Sie sind auf Seite 1von 8

ARTICLE IN PRESS

International Journal of Mechanical Sciences 50 (2008) 124131


www.elsevier.com/locate/ijmecsci

The Merchants model of orthogonal cutting revisited:


A new insight into the modeling of chip formation
A. Molinari, A. Moufki
Laboratoire de Physique et Mecanique des Materiaux, Universite Paul Verlaine, Ile du Saulcy, Metz, France
Received 9 April 2007; received in revised form 17 July 2007; accepted 18 July 2007
Available online 7 August 2007

Abstract

The Merchants model, as the most famous approach of orthogonal cutting, is widely used in introductive courses on machining.
However, the shear angle predicted by the Merchants model from the criterion of minimization of the cutting energy, does not generally
agree with experimental data and numerical simulations. The aim of this paper is to elucidate the theoretical reason for which the
Merchants model fails to predict the correct orientation of the primary shear zone. It is shown that the principle of minimum of the
cutting energy must be supplemented by a stability criterion of the chip morphology. A modied Merchants formula is then obtained for
the value of the shear angle.
r 2007 Elsevier Ltd. All rights reserved.

Keywords: Orthogonal cutting; Merchants model; Shear angle; Energy minimum; Stability criterion; Free surface perturbation

1. Introduction hardening, no thermal and no rate dependence). Friction at


the toolchip interface is governed by a Coulomb law, with
Most of orthogonal cutting models assume the existence friction coefcient m. The Merchants shear angle is
of a primary shear zone which has the aspect of a thin band obtained by minimization of the cutting energy:
where the chip is formed by intense shearing of the work
material, see for instance, Refs. [13]. These models are p al
fM , (1)
extensions of the seminal work of Merchant [4,5] who 4 2
assumed that the primary shear zone was innitely thin.
Representing the primary shear zone as a narrow band where a is the rake angle and l is the friction angle dened
seems to be a realistic schematization in many cases, such by m tg(l).
as high-speed machining and cutting operations of However, some discrepancy is generally found between
materials with negligible strain hardening [6]. the value of the shear angle measured in experiments and
For the working of cutting models, the shear angle f, the theoretical prediction (1), see for example, Fig. 8.12 in
characterizing the inclination of the primary shear band Kalpakjian [8]. Indeed, experimental data are generally
with respect to the cutting velocity, must be accurately better described by the empirical Zvorykins law [9]:
dened. This requirement is essential to obtain realistic fZ A1 A2 a  l, (2)
predictions of cutting variables such as cutting forces and
temperature distribution within the cutting zone [3,7]. In where the constants A1 and A2 are determined from
the Merchants model, plastic shearing is localized along orthogonal cutting experiments. Using Zvorykins law (2)
the shear line AB (Fig. 1). Across this line the tangential in the cutting model [3], provided a good agreement with
component of the particle velocity is discontinuous. The experimental results for cutting forces and temperature
material is assumed to be perfectly plastic (no strain distributions.
To explain the discrepancy between the value of the
Corresponding author. Tel.: +33 387315369; fax: +33 387315366. shear angle predicted by Eq. (1) and experimental data, the
E-mail address: molinari@lpmm.univ-metz.fr (A. Molinari). hypothesis of perfect plasticity made in the Merchants

0020-7403/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijmecsci.2007.07.015
ARTICLE IN PRESS
A. Molinari, A. Moufki / International Journal of Mechanical Sciences 50 (2008) 124131 125

in terms of the rake angle, the friction angle and the free
Vc surface perturbation shape.
t2
2. Effect of the free surface geometry on the shear angle
c

