). In the case of the test plan for variation of
cutting conditions, cutting speed and feed were modified, whereas the
rake, inclination, and cutting edge angles were kept constant (6◦ , 6◦ , and
75◦ , respectively). Table 3 shows the values taken for each variable in the
machining tests for the proposed factorial design of experiment (DOE).
Because of the multiple uncertainty sources of the FEM model
(constitutive models, numerical parameters, etc.), a straight comparison
between experimental and FEM results could be misleading. Therefore,
the analysis of the results will be carried out from a qualitative point of
view, comparing the behavior of the FEM model and the experimental
setup due to an input variable modification. In this sense, the effect
and interaction graphs were employed for the graphical representation
of such behavior (Box et al., 1978). These graphs show the main effects
and interactions of the input parameters over the output variables as
percentage values (with respect to the average value).
RESULTS
Next, the results obtained in each test are shown. First, the analyses
done on the FEM model will be shown, trying to identify the
possible uncertainties introduced into the numerical results due to
the simplifications taken for the model. Subsequently, the results will be
compared with the experimental ones to validate the behavior of the
model. All the cutting force values will be given per unit depth of cut, in
Finite Element Modeling of Oblique Machining 395
TABLE 4 Values Obtained for the FEM Model with Different Friction Coefficients
FEM
Values obtained while using mass scaling, 1 mm doc, 6◦ rake, 6◦ inclination, and
75◦ cutting edge angles, 300 m · min−1 cutting speed and 0.2 mm · rev−1 feed. The
cutting force values are normalized per unit depth of cut.
396 I. Llanos et al.
TABLE 5 Values Obtained for the FEM Model with Different Depth of Cut
FEM
Values obtained while using mass scaling, 0.23 friction coefficient, 6◦ rake,
6 inclination, and 75◦ cutting edge angles; 300 m · min−1 cutting speed and
◦
0.2 mm · rev−1 feed. The cutting force values are normalized per unit depth
of cut.
values at low depths of cut because the FE models would tend to predict
higher chip flow angle values at such conditions.
Figure 9 shows the contact pressure (CPRESS) contour maps on the
secondary shear zone of the chip for (a) the 1 mm thickness simulation
and for (b) the 3 mm thickness model. Two zones could be recognized
in both models: the one corresponding to the lateral edges of the chip
(areas denoted by ellipses) with low contact pressure, and a continuous
high contact pressure zone in the center of the model. Because the
contact pressure was higher in the central zone, this one will be the main
contributor to the frictional forces in comparison to the lateral zones.
Taking into account that the lateral zones with low-contact pressure were
of similar dimension in both models, the ratio of central zone/lateral zone
was higher in the model with 3 mm depth of cut. This way, the lower chip
flow angle estimated with this model in comparison to the 1 mm thickness
model could be related to such a higher ratio.
An additional analysis was carried out around the chip cross-sections
to verify the capability of the adaptive meshing technique employed for
the reproduction of the material flow during oblique cutting for both the
FIGURE 9 Contact pressure contour maps in the tool-chip contact zone for the (a) 1 mm and
(b) 3 mm depth of cut models.
Finite Element Modeling of Oblique Machining 397
FIGURE 10 Chip cross-sections for experimental and numerical tests under (a) orthogonal and
(b) oblique conditions.
1 mm and the 3 mm depth of cut models. Figure 10 shows the chip cross-
sections for experimental and numerical tests in both (a) orthogonal and
(b) oblique conditions. The images above on the figure correspond to the
experimental chips, the images in the middle to the 3 mm depth of cut
FEM model, and the ones below to the 1 mm depth of cut model.
The figures show how the chip sections obtained on orthogonal
conditions are symmetrical, whereas the ones for oblique cutting are
not, modifying the shape of the laterals. It can be seen that both FEM
models are able to reproduce this effect, showing the capability of the ALE
formulation to reproduce the modification of the material flow due to the
cutting geometry.
FEM
The cutting force values are normalized per unit depth of cut.
398 I. Llanos et al.
TABLE 7 Numerical and Experimental Results for the Cutting Geometry Test Plan
FEM Experimental
Test plan was conducted on the FEM model with mass scaling, 1 mm doc, 0.23 friction
coefficient, 300 m · min−1 cutting speed, and 0.2 mm · rev−1 feed. The cutting force values are
normalized per unit depth of cut.
Finite Element Modeling of Oblique Machining 399
FIGURE 11 Effect and interaction graphs for (a) cutting force, and (b) feed force for the
experimental plan involving cutting geometry.
FIGURE 12 Effect and interaction graphs for (a) tangential force, and (b) chip flow angle for the
experimental plan involving cutting geometry.
400 I. Llanos et al.
TABLE 8 FEM and Experimental Results for the Cutting Conditions Test Plan Conducted on
the FEM Model with Mass Scaling, 1 mm doc, 0.23 Friction Coefficient; 6◦ Rake, 6◦ Inclination,
and 75◦ Cutting Edge Angles
FEM Experimental
Vc200_f0.2 1178 4252 1343 −67 80 4188 1994 −229 3.6
Vc200_f0.3 1278 6189 1621 −36 120 5763 2239 −163 4.5
Vc300_f0.2 1308 425 141 −9 79 389 171 −18 1.7
Vc300_f0.3 1415 6165 1653 −36 113 5647 2021 −58 4.9
Average value 1295 5214 1507 −57 98 4872 1991 −158 3.7
The cutting force values are normalized per unit depth of cut.
