Machining of composite
materials. Part I : Traditional
S. Abrate and D.A. Walton
Composite materials are more difficult to machine than metals mainly because they are anisotropic,
non-homogeneous and their reinforcing fibres are very abrasive. During machining,
introduced into the workpiece, and tools wear rapidly. Traditional
as drilling or sawing can be used with proper tool design and operating
of traditional machining methods applied to organic and metal matrix composites
in this article. The use of non-traditional machining methods such as waterjet, laser and ultrasonic
machining will be discussed in the second part.
Keywords: composites; cutting; drilling; machining; grinding;
INTRODUCTION
Composite materials are used extensively because their
higher strength to weight ratios and modulus to weight
ratios, when compared to metals, offer new opportunities
for design. However, being non-homogeneous, anisotropic,
and reinforced with very abrasive fibres, these materials
are difficult to machine. Significant damage to the
workpiece may be introduced and high wear rates of the
tools are experienced. Traditional machining methods
such as drilling, turning, sawing, routing and grinding,
can be applied to composite materials using appropriate
tool design and operating conditions. This article is
concerned with the application of traditional methods
to machining of polymeric matrix and metal matrix
composites. Non-traditional techniques, including water-
jet, laser, electro-discharge, and ultrasonic machining,
have also been applied to these materials and will be
reviewed in Part II.
Drilling is the most common composite machining
operation, since many holes must be drilled in order to
install mechanical fastenersl. Poor hole quality accounts
for an estimated 60% of all part rejections2, and since
holes are drilled in finished products, part rejections due
to poor hole quality prove very costly. The mechanics
of drilling composite materials will be examined along
with the quality of the hole and the effect of tool design
parameters. Band saws and circular saws with special
blade designs are used for cutting composite materials,
and special router bits have also been designed for
contouring composite parts. These developments will be
reviewed since they have direct industrial applications.
Other investigations which are more concerned with
fundamental understanding of the machining process will
also be considered.
Two types of ultrasonic machining are used for
composite materials. The first one, in which the tool is
operated as in normal drilling or turning operations but,
in addition, is subjected to high frequency, small
amplitude vibration, will be discussed here. In this case,
ultrasonic vibration of the tool are used to facilitate
0956-7143/92/020075-09 @ 1992
Composites Manufacturing Vol 3 No 2 1992
methods
defects are
machining techniques such
conditions. A review
is presented
delamination
the traditional machining process. With the second
approach, to be reviewed in Part II, an otherwise
stationary tool is vibrated ultrasonically, and a liquid
slurry containing grit placed between the tool and the
part actually performs the cutting.
DRILLING
Mechanics of drilling composite materials
The thrust and torque applied on a bit during drilling
operations depend on speed, feed rate, tool geometry and
tool wear. Experiments2 showed that thrust increases
steadily until a constant value corresponding to steady
drilling through the thickness of the laminate is reached,
and is followed by a sharp drop as the tool exits the
opposite side (Figure 1). A sharp decrease in normal
force as the bit enters the workpiece is always associated
with the introduction of delamination by mechanical
action of the tool peeling up the top layer of the laminate.
Delamination of the top layer can also be produced by
high thermal stresses generated by drilling, but, in that
case, no discontinuities are observed in the normal force
history. Delaminations near the exit side are introduced
when the tool acts like a punch separating the thin uncut
layer from the remainder of the laminate. This action is
associated with an almost instantaneous drop in normal
force from its steady state value down to zero.
Delaminations can be greatly reduced or eliminated by
reducing feed rates near the exit and using backup plates
to provide support and prevent deformations leading to
exit side delaminations. Analysis of the delamination
process are discussed in the next section.
During the steady portion of the drilling process,
significant oscillations of the normal forces are
observed3*4. The main component of this dynamic signal
corresponds to <he lamination frequency, defined as the
ratio of the number of piles per inch to the time taken
by the drill to penetrate one inch of materia13. This
frequency is generally quite small compared to the one
Butterworth-Heinemann Ltd
75
corresponding to tool rotation, and the magnitude of the
dynamic signal shows good correlation with the
waviness of the surface of the hole. Therefore, the normal
force signal contains important information concerning
possible damage introduced into the workpiece, tool
wear, which can be monitored to determine when the
tool should be replaced, and surface quality.
During drilling torque increases rapidly until the
cutting edges of the tool are completely engaged and then
increases linearly until a maximum value is reached,
followed by a slight drop after hole completion (Figure
2). The cutting torque corresponds to the end of the
engagement phase (Figure 2). The high difference
between the cutting torque and the maximum torque is
attributable to high frictional forces between the lands
of the drill and the wall of the hole. As drilling progresses,
the tool is in contact with the side of the hole over an
increasing area so that frictional forces at the interface
create increasingly higher resistant torque. After complete
penetration has occurred only a small decrease in torque
is observed which indicates that friction is the major
contribution to total torque. High temperatures and
slightly negative coefficients of thermal expansion
compound the problem by squeezing the dril14.
Increasing the feed rate and point angle of the
drill, respectively, produces a corresponding increase
and decrease in torque levels respectively5V6. With
d Knee
h of the number of holes drilled (Figure 3). Increases in
ro \
E
b investigations and were shown to be more important
z when drilling graphite-epoxy than with glass-epoxy due
-l-
Figure 1 Typical axial force history during
laminates
Axial displacement
Figure 2 Variation of torque during drilling:
maximum torque ; T,, torque after penetration
76
I I I I I I50
1oL
10 20 30 40 50 60
Number of holes drilled
Figure 3 Increase in axial force with number of holes drilled
unidirectional composites, a variation of torque at a
frequency corresponding to twice the drill rotation rate
reflects the variation of the stiffness of the workpiece with
direction7,8.
Thrust levels increase as feed rate, cutting speed or
point angle of the drill are increased5g6. Maximum
normal force and maximum torque both increase very
significantly with the number of holes drilled due to
chipping and wear of the cutting surfaces3s4,9-1’. Using
8 mm split point carbide drills to drill 4.5 mm thick
graphite-epoxy composite slabs at 2800 rev min- ’ and
a feed rate of 0.0152 mm min- ‘, Radhakrishnan and
Wu3 determined the variation of peak thrust as a function
thrust due to tool wear were reported in many
to the more abrasive nature of graphite fibres”. Higher
- Time normal forces usually introduce more damage to the
workpiece, particularly delaminations. However, the
maximum torque or maximum normal force does not
correlate well with surface finish3.
An analytical analysis of temperature distributions in
the workpiece and the tool when drilling laminated
glass-epoxy printed circuit board showed good agree-
ment with experimental results”. Heat generated during
drilling of composite
drilling is distributed differently than for metals where,
typically, 75% of the thermal energy is eliminated with
the chip material, 7% is absorbed by the workpiece and
18% by the tool. For carbon-epoxy, approximately 50%
of the energy is absorbed by the tool ; the remainder is
absorbed almost equally by the workpiece and the chips*.
Therefore, a larger fraction of the energy dissipated is
absorbed by the workpiece and also by the tool.
Temperatures as high as 200°C near the hole are
reported’. Spatial and temporal temperature gradients
are strongly affected by the thermal conductivity of the
material. Smaller temperature gradients are observed in
carbon-epoxy than in glass-epoxy or aramid-epoxy
materials under the same conditions. The upper limit on
the cutting speed is limited by the risk of introducing
thermal damage to the workpiece materia17*“, while a
lower limit is governed by the surface quality which
becomes poor as fibres recede in front of the cutting
T,, cutting torque; T,,,, edge”. Feed rates are limited to avoid excessive forces
that cause delaminations. Typical values of cutting speeds
Composites Manufacturing No. 2 1992
Table 1 Typical machining parameters for drilling composite materials