The pioneered work of Merchant was an important step
in the modelling of the orthogonal cutting process. The
Chip
B Tool workpiece material is taken as perfectly plastic and the chip
is assumed to slide along the tool rake face. The contact is
V t1  governed by a Coulomb friction law with constant friction
angle l. The shear angle f that minimizes the cutting
A energy is given in Eq. (1). In this model the chip is formed
Fig. 1. Schematic view of the orthogonal cutting process.
by shearing along the line AB with inclination f fM with
respect to the cutting direction (Fig. 1). Note that the chip
thickness predicted by the Merchant model is
modeling could be incriminated. Indeed, the workpiece
material behaviour is in general not perfectly plastic. t2M t1 cosfM  a= sinfM , (3)
However, strain hardening, strain rate sensitivity and
where t1 is the uncut chip thickness (or depth of cut)
thermal softening of the work material can be accounted
(Fig. 1).
for in the thermomechanical model of Moufki et al. [3].
The Merchants model is related to the limit analysis
Then, if the hypothesis of minimization of the cutting
theory [11]. For a material obeying to the principle of
energy is used, the value of the shear angle is found to be
maximum plastic work, the component F of the cutting
close to Merchants result (1). Thus, the nature of the
force in the cutting direction is given by
plastic response cannot be viewed as the main cause of
disagreement between experiments and modelling. More- FV inf PV^ . (4)
over, nite element simulations of orthogonal cutting V^

predict that the shear angle may be quite different from In this expression PV^ is the energy dissipated in the
the value given by cutting models based on cutting energy system per unit time for the admissible virtual velocity eld
minimization (for the same constitutive law of the work V^ . To be admissible, a virtual velocity eld must be volume
material). This was clearly demonstrated by Baker [6] who preserving and must satisfy the kinematic boundary
considered a perfectly plastic material as in the Merchants conditions. PV^ can be viewed as the virtual cutting
approach. This was also shown by Miguelez et al. [7] by energy per unit time (cutting power). The right-hand side of
considering a viscoplastic thermal-sensitive material. Baker Eq. (4) is the minimum of PV^ for all admissible virtual
[6] concluded that the minimization of the cutting energy is velocity elds V^ . In the Merchants model, the virtual
not an adequate criterion to estimate the shear angle. Note velocity elds considered are dened by the rigid body
that this conclusion was also made by Hill [10] who motion of two blocks with a velocity jump along AB
attributed the failure of the theory when compared to (Fig. 1). These elds are solely depending on the shear
experiments to almost certainly the inadequacy of the angle f. They are denoted by V~ f and constitute a
minimum work hypothesis. subclass of the set of admissible velocity elds. Thus, from
To investigate the theoretical reasons for which the Eq. (4) we have
Merchants approach does not predict the correct value of
the shear angle, a new analysis of the chip formation is FV p inf PV~ f. (5)
f
developed here. The context is the same as in the
Merchants paper. Orthogonal cutting of a perfectly plastic In Eq. (5) the minimum of the virtual cutting energy is
material is considered, and the cutting energy is minimized realized, according to Merchant, for the value fM of the
for virtual velocity elds dened by the rigid body motion shear angle. This minimum denes an upper bound of the
of two blocks. It is shown that minimizing the cutting cutting force:
energy is not a sufcient condition to get an appropriate 1 1
solution of the problem of chip formation. Indeed, the F upp inf PV~ f PV~ fM . (6)
V f V
stability of the chip morphology to a geometrical
perturbation is an additional condition to be considered. This expression can be written as
The stability of the chip morphology is analysed by  
p la
introducing little perturbations of the free surface. For a F upp 2t1 wk tan , (7)
4 2
given perturbation shape, one can characterize the optimal
value of the shear angle for which the cutting energy is where w is the width of cut and k is the yield stress in shear.
minimized and the stability of the chip thickness is ensured. In the Merchants model, the upper bound (7) is considered
An explicit expression for the optimal shear angle is found as being an estimate of the cutting force.
ARTICLE IN PRESS
126 A. Molinari, A. Moufki / International Journal of Mechanical Sciences 50 (2008) 124131

