Anda di halaman 1dari 9

composite .%uchues vol. 38, No.

l-4, pp, 531-539, 1997


0 1997 Published by Elsevier Science Ltd. All rights resewed
Printed in Great Britain
0263-8223/97/$17.00 + 0.00
ELSEVIER
PII:SO263-8223(97)00089-S

Development of a composite boring bar


Shuzo Nagano, Takayuki Koizumi, Toru Fujii, Nobutaka Tsujiuchi, Hiroki Ueda
& Kobe Steel
Doshisha University,Tanabe, Kyoto 610-03, Japan

A composite boring bar, whose stability against chattering is superior to not


only conventional steel bars but also cemented carbide bars, has recently
been developed. The main material of this composite bar is pitch-based
carbon fiber reinforced plastic. Carbon fibers aligned unidirectionally in the
longitudinal direction of the bars give high bending stiffness. Four types of
bar having different shaped steel cores were designed by FEM analysis and
produced for actual testing. A bar having a cross-shaped steel core shows
the best cutting capability and stability amongst all bars designed. This bar
can be used when the length (~5) and diameter (0) ratio L/D is 7 or even
at severe conditions while a cemented carbide bar cannot control the
chatter vibration even if the L/D is less than 6. Emphasis should be placed
on the fact that the cross-shaped steel core can increase the bending
stiffness of the bar in both tangential and radial directions by constraining
the shear deformation of the fiber layers without sacrificing the increase of
resonant frequencies. 0 1997 Published by Elsevier Science Ltd.

INTRODUCTION the stability against chattering for boring bars,


one must use materials which have a high stiff-
When cutting long holes into the internal sur- ness but low density. Carbon fiber reinforced
face of machine components, chattering plastics (CFRP) could meet such requirements
frequently occurs due to the low bending stiff- if high modulus fibers were to be used. Not only
ness and low damping ability of the boring bar high stiffness and high resonant frequencies, but
when the arm of the bar is long. Once the chat- also high damping can be expected because of
tering occurs, the roughness of the surface of the polymer matrix of CFRP [4].
the machine components becomes unacceptable The objective of the present work is to
to any standard, including the dimensional develop a new boring bar using CFRP which
accuracy, and the cutting edge is often broken. can be used at a wide range of L/D over 6. How
In the case of conventional steel boring bars, it to design the bar and its performance are dis-
becomes impractical to cut any metal compo- cussed and shown in this paper.
nents due to chattering when the ratio of bar
length L to bar diameter D (L/D) is larger than
4-5 [l]. Therefore, bars made of cemented car- DESIGN OF CFRP BORING BAR
bide are usually used when L/D is 4-6. When
L/D exceeds 5-6, specially devised boring bars Chattering
having anti-vibration mechanisms are some-
times required [2,3]. However, these are Vibration during operation is classified as
expensive. forced vibration and self excited vibration.
Chattering during cutting occurs as self Chattering of boring bars during internal
excited vibration. According to the chattering cutting is due to the self excited vibrations
theories based on the l-DOF vibration system, having a regenerative feedback, generally called
the bending stiffness and resonant frequencies regenerative chattering. In practice, boring bars
of boring bars govern chattering as well as the are designed to be stable against regenerative
damping ability of the bars. In order to increase chattering. Figure 1 shows the regenerative

531
532 S. fVugun0 et al.

Horizontal
Workpiece
/-

Regenerative effect
i- Tool
Fig. 3. Conceptional illustration of the regenerative effect.
Fig. 1. Merritts model.