Hole Material Cutting Feed

Workpiece Tool diameter thickness speed rate
material material (mm) (mm) (m min-r) (mm rev-‘) Source

Unidirectional Carbide 4.85-1.92 O-12.7 42.1 0.0254-0.0508 13,19

graphite-epoxy 4.85-7.92 12.7-19.1 33.5 0.0254
PCD 4.85-7.92 O-12.7 61.0 0.0508-0.0889
4.85-1.92 12.7-19.1 51.8 0.0508-0.0889
Multidirectional Carbide 4.85-7.92 O-12.7 61.0 0.0254-0.0508
graphite-epoxy 4.85-7.92 12.7-19.1 42.7 0.0254
PCD 4.85-7.92 O-12.7 68.6 0.0508-0.0889
4.85-7.92 12.7-19.1 61.0 0.0508-0.0889
Graphite-epoxy Carbide 4.85 6.35 60.9 0.0254 20
Glasssepoxy HSS 12.5 15.0 0.028 5
Glass-epoxy HSS 3 10 33.0 0.05 11
Carbon-epoxy Carbide 3 10 33.0 0.05
Glass-epoxy HSS 8 1.2 O-40.2 20-460 mm min-’ 14
Boron-epoxy PCD 6.35 2.0 91-182 25.4 mm min- r 24
6.35 25.4 91-182 25.4 mm min- r
PCD
MMC 6 19.2 15-75 22
Carbide
Boron-epoxy PCD 6.35 10.4 79 41.91 mm min-’ 8
Kevlar-epoxy Carbide 5.6 158 0.05 10

and feed rates used for drilling composite materials are
given in Table 1.
Damage induced by drilling
Several types of damage are introduced during drilling
operations : matrix cratering and thermal alterations,
fibre pullout and fuzzing, interlaminar cracks and
delamination, in addition to geometrical defects commonly
found in metal drilling. Drill wear and delamination are
both influenced by the type of drill used. A delamination
factor 6 can be defined as the ratio between the maximum
diameter of the damaged zone and the diameter of the
hole’ : 6 reaches an upper limit as the number of holes
drilled increases. For a spiral point drill, the delamination
factor tends towards 1.2 after just three holes2. For solid
carbide split point drills, the delamination factor settles
around 1.8 after five holes, while the High Speed Steel
(HSS) split point drill settles at 2.5 after 4-5 holesi3.
The size of the damaged zone was related to drilling
parameters in a recent study i4 . The width of the damaged
zone in glass-epoxy laminates was determined by first
infiltrating a liquid penetrant through the cut surface and
measuring D, the width of the damaged zone, with an
optical microscope. D is shown to depend on the ratio
between the cutting speed V, and the feed rate V,. A sharp
decrease in damage width is observed first as V,/V,
increases ; beyond a critical value of I’,/ V, damage size
remains constant. The critical value, in the range of
100- 150, is independent of resin type, fibre format,
composite thickness and fabrication procedure. The
extent of damage, however, may depend on material
properties and lay-up.
The use of vision systems to detect fibre fraying on the
surface of the hole and to determine the size of
delaminations has been demonstrated15. Algorithms
required for image analysis were described.
Drill bit material and geometry
To machine fibre reinforced composites the material the
tool is made of must be carefully selected. Because glass
and carbon fibres are very abrasive drill bits made out
of High Speed Steel fail after just a few holes16,i7.
Tungsten carbide tools possess adequate life16-20,
particularly when submicron carbide is used, since its
resistance to rupture is 50% higher and it is harder than
the standard C2 grade16. A much higher number of holes
can be drilled with tungsten carbide tools coated with
polycrystalline diamond (PCD)‘~~~‘,~~. PCD coated tools
can easily be chipped, particularly when used with
portable drills, and cannot be sharpened. Another
approach involves grinding the point angle into a
carbide-tipped blank ; PCD veins are then sintered in that
groove, and the drill point geometry and flutes are ground
in’ 3. These diamond veined drills can be sharpened’3*21.
The type of drill to be used is then selected depending
on the type of composite material to be drilled and
whether it is already bonded to a metal part. In many
cases composites are bonded to aluminium or titanium
parts, and the hole to be drilled must go through both
the composite and the metal. Drilling aluminium with a
tool designed for drilling glass-epoxy calls for a very
different shape of tool in order to remove the long
stringy aluminium chipslg. Drilling titanium with the
same tool as composites is difficult because tool wear
makes drilling through titanium difficult. It is preferablelg
to drill the composite first, remove the tool, and drill the
titanium part with a different tool. Then, because the
titanium chips damage the surface of the hole in the
composite section as they are removed, the hole must be
reamed. This procedure, involving three tools and three
operations, is expensive. Ways are sought to provide
combination tools21 or to design drills that automatically
change operating conditions as the drill bit enters
different materials23. Many types of drill bits have been