t2 = BH
Vc

C D
t2*=BH
Vc
B B'
H

Tool B H
t1 = AH V
M
t1=AH V  Tool
*
A
workpiece
A
F workpiece

Fig. 2. Virtual admissible velocity eld at the beginning of the cutting


process. The optimal shear angle that minimizes the cutting energy, is Fig. 3. Workpiece with an excrescence BCDH of width t2 . The virtual
given by fM (Merchants model). velocity eld is dened by the rigid body motion of two blocks, the rst
with cutting velocity V, the second with sliding velocity Vc along the tool.
The particle velocity undergoes a tangential discontinuity across the
The theoretical framework upon which the Merchants separation line AB0 . The optimal value of the shear angle f (inclination of
AB0 ) minimizing the cutting energy is fopt fM (Merchant model) for
model is based is questionable because no attention was f*XfM, and fopt f* for f*pfM (see Fig. 4).
paid to the sensitivity of results with respect to the free
surface geometry. Indeed, it will be shown that the value of
the shear angle obtained by minimization of the cutting expression for fpf*:
energy can be dramatically affected by a little perturbation cosl
of the free surface. F f kt1 w fpf . (8)
sinfcosf l
In the following, the same assumptions as for the
Merchants model are adopted. The same virtual velocity For fXf*, it is sufcient to replace t1/sin(f) in Eq. (8)
elds are considered (rigid body motion of two blocks). (the length of AB0 for fpf*) by t2 =cosf (the length of
The criterion of minimization of the cutting energy is still AB0 for fXf*) to obtain
used to characterize the shear angle. The analysis of the
cosl
effect of a little perturbation of the free surface geometry is F f kt2 w fXf . (9)
the additional and crucial input of this paper. cosfcosf l
Results obtained from limit analysis theorems, may be By using the relationship
affected by changes of the free surface geometry. As a rst
illustration we consider two cases where the free surface is t1 t2 tanf , (10)
identied to
we have
cosl
(i) the initial at surface of the workpiece (Fig. 2), F f kt1 w  fXf . (11)
tanf cosf cosf l
(ii) the at surface of the workpiece with an additional
excrescence (Fig. 3). For a given cutting velocity V, the shear angle has an
optimal value, fopt, for which the cutting power F(f)V
For the purpose of discussion, it is enough to take a zero (and the force F(f)) is minimum. The dependence of the
rake angle, a 0. normalized force
Case (i) is implicitly considered in the Merchants model F
F~ (12)
and represents the conguration existing at the beginning kwt1 cosl
of machining (Fig. 2). The workpiece surface is at, no chip
has been yet formed. The optimal shear angle for which the with respect to the shear angle f is illustrated in Fig. 4.
cutting power is minimum is given by fM (Eq. (1)). Note that in Fig. 4, the slope dF/df has a discontinuity at
The case (ii) is shown in Fig. 3. The workpiece surface f f*, where the expressions (8) and (11) of F(f)
has an excrescence BCDH of width t2 . Admissible virtual coincide.
velocity elds are dened as in the Merchants approach by Two cases can be distinguished:
the rigid body motion of two blocks separated by the line
AB0 across which the velocity is discontinuous. Denoting (a) for f*XfM, the optimum angle has the value
by f the inclination angle of AB0 , the force F work- fopt fM predicted by the Merchants model (1), see
conjugate to the cutting velocity V has the following Fig. 4a. Note that the result is affected by friction;
ARTICLE IN PRESS
A. Molinari, A. Moufki / International Journal of Mechanical Sciences 50 (2008) 124131 127

model (3) (here a 0). However, it might be possible to


5 have a chip thickness larger than t2M. This discussion shows
4.8 that the outcome of the limit analysis is quite sensitive to the
NORMALIZED CUTTING FORCE

geometry of the free surface considered. Therefore, it is


4.6 worth analysing how the predictions of the chip thickness
4.4 (or equivalently of the shear angle since both are related by a
one-to-one relationship), can be affected by small perturba-
4.2
tions of the free surface.
4