chattering model based on the l-DOF vibration cutting force F is a function of u(t) and is given
system gived by Merritt [5]. Considering outer as
cutting in this model, the short tool is fixed
while the workpiece is sustained by a spring and F(t) = k,(u)t (2)
a damper. For inner hole cutting using a boring
bar, a workpiece is large enough to be fixed where k, is a cutting stiffness varying according
while the bar should be represented by a mass- to cutting conditions such as workpiece
spring-damper combination as shown in Fig. 2. material, cutting speed, feed and stiffiness of
In the figure, F, is the cutting force acting in the lathe. Based on the l-DOF system shown in
the tangential direction. Due to this force, the Fig. 2 the governing equation for a boring bar is
bar is bent in the vertical plane. FH is the radial given by
force acting in the radial direction of the hole
(radial force is known as shear force in outer F( t ) = mji( t)+cj( t)+ky( t) (3)
cutting). This force causes the bar to bend in
the horizontal plane. Figure 3 illustrates how where m is an equivalent mass of the boring
the regenerative effect due to former cutting bar, c and k are an equivalent damping coeffi-
traces occurs. cient and an equivalent stiffness of the bar,
Actual cutting depth u(t) at time t is defined respectively. By solving eqn (3) in conjunction
as with eqns (1) and (2), it is obvious that high
n equivalent stiffness and damping ability give
u(t) = uo- y(t)+py(t - (1)
high chattering stability to boring bars.
where ug is the initial cutting depth: y is the It is known that the frequency of chattering is
displacement of the cutting edge in the direc- almost equal to the first resonant frequency of a
tion of the radial force: p is all overlap ratio of boring bar. From eqn (4), if the maximum
the cutting edge determined by cutting edge acceleration of a cutting edge, a, is constant, the
shape, cutting depth and feed: i_~governs the displacement of the cutting edge 6 decreases
magnitude of the regenerative effect; T is a with increasing chattering frequency f
periodic time for the workpiece revolution. The

Boring-Bar
Equation (5) shows the well-known relation-
ship betweenf, m and k

Workpiece/ So, f increases with decreasing m. Thus, light-


Fig. 2. Modified model. weight bars have the advantage of stability
Development of a composite boring bar 533

against chattering. In this point CFRP is one of


most attractive materials for a boring bar.

Carbon fibers

The previous section reveals that high stiffness


k, high damping c and small equivalent mass m 4 Cuttinghead steel
give a high stability of the bar for regenerative 3 Adhession mw
chattering of the system. These qualities are 2 Adapter Steel-.
also preferable for anti-chattering. Therefore, 1 Rod CFRP
high stiffiness and light weight CFRP is an Stnl Abptor md CFRP rod are assembled
attractive material for boring bars. Today, by adhesstve and cutting head la scmwed
on rdaptor.
graphite/epoxy composites are widely applied
Fig. 4. Total assembly of the CFRP boring bars.
not only on aero or astronomic structures but
also general products such as sports goods. For
such applications, virtually only pan-based fibers shows the schematic view of the CFRP boring
are used. The cost of these fibers is reasonable bars manufactured by way of trial. The adapter
but the Youngs moduli of such fibers are not whose inner hole was tapered, was bonded to
enough to give bars enough bending stiffness as the bar using an epoxy adhesive. The commer-
compared to conventional steel bars. The cially available cutting head is fixed to this
Youngs modulus of CFRP applied to boring adapter using three bolts. When the CFRP bor-
bars must he higher than 200 GPa (the Youngs ing bars are fixed to a lathe, a specially devised
modulus of steel). The longitudinal Youngs bar holder is used because the bars cannot be
modulus EL of CFRP can be estimated by the subjected to concentrated loads given by
law of mixtures as follows when fibers are standard fastening bolts. For sufficient fixing of
aligned unidirectionally. boring bars made from unidirectional CFRP,
four types of different shapes of steel core were
E, = E,V,+E,( 1 - V,) = E,V, (6) considered.
where EF and E, are the Youngs moduli of the
fibers and the matrix, respectively; V, is a fiber Type p
volume content. Usually, E,,, is much lower than In the case of a Type P boring bar, a steel pipe
EF and V, is higher than 50%. Therefore was used as the center core of the bar (as
E,(l - V,) in the above equation is negligible. shown in Fig. 5(a)) since lubricant is often used
According to eqn (6), E, must be higher than during the cutting operation. The outer diam-
400 GPa which implies the same stiffness as eter of the pipe is 15 mm while inner diameter
steel bars. Pitch-based carbon fiber, whose is 8 mm. This steel pipe core is also useful to
Youngs modulus is higher than 700 GPa is now give an accurate diameter of the bar by machin-
commercially available. Considering the varia- ing after fabricating the bar in an auto clave
tion of material data, pitch-based carbon fibers using unidirectional prepreg.
which has a nominal Youngs modulus of The effect of shear deformation is appreci-
700 GPa was adopted for the prototypes of able when the cantilever beam bar is relatively
CFRI boring bars. The average fiber volume short since the shear modulus of CFRP is over
content was about 55% and the observed 100 times lower than the longitudinal modulus
Youngs modulus of CFRP was 300-350 GPa. of CFRP. It is expected that this small center
core cannot restrict shear deformation due to
Construction of boring bars the cutting and radial forces.