Composites Manufacturing No. 2 1992 77

tested on glass-epoxy laminates”. The general geometry,
helix angle, and point angles for twist drills are shown
in Figure 4. The rake angles, relief angles, and chisel
angles of the cutting edge are defined in Figure 5.
For graphite-epoxy or glass-epoxy, best results are
obtained with a solid tungsten carbide dagger dri1116,r9
or with PCD coated twist drills19. Positive rake angles
are needed to generate the least amount of heat during
cutting7s’0,‘6. However, the more positive the rake angle,
the more fragile the cutting edge becomes. A small chisel
angle is the second element of good tool geometry and
serves to improve the penetration rate16. Helix angles
for drills used to machine composites with metal backing
are selected in order to remove the chips produced
effectively. Helix angles of 12” with titanium backing and
Figure 4 Geometry of twist drill : c(, helix
a b
Figure 5 Geometry of cutting edge for tool
cutting speed; t, depth of cut)
78
up to 35” with aluminium backing are selected19. The
different angles are necessary because of the difference in
shape of the chips produced.
Machining materials containing aramid fibres requires
special tooling”. Because of their low compressive
strength, aramid fibres have a tendency to recede within
the matrix instead of being shorn ~fflO*~*. Frayed fibres
will protrude from the hole surface to create what is
commonly known as ‘fuzz’. To drill this type of material,
fibres should be pulled from the periphery of the hole
towards the centre and then sheared. This can be
accomplished by tools with protruding peripheral cutting
edges and positive axial and radial rake angle@. Rake
angles and relief angles in the range of 29-35” and
12-30”, respectively, were found adequatelO. Self-
centering drills were designed so that serrations are
alternatively oriented upward and downward around the
circumference, introducing a cutting action in both
directions to cut unsheared aramid fibres”. A serrated
spade drill for a drill press operation and a serrated
countersink are also recommended’*. For these three
types of drill bits, serrations are designed to trap the
uncut fibres and shear them.
Drilling of composite materials with boron fibre
reinforcement requires the use of diamond impregnated
toolingz4. The major problem encountered with
boron-epoxy laminates comes from the heat generated.
Material thinner than l/8 in can be drilled without fluid
cooling. Above that thickness, coolant must be supplied
through the tool at pressures up to 75 psi to prevent
damage to the workpiece. When a layer of boron-
aluminium is bonded to one piece of titanium, a two tool
approach is preferable. A multi-layered section would
require too many tool changes and must be drilled with
one tool with a reduced feed rate when cutting through
a titanium layer. Tool wear was shown to depend on
prior heat treatment when drilling boron-aluminium
composites 24. While ultrasonically vibrating the tool was
not effective when drilling boron-epoxy, it reduced
friction, tool wear and the tendency of the aluminium to
accumulate on the tool when drilling boron-aluminium.
Drilling with an ultrasonically vibrated tool was also
successful in producing holes in boron-aluminium/
angle; p, point angle titanium laminatesz4.
a
C
C
with (a) positive, (b) neutral, and (c) negative rake angles: ara (are, relief angle; a,, chisel angle; V,
Composites Manufacturing No. 2 1992
 Experiments were performed22 drilling metal matrix
composites (MMC) with continuous and short fibre
reinforcement of aluminium matrices by alumina, silicon
carbide and boron carbide. Comparison of hardness of
tool materials and fibre reinforcement indicated that
High Speed Steel is not appropriate for drilling metal
matrix composites. The hardness of a-Al,O,, Sic, and
B,C fibres is higher than that of solid carbide tool
materials but significantly lower than that of PCD. Solid
carbide and PCD tools were used, and some twist drills
had interior channels for coolant. Drilling a-Al,O, short a f
fibre reinforced aluminium, the tool life of high-speed
steel tools was 100 times lower than that of solid carbide
tools. The life of PCD tools was 20 times that of solid
carbide tools when drilling continuous fibre reinforced
aluminium, and 100 times that for short fibre
reinforcement. Because of tool wear, machining forces
increase ten-fold over their initial values, but even then
they remain at a level corresponding to only one third
or one fourth the forces involved in drilling steel. The
use of coolant did not reduce tool wear in the experiments
performed in that study.
With abrasive tools, the cutting edge is not well defined
since the individual cutting particles are randomly
oriented, many of them with negative rake angles. Only b
small spaces are available for chips between the grains.
Therefore, intense heat is generated rendering these tools
unsuitable for drilling aramid reinforced plastics”.