3.8
3. Stability of the chip morphology

3.6 A free surface geometry with a sharp rectangular angle


3.4 at the chip root (point B of Fig. 3) is neither seen in the
experiments nor in nite element simulations. A smooth

3.2 M transition is rather observed. In Fig. 5 it is assumed that the
3 free surface morphology HIJK presents a progressive
20 22 24 26 28 30 32 34 36 38 40
transition IJ from the workpiece surface HI to the chip
SHEAR ANGLE (deg) boundary JK. This conguration can be viewed as a
perturbation of the surface HBK by an additional
5 triangular material element IBJ characterized by the
inclination angle (or shape angle) y and the (normalized)
4.8
perturbation amplitude dened by
NORMALIZED CUTTING FORCE

4.6
l1
d1 , (13)
4.4 t1
4.2 where l1 IB.
In the following discussion, we take a 0, the effect of a
4
non-zero rake angle being analysed later. Consider a point
3.8 P on the free surface HIJK. The virtual velocity elds are
3.6 dened by the rigid body motion of two blocks separated
by the line AP across which the tangential velocity
3.4 is discontinuous. The work-conjugate cutting force is
3.2 given by
 M
3 cosl
20 22 24 26 28 30 32 34 36 38 40 F f kw AP; (14)
cosf l
SHEAR ANGLE (deg)
where f is the inclination of the shear line AP (Fig. 5).
Fig. 4. Evolution (bold solid line) of the normalized virtual cutting force
in terms of the shear angle f (inclination of the shear line AB0 , Fig. 3). The
geometry of the free surface, the angle f* and the shear angle f are K
dened in Fig. 3. Here the friction angle has the value l 301, and the t2
Merchant shear angle is fM 301 for a 0. It appears that the minimum
of the cutting force (i.e. the minimum of the cutting energy) is realized for 
the value of the shear angle f given by (a) the Merchant model
J
(fopt fM) for f*XfM, (b) fopt f* for f*pfM. Note the disconti-
nuity of the slope dF/df at f*, where the expressions (8) and (11) of F(f) P
are merging. The dashed line in (a) and (b) shows the evolution of the
cutting force with f for a at free surface as in Fig. 2. Thus, by H I 
B
comparison with the bold line, the effect of the excrescence BCDH in l1
Fig. 3 is clearly illustrated. It produces the large increase of the slope Tool
dF/df for fXf*. t1

*
(b) for f*pfM, we have fopt f* , see Fig. 4b. From
A
Eq. (10), the optimum angle fopt arctant1 =t2 workpiece
appears to be solely governed by the conguration
Fig. 5. Free surface with a smooth transition IJ from the workpiece
geometry.
surface HI to the chip surface JK. The triangular material element IBJ is
characterized by the shape angle y and the length l1 IB (or by the
Thus, it appears that the chip thickness cannot be smaller relative amplitude d1 l1/t1). Without smooth transition, the free surface
than the value t2M t1/tan(fM) predicted by the Merchant would be HBK. For the rake angle a 0, IBJ is a rectangular triangle.
ARTICLE IN PRESS
128 A. Molinari, A. Moufki / International Journal of Mechanical Sciences 50 (2008) 124131

For a 0 the normalized force F~ , dened in Eq. (12),


can be written as 4

1 3.8
F~ f for fpfI , (15a) workpiece surface
sinfcosf l

NORMALIZED CUTTING FORCE


3.6
J chip surface
1 3.4
F~ f  for fXfJ , (15b) I
tanf cosfcosf l 3.2
B
1 tanyd1 cotf 3
F~ f
tanycosf sinfcosf l 2.8
for fI pfpfJ , 15c 2.6
where f*, fI and fJ are, respectively, the inclination angles 2.4
of AB, AI and AJ (Fig. 5). The chip thickness t2 and the
2.2 *
depth of cutting t1 are related by t1 t2 tan(f*). Note that
for y 0 expression (15c) reduces to expression (15a) as it 2
20 22 24 26 28 30 32 34 36 38 40
should be. For l1 0, we have tan(y) N, and Eq. (15c) is
SHEAR ANGLE (deg)
identical to Eq. (15b).
The evolution of F~ with respect to f fP is represented
by the bold line in Fig. 6 for different values of f*. The 7
friction angle is l 201, thus the shear angle predicted by 6.8
Merchants theory is fM 351. The geometry of the
NORMALIZED CUTTING FORCE