FIGURE 13 Effect and interaction graphs for (a) cutting force and (b) feed force for the
experimental plan involving cutting conditions.
Finite Element Modeling of Oblique Machining 401
FIGURE 14 Effect and interaction graphs for (a) tangential force, and (b) chip flow angle for the
experimental plan involving cutting conditions.
DISCUSSION
When comparing the FEM estimates for the force components and
chip flow angles with the measured experimental values, higher chip flow
angles and cutting forces are obtained numerically, whereas the values for
the feed and tangential forces stay below the experimental ones.
Considering the simplifications implemented for the decrease of the
calculation time for the numerical model, higher cutting forces and chip
flow angles have been reported when employing the 1 mm depth of cut
model and mass scaling. Thus, the higher values obtained for the cutting
force and the chip flow angle on the FE model in comparison to the
experimental results could partially be justified.
Moreover, it must be taken into account that in the experimental
tests, the thickness of the tube shows a variable cutting speed profile
(Figure 15a); whereas in the FE model, it is uniform (Figure 15b), which
would lead to even lower experimental values for the chip flow angle in
comparison to the numerical ones. In contrast, taking into account that
the tangential and feed forces would oppose the sliding of the chip over
402 I. Llanos et al.
FIGURE 15 Cutting speed profiles for (a) experimental tests and (b) numerical tests.
the tool rake face, it seems logical that lower values of such forces would
lead to higher chip flow angles.
Regarding the analysis carried out on the modification of the cutting
geometry, the developed model has shown a good behavior in comparison
to the experimental results. Although showing differences in the case of
the effect of the modification of the rake angle, such effect has been
already shown by Moufki et al. in their work on the analytical modeling of
the oblique cutting process (Moufki et al., 2000). In fact, they have shown
that when employing a simple Coulomb friction law, as the rake angle
increases, the chip flow angle would increase also (Figure 16a). In contrast,
when employing a more sophisticated temperature-dependant friction law,
the chip flow angle would change its trend; increasing when the rake angle
would turn into a more negative value (Figure 16b).
In the case of the experimental plan conducted to study the effect
of the cutting conditions on the cutting forces and chip flow angle, the
behavior of the presented model agrees qualitatively with the experimental
results except for the cutting speed. This point could once again be
related to the simple friction law employed in the FEM model, unable
to account for a “friction softening” due to a temperature rise. Although
FIGURE 16 Effect of the rake and inclination angles on the chip flow angle for different friction
characteristics: (a) = constant and (b) = f (). (Moufki et al., 2000).
Finite Element Modeling of Oblique Machining 403
NOMENCLATURE
REFERENCES
ABAQUS V6.6 (2006) Analysis User’s Manual, Vol. II: Analysis.
Albrecht, P. (1960) New development in the theory of metal-cutting process. Part I: The ploughing
process in metal cutting. Journal of Engineering for Industry, pp. 348–358.
Armarego, E.J.A.; Brown, R.H. (1969) The Machining of Metals, Prentice-Hall Inc, Englewood Cliffs,
New Jersey.
Arrazola, P.J.; Meslin, F.; Marya, S. (2004) Sensitivity study in numerical cutting modeling. Proceedings
of the 7th CIRP International Workshop on Modelling of Machining Operations, Cluny, France, pp.
83–90.
Arrazola, P.J.; Pujana, J.; Llanos, I.; Villar, A.; Ugarte, D.; Aguirre, A.; Gallego, I.; Maître, F. (2006)
Finite element modeling of oblique cutting. Proceedings of the 9th CIRP International Workshop
on Modelling of Machining Operations, Bled, Slovenia, pp. 101–112.
Arrazola, P.J.; Ugarte, D.; Domínguez, X. (2008) A new approach for the friction identification
during machining through the use of finite element modelling. International Journal of Machine
Tools and Manufacture, 48: 173–183.
Aurich, J.C.; Hill, B. (2006) 3D Finite element modeling of segmented chip formation. Annals of
the CIRP, 55(1): 47–50.
Box, G.E.P.; Hunter, W.G.; Hunter J.S. (1978) Statistics for Experiments, John Wiley & Sons, Hoboken,
New Jersey.
Chuzoy, L.; DeVor, R.E.; Kapoor, S.G. (2003) Machining simulation of ductile iron and its
constituents. Part 2: Numerical simulation and experimental validation of machining. Journal
of Manufacturing Science and Engineering, 125: 192–201.
CIRP International Institution for Production Engineering Research (2004) Dictionary of Production
Engineering, Vol. 2, Material Removal Processes. 1st edition, Springer, Berlin, Germany.
Grolleau, V. (1996) Approche de la validation expérimentale des simulations numériques de la
coupe avec prise en compte des phénomènes locaux à l’arête de l’outil. PhD thesis, École
Centrale de Nantes.
Johnson, W.K.; Cook, R. (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,
Hague, The Netherlands, pp. 541–547.
Klocke, F.; Kratz, H. (2005) Advanced tool edge geometry for high precision hard turning. Annals
of the CIRP, 54(1): 45–50.
406 I. Llanos et al.