Hole quality
The quality of the hole produced can be described in
many ways7. The surface microstructure can be assessed
by means of standardized roughness parameters. The
dimensional error is defined as the difference between the
diameter of the hole and that of the tool. The roundness
of the hole is characterized by a parameter equal to one
half the difference between the maximum and minimum
diameter of the hole. Another quality criterion is damage
to the material, which can be described by the width of
the heat affected zone and the size of the delaminated
area. C
With unidirectional laminates, the surface quality of
the hole varies considerably around the circumference7.
When the cutting direction is parallel to the fibre
orientation (0” ), individual fibres are pulled-out. When
this angle increases, compression and bending occur in
the 20-45” range, the fibres are pulled out of the cut
surface and diverted into the cut direction. The worst Figure 6 (a) entrance or peel-up delamination ; (b) exit or push-out
surface quality occurs in this range. When the cut delamination; (c) simplified modeP
direction is perpendicular to the fibre orientation, fibres
are subjected to shear and bending, and surface quality
improves’. Hole quality depends greatly on the part and must be avoided. During drilling, delaminations
mechanical properties of the reinforcing fibres7*10, can be introduced by three mechanisms: peeling up of
Aramid fibres have very low compressive strength, tend the top layer, punching out of the uncut layer near the
to recede into the matrix during machining, and then exit (Figure 6), and the thermal stress mode’.
appear frayed on the surface of the hole. Since, for A simple model for predicting thrust levels that will
unidirectional composites, the strength of the fibres is induce ‘push-out at exit’ or ‘peel-up at entrance’
usually higher in tension, cutting forces are greater when delaminations has been proposed2’. The delaminated
the cutting direction is parallel to the fibre direction19. area is assumed to be circular, and the uncut portion is
Angle-ply laminates are reported to be easier to machine modelled as an isotropic circular plate clamped on its
than unidirectional oneslo. contour to the cut portion of the laminate, which is
assumed to be rigid (Figure 6(c)). From the energy
DELAMINATION balance equation, relating the energy required to extend
the crack to the work done by the normal force applied
In laminated composite materials, delaminations cause by the drill and the strain energy and assuming Mode I
severe reductions in the load carrying capacity of the crack propagation, the critical load at the onset of