junction IJ is dened by y 451 and d1 0.1. The thin 6.6


line in Fig. 6 corresponds to the evolution of F~ for the free 6.4
surface as HBK in Fig. 5.
6.2
In Section 2 it was shown that the chip thickness could
not be smaller than t2M t1 cot(fM). Therefore the chip 6
thickness considered here is t2 t1 cot(f*) with f*pfM. I J
5.8
Fig. 6a shows that, for f* 301, the minimum of F~ is
5.6
obtained at f fI 28.61. Thus shearing occurs along the B
line AI, and the chip thickness is increased from the value 5.4
t1 cot(f*) to t1 cot(fI) providing that fIof*. The chip
5.2 *
morphology appears to be unstable with respect to the free
surface perturbation introduced here. Note that no 5
10 10.5 11 11.5 12 12.5 13 13.5 14 14.5 15
assumption of any kind was made about the origin of the
SHEAR ANGLE (deg)
free surface perturbation.
The following results are now demonstrated. For a given
9.6
value of the perturbation shape angle y, there exists an
optimal chip thickness topt 
2 y function of fopt but
NORMALIZED CUTTING FORCE

insensitive to the perturbation amplitude d1, such that 9.5

(a) the chip thickness t2 is unstable (i.e. has a tendency to 9.4


increase) for t2M pt2 otopt
2 y, with t2M t1 cot(fM ) I
J
9.3
Fig. 6. Evolution of the normalized cutting force F~ in terms of the shear
angle f (bold line) for the free surface geometry HIJK (see Fig. 5). Here B
9.2
a 0, and the geometrical parameter of the junction are y 451 and
d1 0.1. The friction angle is l 201. For a free surface dened by HBK,
the evolution of F~ would follow the thin line IBJ in place of the bold
9.1
segment IJ. Note that f* denes the inclination of AB in Fig. 5. For a *
given junction geometry, the results show the existence of an optimal shear
angle fopt for which the chip morphology is stable (here fopt 12 (b)). 9
For f 4fopt the minimum of the cutting force is achieved at fI (see (a)). 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10
Thus, for f 4fopt shearing occurs along the line AI, with the result of SHEAR ANGLE (deg)
increasing the chip thickness. The stable conguration is attained at the
shear angle fopt 12 for which the cutting force shows a plateau IJ (b).
For f ofopt , as in (c), the minimum of the cutting force is achieved at fJ.
Shearing occurs preferentially along AJ and the chip thickness is not
changed (stability).
ARTICLE IN PRESS
A. Molinari, A. Moufki / International Journal of Mechanical Sciences 50 (2008) 124131 129

being the chip thickness predicted by the Merchants


model, 60 60
(b) the chip thickness topt
2 y is stable (i.e. remains
M = 45 M = 45
unaffected by perturbations with shape angle y).
50 =0 50

OPTIMAL SHEAR ANGLE (deg)

OPTIMAL SHEAR ANGLE (deg)