The diameter of the boring bar is to be 32 mm, Type v


considering fabrication and the standard tool In order to improve the bending stiffness, addi-
size. As the cutting head, which holds a cutting tional deformation due to shear deformation of
tip, cannot be attached directly to the CFRP CFRP layers should be reduced. For such a
bar, a steel adapter was developed. Figure 4 requirement, different types of steel core were
534 S. Nuguno et al.

considered. Near the neutral axis of a beam, the cal plane. The vertical plate is effective for the
shear stress becomes high when the beam is shear deformation due to cutting force.
subjected to not only a bending moment but
also a shear force. Therefore, if a material with
Type H
a high shear rigidity is used as the center plate Type H CFRP composite boring bar has a
core of the composite bar, the total deflection 2 mm center plate core horizontally embedded.
of the bar can be reduced. Figure 5(b) shows a The horizontal plate is effective for shear defor-
Type V CFRP composite boring bar with a mation due to the radial force. Usually, the
2 mm center plate core embedded in the verti- radial force is smaller than the cutting force.
Figure 5(c) shows Type H bar. This bar is made
by rotating the Type V bar on its longitudinal
CFRP axis by 90.

Type V and Type H bars can constrain only


unidirectional shear deformation. However, two
plates should be used in both directions as the
core of a composite bar if both forces are rela-
/ tively large. For such a case, a cross-shaped
(a) Type P steel core as shown in Fig. S(d) must be effect-
ive even if the total weight of the bar is
&FRP sacrificed to some extent.

FEM ANALYSIS

The bending stiffness and the first resonant fre-


quency of all types of CFRP boring bars were
/Steel core calculated by FEM analysis using SDRC
l-DEAS.
(b) Typev For a Type C boring bar, the stiffness is esti-
mated when the thickness of the steel core
CFRP varies. Figure 6 shows FE divisions for all bor-
ing bars. Both steel and cemented carbide
boring bars are also analyzed using the same FE
division of Type C. Solid elements having eight
nodes were used. Each node has three degrees
of translation. The material data for calculation
for composite bars are given in Table 1. These
/ Steel core values were estimated by material test. Bound-
(~1 Type H ary conditions of the FE model is shown in
Fig. 7. The overhang of the boring bars is
CFRP 224 mm and the diameter of the bars is 32 mm
(L/D = 7). All degrees of freedom of the sur-
face nodes contacting the internal surface of the
boring bar holder are constrained, as shown by
the gray area in Fig. 7.

Bending stiffness

(4 Type C The bending stiffness of the composite bars is


Fig. 5. Constructions of CFRP boring bars. (a) Type P. (b) calculated by concentrating a force on the node
Type V. (c) Type H. (d) Type C. located at the free edge of the boring bar when
Development of a composite boring bar 535

(a) Type P CFFLP boring bar (b) Type V CFRP boring bar

0.80 I
01234 5670
Thickness of the steel part of
the CFRP boring bars, mm
Pig. 8. Relationship between the thickness of the steel
core of the CFRP boring bars and bending stiffness (Ll
D = 7).