Composites Manufacturing No. 2 1992 79

delamination is obtained as
F,* = lr{(8G,,Eh3)/[3(1 - u~)]}“~
where G,, is the critical strain energy release rate, F,* is
the applied thrust force, h is the uncut depth under the
tool, E is Young’s Modulus, and u is Poisson’s ratio.
Because of the assumption made, the critical load is not
dependent on the cut portion of the workpiece, but it
decreases with h, since the uncut portion of the laminate
becomes more flexible. Good agreement with experi-
mental results taken from the literature is reportedz5.
The same model was used for peeling-up type
delaminations, assuming that the peeling force is related
to the horizontal cutting force by a constant called the
peeling factor.
The same problem was studied using a two-
dimensional axisymmetric finite element model of a small
zone around the hole and the too126. The tool end is
assumed to be flat, and the workpiece is clamped on its
outside surface. Failure is assumed to occur when
interlaminar tensile stresses reach a critical value at the
interface between the cut and uncut regions. Both regions
were modelled as either isotropic or orthotropic materials
using apparent mechanical properties. The model can
also account for a metal backing plate. Results confirm
the trends predicted by Equation (1). The push-out at
exit delamination mode in sawing operations was also
studied using a finite element mode127. Good agreement
with experimental results is reported.
CUTTING
Several experimental studies were conducted to improve
understanding of the cutting process in machining
composite materials. Shaping experiments were con-
ducted2*~2g on carbon-epoxy laminates in order to study
the shape and dimensions of the chips, the surface of the
workpiece and cutting forces. In ‘quick-stop’ experiments,
the shaping tool was brought to a stop in a very short
distance, corresponding to approximately two fibre
diameters28’2g, in order to examine the area near the tool
tip and determine the failure mechanisms involved.
Machining of unidirectional graphite-epoxy, both
parallel and perpendicular to the fibre direction, was
considered. When machining perpendicular to the fibres
the tool presses on the composite in front of it causing
it to fracture and creating a chip, while at the same time
it presses down on the material below the tool, producing
cracks 0.1-0.3 mm deep into the material. When
machining parallel to the fibre direction a crack often
runs ahead of the tip of the tool and cracks on the surface
reach ‘depths of only ‘one to two fibre diameters.
Examination of the chips indicated that they have not
been subjected to large plastic deformations and that the
cutting process consists of a series of fractures****‘.
Increasing the rake angle of the tool slightly reduced the
horizontal cutting force, while no definite trend was
noticed for the vertical force. Increasing the relief angle
dramatically decreased the vertical force, and increasing
the cutting depth increased the horizontal cutting force
significantly, while leaving the horizontal force essentially
unchangedZ9.
Orthogonal cutting of unidirectional glass-epoxy
composites with fibre oriented between 0 and 90” from
the cutting direction was studied, starting with a series
of turning experiments 30. Results indicated that cutting
80
forces are relatively low when cutting parallel to the fibre
direction, decrease slightly to a minimum when
machining at a 30” orientation, and then increase
significantly as the cutting direction becomes normal to
the fibre direction. A mathematical model is developed
to predict the cutting force resultant as a function of fibre
orientation and material properties, based on the
assumptions that chip formation occurs by quasi-
continuous shear and that shearing occurs in the
direction in which cutting energy is minimum. Good
agreement is obtained with experimental results over the
whole range of fibre orientations3’. Ultrasonic vibrations
of the tool at 18 kHz in the cutting direction were shown
to dramatically reduce cutting forces and the depth of
the damage zone over the range of fibre orientation from
0 to 18003’.
A detailed experimental study of tool wear and surface
finish when machining graphite-epoxy laminates with
PCD tools was conducted by Ramulu et aL31. Circular
plate specimens with a 152.44 mm diameter, a 25.4 mm
thickness, and a [(O/90/45/45),,], lay-up were
machined by face turning. The cutting edge of the tool
had a zero rake angle and a 12” relief angle. Cutting
speed was set at 228.6 m min-’ with a feed rate of
0.1778 mm rev-’ and a depth of cut of 4.76 mm. Flank
wear of the tool was shown to increase sharply with
machining time at first and then to stabilize. The average
peak to valley surface roughness parameter R, varied in
the 23-37 pm range for different grades of PCD tool
materials. The variation of surface roughness with fibre
orientation previously reported for drilling operations6,7
was also observed.
Most of the energy spent in cutting is converted into
heat, which causes temperature rises in the workpiece,
the tool and the chips. Turning experiments on
glass-epoxy materials with thermocouples embedded