slope = -0.5
Case (a) is illustrated in Fig. 6a. Fig. 6b, shows the
existence of fopt ( 121) for which F~ f has a plateau in 40 40
the range fIpfpfJ. F~ is minimum on this plateau.
Denoting the chip thickness associated to fopt by 30 30
topt  opt
2 t1 cotfopt , it appears that t2 y is stable with =45
respect to perturbations with shape angle y. Indeed any
chip thickness t2 t1 cot(f*) larger than topt 
2 t1 cotfopt
20 20
is stable. This is illustrated in Fig. 6c with
f* 71(f ofopt ). In that case we haveF~ fI 4F~ fJ . 10 =0 =0 10 1=0.01
Thus shearing occurs preferentially along AJ. This
deformation mode leaves the chip thickness unchanged.
The dependence of fopt with respect to y and the invariance 0
10-4 10-3 10-2 10-1
0
0 10 20 30 40 50
of fopt with respect to d1 will be analysed later. The results DEFECT AMPLITUDE 1 (deg) SHAPE ANGEL (deg)
obtained for fopt can be transposed to the optimal chip
thickness through the relationship topt 
2 t1 cotfopt .
 Fig. 7. (a) Effect of the perturbation relative amplitude d1 l1/t1 on the
It is worth noticing that fopt y is characterized by two
optimal shear angle fopt , for perturbations with shape angle y 451 and
conditions: (i) the stability of the chip thickness with
y 0 Note that the amplitude has no inuence on fopt , except for
respect to perturbations with shape angle y, (ii) the unrealistic large values of d1 (40.2). (b) Effect of the shape angle y on fopt
minimization of the cutting energy among all stable for d1p0.2. Note that fopt varies as y/2. Results in (a) and (b) are
congurations: F~ fopt oF~ f 8f ofopt . The later result obtained for the friction angle l 0 and the rake angle a 0.
comes from the fact that F~ f is a decreasing function of
f*. Indeed the point B associated to f* in Fig. 6 always
belongs to the branch denoted as workpiece surface in 45
Fig. 6a. The left parts of Fig. 6b and c belong to the same 40
branch as the workpiece surface branch in Fig. 6a. Since Merchant model
OPTIMAL SHEAR ANGLE (deg)

this branch is decreasing, so is F~ f . 35


=0
The perturbation amplitude d1 has no effect on fopt and 30
consequently on the value of the optimal chip thickness
topt 
2 . This is illustrated in Fig. 7a where fopt is shown to be
25
independent of d1 l1/t1 except for unrealistic large values
20
of d1 (above 0.1). Perturbations with shape angle y 451  = 30
and y 0 are considered. Note that for y 0, fopt is equal 15
 = 45
to the shear angle fM of the Merchants model ( 451 for
10
a 0 and l 0). Fig. 7b shows that the dependence of fopt present model
with respect to the perturbation shape angle y is quite well 5
represented by the law:
0
fopt  fM  y=2. (16) 0 5 10 15 20 25 30 35 40
FRICTION ANGLE (deg)
This result is true for small enough perturbation
amplitudes. Henceforth, the value of the perturbation Fig. 8. Dependence of the optimal shear angle upon the friction angle, for
different values of the perturbation shape angle and for a 0. Note that
amplitude d1 will be taken small enough (say o0.05)
the trends are the same as for the Merchant model (retrieved for y 0)
so that the results become insensitive to d1 according to with a vertical shift of y/2.
Fig. 7a.
So far the results have been obtained for l 0 and
a 0. The effect of these parameters is investigated in the
following. Fig. 8 shows the dependence of the optimal When y40, the trends are the same as for the
shear angle with respect to the friction angle, for three Merchants model except for a vertical shift by y/2 which
shape angles (y 0, 301, 451). The Merchants model is clearly apparent in Fig. 8. Therefore, from Eq. (17) we
(with a 0) is retrieved for y 0: obtain

p l p l y
fopt   . (17) fopt    . (18)
4 2 4 2 2
ARTICLE IN PRESS
130 A. Molinari, A. Moufki / International Journal of Mechanical Sciences 50 (2008) 124131

The effect of a60 is analysed now. By dening the


normalized cutting force as 30 30
F
F~ (19)
kwt1 cosl  a 25 25
the following relationships are obtained:  = 30 MM

OPTIMAL SHEAR ANGLE (deg)


1 20 20
F~ f f for fpfI , (20a)
sinf cosf l  a

cosf  a 15 15
F~ gf 
sinf cosf  a cosf l  a
 = 45
for fXfJ . 20b 10 10
These expressions coincide with Eqs. (15a) and (15b)
when a 0. The inclination angles fI and fJ of the lines,  = 20
5 5
respectively, AI and AJ (Fig. 5) are dened by =0  = 45