CFRP, steel and cemented carbide bars at Ll


(c) TypeH CFRP boring bar (d) TypeC CFRP boring bar D = 7. The variation of bending stiffness with
Pig. 6. FE divisions of CFRP boring bars. (a) Type P respect to core thickness is also given for a
CFRP boring bar. (b) Type V CFRP boring bar. (c) Type Type C boring bar. The ordinate is normalized
H CFRP boring bar. (d) Type C CFRP boring bar. by that of the steel boring bar.
If a whole bar is made of CFRP instead of
set-up as a cantilever beam in the vertical direc- steel, the tensile stiffness or bending stiffness
tion. under a pure bending moment would be 160%
Figure 8 shows a comparison of bending stiff- (320 GPa) higher than those of steel. However,
ness in the direction of cutting force among the bending stiffness under both bending
moment and shear force becomes 87% of that
lhble 1. Material properties of the CFRP used in FEM
for the steel boring bar. Use of CFRP does not
analysis
always increase the bending stiffness of canti-
Youngs modulus EX EZ
E, lever beam bars. The reason of this low stiffness
@Pa)
was explained in the above section. From Fig. 8,
6.85 6.85 320 even a 1 mm thick cross-shaped steel core is
Shearing modulus GYZ GLX effective to constrain shear deformation. Up to
@Pa) 6 mm, the bending stiffness of the composite
0.533 2.54 2.54 bar increases with increasing thickness. Shear
Poisson ratio % YYZ V7X
deformation is much constrained with increas-
0.23 0.004 0.004
ing thickness of the steel core, but a thicker
core also reduces the bending stiffness more,
due to the low Youngs modulus of steel. The
maximum increase in bending stiffness is
around 20% using CFRP. The Type H com-
bination has a lower bending stiffness than that
of the steel bar since the steel plate core does
not constrain the shear deformation of CFRP
layers in the Y-Z plane. For a Type P boring
bar, the bending stiffness is only 4% as high as
that of the steel bar. The difference in bending
stiffness between a Type P bar and steel
becomes less with a decrease of L/D.Finally,
224 140 the bending stiffness for a Type P bar becomes
420
lower than that of the steel bar. The Type V bar
has a similar stiffness to Type C bar since the
Fig. 7. Boundary conditions of the FEM models. core plate vertically aligned is effective in con-
536 S. Nugunoet al.

straining shear deformation of CFRP layers in for bending stiffness, the maximum gain is
the Y-Z plane. On the other hand, the cemen- obtained at 6-7 mm. In the present study, the
ted carbide boring bar has a bending stiffness thickness of all steel plate cores is 2 mm as it is
2.37 times higher than that of the steel boring expected that the resonant frequency is effective
bar. One resolution to improve the bending for the chattering stability.
stiffness is to use carbon fibers with a higher
Youngs modulus. Although some fibers have
Youngs modulus higher than 1000 GPa, they CUTTING EXPERIMENTS
are not yet commercially available.
Cutting experiments were conducted to evaluate
Resonant frequency cutting performance and stability against chat-
tering for CFRP bars (Type P, Type V, Type C,
Modal analysis of the models was also con- Type H and Type C) as well as conventional
ducted to estimate the first resonant frequencies steel and cemented carbide bars. The diameter
of the bars (bending mode I). In Fig. 9, a com- of all bars is 32 mm. A conventional lathe (not
parison of the first resonant frequency for a NC lathe) was used for the tests. Here, all
CFRP, steel and cemented carbide bars at Ll boring bars were fixed using the specially
D = 7 is given. The variation of the first devised bar holder to provide equal conditions
resonant frequency with respect to core thick- of holding in the tool fixture as shown in
ness is also given for the Type C boring bar. Fig. 10. An overhang of 224 mm gives an L/
The ordinate is normalized by that of the steel D = 7. Thick cylindrical pipes (inner diameter:
boring bar as well as in Fig. 9. 60 mm, outer diameter: 100 mm) whose
In contrast to bending stiffness, the resonant material was mild steel were used as a work-
frequencies of CFRP boring bars are always piece for the tests. Before the cutting
higher than that for the steel bar. They are also experiment, the surface of the workpiece was
comparable to the resonant frequency of the smoothed to give a constant testing condition.
cemented carbide bar because the composite Two tiny accelerometers were attached on the
bars are one third to one quarter lighter than cutting head near the cutting edge to measure
the steel bar. The cemented carbide bar has a accelerations in the tangential and radial direc-
higher bending stiffness but it is extremely tions during the cutting operation. Cutting
heavy. From a viewpoint of resonant frequen- conditions are given in Table 2.
cies, high performance of CFRP boring bar can As noise occurs and chatter marks can be
be expected. It is obvious that the first resonant distinguished on the cutting surface of the
frequency increases once with increasing in workpiece when chattering occurs, it is easy to
thickness and then it decreases from the Type C identify whether chattering has occurred or not.
bar result. The maximum increase in resonant
frequency is about 30% at 2 mm thickness while
RESULTS AND DISCUSSION