below the cut surface were performed3* in order to
examine the relation between cutting temperature and
tool wear. The heat generated in cutting is concentrated
near the tool nose and causes tool wear in this area. A
detailed study of wear of HSS tools in face turning
operations is available33. Cutting temperatures depend
largely on the thermal conductivity of the too13*, which
is consistent with a study on drilling of composites7
showing that a significant fraction of the heat generated
during machining of composites is evacuated by the tool.
Another study of tool wear in the turning of glass-epoxy
with different tool materials, including cemented carbide
and ceramics34, confirmed the importance of the thermal
conductivity of the tool. This factor determines cutting
temperatures and wear rates.
An experimental study35 of the machinability of
aluminium alloys, reinforced by graphite particles with
an average diameter of 60 pm with volume fractions up
to 22%, indicated that lowering of cutting forces is
attributable to the reduction in shear flow stress rather
than to lower coefficients of friction at the chip-rake
interface. Similar reductions in cutting forces as
graphite particles are added to aluminium alloys were
reported36*37.
Cutting forces and chip formation in face turning of
MMCS with short alumina fibre reinforcement of
aluminium matrix was studied by Andersson et aL3’. For
small volume fraction reinforcements (6 and 12% ), chip
formation is dominated by the plastic flow of the matrix.
A model is presented to predict the tangential force on
Composites Manufacturing No. 2 1992
the tool. Experimental results verify the theoretical
prediction of a linear increase of tangential force with
increasing undeformed chip thickness.
The mechanics of machining fibre reinforced materials
was first studied analytically by Everstine and Rogers39.
Orthogonal machining parallel to the fibre direction was
considered, modelling the fibre reinforced material as
homogeneous, incompressible and inextensible in the
fibre direction.
SAWING
One set of experiments was conducted, cutting
graphite-epoxy laminates with a circular saw9T40,41.The
high speed steel slotting saw used had an outside diameter
of 57.15 mm, a thickness of 0.71 mm and 60 teeth. Feed
rate was held constant at 50.8 mm min-‘, while cutting
speeds of 1.5,3,6,9 and 12 m s-l were tried. The force,
applied perpendicular to the plane of the laminate, and
the feed force decreased, while temperatures near the
cutting zone steadily increased as cutting speed increased.
Near the cut surface, fibre orientation was altered over
a depth d as the tool pushed the fibres in the feed
direction. With increasing speed, d decreased the same
way cutting forces did. Delaminations and cracks leading
to delaminations were observed on the cut surface.
a b
Figure 7 Bandsawing of aramid laminates:
direction; (b) reverse direction
Interlaminar tensile strength of a small specimen with
one edge sawed increased with increasing cutting speed,
because damage decreased as did tool forces. The cutting
speed is limited to a maximum for which melting of the
matrix material occurs. On the other hand, damage size
decreases with increasing speed as cutting forces become
lower. Therefore, cutting speed should be as high as
possible without exceeding the acceptable temperature
level.
Composites with aramid fibre reinforcement up to
l/4 in thick can be cut with band saws and saber saws
with only slight fuzzing on the exit side”. However, the
blades are run in reverse (Figure 7) so that the heel of
the tooth enters the composite first, which gives a more
efficient shearing action. The edges of the teeth should
be honed to remove sharp edges which tend to pull the
Kevlar fibres and create fuzz.
Diamond plated saw blades are used for both
straight sawing and contour sawing of graphite-epoxy
materials4’. Boron-epoxy and boron-aluminium com-
posites can be cut to close tolerances using diamond
coated circular saws or band saws24. However, when
boron-epoxy is bonded to titanium, flood coolant is
required to prevent damage to the epoxy matrixz4.
ROUTING
-
Routing is often required to trim parts. With glass and
carbon reinforcement tools with multiple cutting edges
made out of cemented carbide or PCD materials are
recommended6. As in drilling operations, cutting forces
and surface quality depend on fibre orientation. Highest
forces and worst quality occur when fibres are oriented
at 45” from the cutting direction. PCD tools provide better
life than carbide tool@.
For composites with aramid fibre reinforcement,
opposed helical tools producing alternating forces across
the entire cut surface gave the best results”. Two router
bits were developed for composites with aramid
reinforcement using the principle of shearing the
outermost fibres towards the interior of the composite18.
The opposed helical cutter is used for thick sections or
honeycomb sandwich panels, while the V-slot router is
designed for thin laminates (Figure 8). With both of these
tools, alignment with the centre of the groove or the
(a) conventional cutting centre of the opposed helices is critical and hard to
achieve with hand held tools.