1
tanfI ,  0 0
cotf d1 0 10 20 30 40 -20 -10 0 10 20
 
cosa siny FRICTION ANGLE (deg) RAKE ANGLE (deg)
tanfJ 1 d1
cosa y
  Fig. 10. Optimal shear angle fopt in terms of (a) the friction angle (b) the
sina siny 1 rake angle. Solid lines correspond to values of fopt obtained by the
 cotf  d1 . 21
cosa y numerical solution of Eq. (22), dashed lines correspond to the value fMM
given by the law (23).
The optimal shear angle fopt associated to a given
perturbation of the free surface with shape angle y and
amplitude d1 is dened as before by
f fI gfJ (22) representing the shear angle fM given by the Merchants
model (Eq. (1)). Thus, the dependence of fopt with respect
with fI and fJ given implicitly in terms of fopt by Eq. (21). to the rake angle is, as in Merchants model, given by the
Fig. 9 illustrates the variation of the optimal shear angle additive term a/2.
fopt in terms of the rake angle a for l 201. The solid line, Finally, by combining the effects of y, l and a described,
corresponding to a free surface perturbation with shape respectively, in Figs. 7b, 8 and 9, the following law is found
angle y 451, is nearly parallel to the dashed line for the optimal shear angle fopt :
p ayl y
fMM fM  . (23)
45 4 2 2
This relationship denes the value of the Modied-
40
Merchant's model Merchant shear angle fMM associated to a free surface
35
perturbation of shape angle y. The values of fMM given in
OPTIMAL SHEAR ANGLE (deg)

Eq. (23) are shown in Fig. 10 to be in excellent agreement


30 with the predictions of fopt obtained directly by solving
Eq. (22).
25

20 4. Discussion and conclusions

15  = 45 For a given free surface perturbation with shape angle y,


we have obtained an optimal value of the shear angle for
10 which the chip thickness is stable and the cutting energy is
present model
minimum. Denoting this optimal shear angle by fMM
5
(Modied-Merchant shear angle), we have found that fMM
0 differs from the Merchants shear angle fM solely by the
-20 -15 -10 -5 0 5 10 15 20 additive term y/2. It is important to emphasize, that fM
RAKE ANGLE (deg) was obtained in the Merchants model from the sole
Fig. 9. Variation of the optimal shear angle fopt in terms of the rake angle criterion of minimum of the cutting energy while this
a for l 201. The solid line corresponds to the value y 451 of the criterion is supplemented in the present paper by a chip
perturbation shape angle. stability criterion.
ARTICLE IN PRESS
A. Molinari, A. Moufki / International Journal of Mechanical Sciences 50 (2008) 124131 131