Cutting limits for all boring bars are plotted on


a feed-cutting depth map (Fig. 11) at L/D = 7.

- 0 Type H boring bal


Type P boring bar

Workpiece Screw of
t Boring bar
E 1.10 -
T tooi post
holder
E
Steel boring bar \ .\
s 1.00
s! Boring bar
jJ 0.90 /
ii 0 1 234 56 70
Thickness of the steel part of
the CFRP boring bars, mm
Fig. 9. Relationship between the thickness of the steel
core of the CFRP boring bars and first resonant frequency
(LID = 7). Fig. 10. System for cutting test.
Development of a composite boring bar 537

Table 2. Cutting conditions in the measurement that the bending stiffness in the radial force
Revolution Cutting Feed direction is also as important as the bending
(rpm) depth (mm/rev) stiffness in the cutting force direction.
(mm) Figures 12 and 13 show the dynamic response
630 0.4 0.3 for Type C CFRP and steel bars during cutting
0.2 :*:
operation. Parts (a) are the acceleration-time
0:2 history plots, parts (b) are the power spectra for
500 0.4 0.3 trace (a) and parts (c) are the Lissajious plots
0.2
of acceleration in X and Y directions. It must be
0.2
::: noted that the surface of the workpiece cut by
the steel boring bar was extremely rough and
chattering occurred. Even for the Type C bar,
oscillation occurred but the magnitude of oscil-
As before mentioned, the steel bar was not able
lation is much smaller than that for the steel
to cut without chattering when the L/D value bar. The period of the main oscillation corre-
was greater than 4. Even when L/D was smaller sponds to the first resonant frequency for both
than 4, it was difficult for the steel bar to cut cases.
without chattering under several conditions. No Here, the maximum acceleration of Type C
marks in the figure mean that smooth cutting and steel bars are 11 and 280 G and the fre-
without chattering was not attained for the quency of oscillation for the bars are 500 and
corresponding bars. Therefore, a mark for the 375 Hz, respectively. Maximum displacement of
steel bar cannot be found in this figure. It is the cutting edge of the Type C and steel bars
found that the Type C CFRP bar has excellent are calculated as 0.011 and 0.43 mm, respec-
performance. Although the cemented carbide tively, from eqn (4). It is found that the
sometimes attained smooth cutting at lighter machined surface cut by the Type C bar is
conditions than those for the Type C bar, even acceptable for the roughness and dimensional
at L/D = 7, it was not always stable. Once chat- accuracy. However, it is realized that the maxi-
tering occurred, it did not stop for the mum displacement of the cutting edge of the
cemented carbide bar while the Type C CFRP steel bar is larger than cutting depth and the
bar was always stable. Good stability and cutting edge was beating the surface of the
cutting performance could not be obtained for workpiece. In this point, the advantage for chat-
both Types V and H bars. In particular, the tering stability of the Type C bar is obvious.
Type V bar shows a high bending stiffness Although the Type C bar has a high cutting
almost equal to the Type C bar and the first ability and chattering stability as above, it has
resonant frequency is higher than that of the no problem entirely. From Fig. 12(b) and
Type C bar in the cutting force direction. How- Fig. 13(b), the resonant frequency for the Type
ever, the bending stiffness of the Type V bar in C bar is higher than that for the steel bar. How-
the radial force direction is equal to the bend- ever, an increase of the resonant frequency is
ing stiffness of the Type H bar and it is much not remarkable, as expected in Fig. 9. No
lower than that of the Type C bar. It is shown apparent reasons could be found for this dis-
crepancy. Holding CFRP boring bars tightly is a
problem and this could be one of reasons for
0.8 the above discrepancy. The method for holding
0 Type C boring bar
rD Type V boring bar
the bar must be considered in the near future as
E 0.6 well as how to attach the cutting head to the
9 Type H boring bar
S-
8 0.4 0 Type P boring bar bar. Surface protection for lubricant and tips is
e Cemented carbide
boring bar
also one of issues to be considered for practical
.-@ 0 Steel boring bar use.
2 0.2