Table 2 Typical machining parameters for sawing of composite materials

Thickness Cutting speed Feed rate Type of

Material (mm) (m s-‘) (mm mint) saw Reference

Graphite-epoxy 4 2-12 50.8 Circular 40

(HSS)
Graphite-epoxy O-25.4 15.24 50.8 Bandsaw 42
(PCD)
Boron-eopxy 2.0 30.48 254 Circular* 24
ND)
Boron-epoxy 25.4 15.24 50.8 Bandsaw* 24
(PCD)
Boron-aluminium 1.52 28.5 152.4 Circular* 24
(Annealed) (PCD)
Boron-aluminium 12.7 28.5 25.4 Bandsaw* 24
(Annealed) (PCD)

*With coolant

Composites Manufacturing No. 2 1992 81

 Because of the abrasive nature of the reinforcing fibres,
carbide or diamond impregnated tools are satisfactory.
Precautions must be taken to avoid introducing
delaminations in the workpiece by limiting tool wear in
order to keep machining forces low and by using back-up
I I material to prevent delaminations as the tool exist the
workpiece. Special problems are encountered when
Aramid laminate machining composite containing aramid reinforcement
a due to the properties of these fibres. Special tool designs
are necessary in order to avoid ‘fuzz’.
The health effects of dust particles in the air produced
when machining reinforced polyesters was studied44,
indicating that if no special precautions are taken
respirable dust levels can exceed tolerable levels. One
solution to this problem is to use liquid coolant which
will carry away these particles. This coolant also serves
to reduce the cutting temperatures and is sometimes
necessary. However, for some materials moisture is easily
absorbed and affect the performance of the part and
should be avoided.