The explicit expression obtained for fMM has the same chip is considered to be sliding along the tool rake face in
form as the empirical Zvorykin law (2) providing that the Merchants model as in the present model). Finally, to
A1 (p/4)(y/2) and A2 (1/2). improve the quality of results, one should enter into the
It is worth mentioning that the results obtained for model the real thermo-viscoplastic response of the work-
y 451 are almost identical to those obtained by Miguelez piece material and consider that the primary shear zone has
et al. [7] with nite element simulations of the chip a nite thickness. Nevertheless, even before fully reaching
formation. In these calculations the workpiece material this objective, a qualitative comparison with experiments
was a medium carbon steel whose behaviour was described can be drawn. In general the shape angle y can be evaluated
by a thermo-sensitive viscoplastic law. It can be argued that from experiments in the range of 201501. Therefore the
any comparison with the present results could be biased by shear angle predicted by our model is found to be below
the fact that the constitutive law used in the FE simulations those of the Merchants model by the quantity y/2, i.e. by
is not perfect plasticity. However, the analysis made in this about 101251. For many materials this observation is in
paper for a perfectly plastic material can be conducted in a agreement with experimental measurements (see Fig. 8.12
similar way in the framework of the thermomechanical in Ref. [8]).
modelling of Moufki et al. [3] by using the same The present theoretical investigation does not provide a
constitutive law as in Ref. [7]. Then, if one just considers practical mean to determine an appropriate value of the
the criterion of minimization of the cutting energy, the junction shape angle y. However, this work has an
value found for the shear angle is close to the Merchants important outcome by illustrating where the Merchants
result fM. However, when a free surface perturbation with model appears to be weak in its theoretical foundation and
shape angle y is introduced, then by using the chip stability how it could be improved. As a pedagogical model, the
criterion jointly with the minimum energy criterion, the Merchants approach has a great importance. It is used in
shear angle is found to be close to fMM (results not all textbooks to provide a simple modelling of the
presented here). Therefore the main conclusions of this orthogonal cutting process. In view of the present analysis,
paper are conserved when general plastic responses are one should have some words of cautious in the presenta-
considered including strain hardening, strain rate sensitiv- tion of the Merchants model. The proposed improvement
ity and thermal softening. of the Merchants model does not discredit the criterion of
The present theoretical investigations show the impor- minimum of the cutting energy. Simply one has to
tance of taking account of the free surface geometry at the supplement this criterion by a stability condition for the
chip-root (junction between the workpiece surface and the chip morphology.
chip) for a realistic calculation of the shear angle. The
junction, dened here by a triangular shape with inclina- References
tion angle y, may have in reality a smooth round shape as
shown by the nite element calculations, see for example, [1] Oxley PLB. Mechanics of machining. Chichester, UK: Ellis Hor-
Refs. [6,7]. The triangular shape may be thought of as a wood; 1989.
[2] Molinari A, Dudzinski D. Stationary shear bands in high speed
rst approximation of the real junction shape. An
machining. Comptes Rendus de l Academie Des Sciences Serie II
important result was that the shear angle was not 1992;315:399405.
inuenced by the amplitude of the triangular junction, [3] Moufki A, Molinari A, Dudzinski D. Modelling of orthogonal
but was only depending upon the shape angle y. cutting with a temperature dependent friction law. Journal of the
The shape of the workpiecechip junction is certainly Mechanics and Physics of Solids 1998;46(10):210338.
function of the work-material properties. In other words, [4] Merchant ME. Mechanics of the metal cutting process, I: orthogonal
cutting. Journal of Applied Physics 1945;16:26775.
the shape angle y should be material dependent. That is [5] Merchant ME. Mechanics of the metal cutting process, II: plasticity
probably the reason why most of the experimental results conditions in orthogonal cutting. Journal of Applied Physics
show that for a given work material, the shear angle is 1945;16:31824.
more or less given by the Merchants prediction fM minus [6] Baker M. Does chip formation minimize the energy? Computational
a constant. This constant is identied to y/2 in our Materials Science 2005;33:40718.
[7] Miguelez H, Zaera R, Rusinek A, Moufki A, Molinari A. Numerical
modelling. modelling of orthogonal cutting: Inuence of cutting conditions and
A comparison with experimental data available in the separation criterion. Journal de Physique IV 2006;134:41722.
literature would be an interesting goal. However, to [8] Kalpakjian S. Manufacturing processes for engineering materials.
achieve this goal with quantitative accuracy, one should Reading, MA: Addison-Wesley; 1984.
[9] Zvorykin KA. In: Proceedings of the Kharko Technological Institute,
not only have precise experimental observations of the chip
Ukraine, 1893.
morphology near the primary shear zone (assuming a [10] Hill R. The mathematical theory of plasticity. 1st ed. Oxford:
continuous chip) but also a good characterization of the Clarendon Press; 1951.
friction law (assuming that no sticking does occur since the [11] Lubliner J. Plasticity theory. New York: Macmillan; 1990.

Das könnte Ihnen auch gefallen