CONCLUSIONS
0 0.1 0.2 0.3 0.4 Nomarkmeans
impossible to cutting
Feed, mm/rev 1. A boring bar made from unidirectional
Fig. 11. Cutting limit of the boring bars. CFRP which has a high stability against chat-
538 S. Nuguno et al.

tering under severe conditions at L/D = 7


was successfully developed.
2. The cross-shaped steel core embedded
CFRP boring bar can constrain shear defor-
mation and improves the equivalent bending
stiffness of the boring bar.
3. An optimized cross-shaped steel core
improves the dynamic characteristics of the
total structure of the boring bar.
4. Compared with the cemented carbide boring
bar, chattering can be completely suppressed,
even in the range of L/D ratios greater than
7.

-20 I I
0 20 40 60 80 100
Time, ms

(a) Acceleration - Time history

12

10

-ml I
0 20 40 60 60 100
Time,ms

(a) Acceleration - Time history

300 ,
250

0 _..J L u 200
0 500 1000 1500 2000
.g 150
Frequency, Hz %
h
100
(b) Power spectrum for (a) 4
50

20 0 500 1000 15CiJ 2000

Frequency.Hz
15
(b) Power spectrum for (a)
10
4001
0
s 5
E 200
E O (3
s
s
8 -!j E
!! 0
s!
a -10
3
-15 4 -200

-20
-20 -15 -10 -5 0 5 10 15 20
-200 0 200 400
Acceleration, G Acceleration, G
Fig. 12. Dynamic response for the Type C CFRP boring Fig. 13. Dynamic response for the steel boring bar
bar (630 rpm, cutting depth = 0.4 mm, feed = 0.3 mm/rev): (630 rpm, cutting depth = 0.4 mm, feed = 0.3 mm/rev): (a)
(a) acceleration-time history, (b) power spectrum for (a), acceleration-time history, (b) power spectrum for (a), (c)
(c) Lissajious plot of acceleration in X and Y directions. Lissajious plot of acceleration in X and Y directions.
Development of a composite boring bar 539

REFERENCES 3. Kitajima, K. and Tanaka, Y., J. Japan Sot. precision


engng, SO(5) (1984) 860.
1. Hoshi, T., Vibration ana&is of mechanical cutting, 4. Kitajima, K. et al., J. Japan Sot. precision engng, 53(10)
Kegyo Chasakai Publishing Co. Ltd, 1990. (1987) 1582.
2. Takeyama, H. et al., J. Japan Sot. precision engng, 5. Merritt, H. E. Trans. ASME, Ser: B, 87(4) (1965) 27.
48(12) (1982) 1628.