REFERENCES
b
1 Sprow, E.E. ‘Cutting composites: three choices for any budget’,
Toolina & Product& 43 No 12 (1987) vv 46-50
2 Wang,” T.L., Wu, SM. and ‘Croy,’ ‘GM. ‘An analysis of
delamination in drilling composite materials’ Proc 14th SAMPE
Tech Conf 1982, pp 471-483
3 Radkakrishnan, T and Wu, S.M. ‘On-line hole quality evaluation
for drilling composite materials using dynamic data’ .Z Eng for
Industry 103 (1981) pp 119-125
4 Di Ilio, A., Tagliaferri, V. and Veniali, F. ‘Tool life and hole
I / quality in drilling aramid and fibrous composites’, in Composite
Material Technoloav 1991. Proc of 14th Annual Energv Sources
Aramid laminate Technology Conf. ahd Exhibition, Houston, TX, Jan 2{123,1991,
ASME Publ. PD-Vol 37, pp 203-207
Gindy, N.N.Z. ‘Selection of drilling conditions for glass fibre
reinforced plastics’znt J Research25 No 8 (1988)~~ 1317-1327
Lambert, B.K. ‘Cutting and drilling of composite materials’
Carbide and Tool J 19 No 5 (1987) pp 31-34
Konig, W., Wulf, C., Grawq P. and Wiiierscheid, H. ‘Machining
of fibre reinforced plastics’ Manufacturing Technology CZRP
Annals 34 No 2 (1985) pp 537-548
8 Konig, W. and GraS, P. ‘Quality definition and assessment in
drilling of fibre reinforced thermosets’ Annals of the CZRP 38 No
Figure 8 Routing of aramid laminates with (a) helical router bit or 1 (1989) pp 119-124
V-slot bit; (b) top view; (c) side view 9 Sadat, A.B. ‘Machining of composites’ Encyclopedia of Composites,
Vol3 (1990) pp 95-102
10 Konig, W., Grass, P., Heintze, A., Okcy, F. and Schmitz-Justin,
C. ‘Developments in drilling & contouring composites containing
GRINDING Kevlar’ Production Engineer 63 No 8 (1984) pp 56-61
11 Sakuma, K., Yokoo, Y. and Seto, M. ‘Study on drilling of
A study of grinding of unidirectional glass-epoxy reinforced plastics (GFRP and CFRP)’ Bulletin of JSME 27
No 228 (1984) pp 1237-1244
composites concentrated on the topography of the cut 12 Nagao, T. and Hatamura, Y. ‘Investigation into drilling laminated
surface and the orientation of the reinforcing fibres43. printed circuit board using a torque-thrust-temperature sensor’
Two failure modes were observed for the glass fibres. For Annals of the CZRP 37 No 1 (1988) vv 79-82
orientation angles between 0 and 60” grinding grains dig 13 Beard, T. ‘Machining compos‘ites -‘new rules and tools’ Modern
Machine Shop 61 No 11 (April 1989) pp 74-85
up fibres, which results in deep fibre-matrix debonding
14 Tagliaferri, V., Caprino, G. and Diterlizzi, A. ‘Effect of drilling
and failure of the entire cross-section of the fibre. The parameters on the finish and mechanical properties of GFRP
end of the fibre forms a hollow pit on the ground surface. composites’ Znt J Mach Tools Manu 30 No 1 (1990) pp 77-84
When 90 < 0 c 180, the tool pushes down on the fibres 15 Griswold, N.C., Massanvek, WA. and Hougk, C.L. ‘Morphological
which then protrude from the surface. structures in analysis and inspection of hole quality in fibre
composites’JTestingandEvaluation 17No 5 (1989)pp281-286
16 Mackey, B.A. ‘How to drill precision holes in reinforced plastics
CONCLUSION in a hurry’ Plastics Engineering (1980) pp 22-24
17 Lamkert, B.K. ‘Find low-cost methodology when machining
Traditional machining methods can be used for drilling, composites’ Cutting Tool Engineering 39 No 7 (1987) pp 20-22
cutting, routing and grinding composite materials. In 18 Miner, L.H. ‘Cutting and machining Kevlar aramid composites’
Proc Kevlar Composites Symposium, El Segundo, CA, Dee 4-6,
order to minimize the amount of heat generated and 1979, pp. 85-93
avoid thermal damage to the part, the proper tool 19 Hickey, J. ‘Drilling graphite composites’ Modern Machine Shop
geometry and operating conditions must be adopted. 59 (March 1987) pp 84-90

82 Composites Manufacturing No. 2 1992

20 Miller, J.A. ‘Drilling graphite/epoxy at Lockheed’ Am Much reinforced plastics (the relationship between fibre orientation and
Auto Mu&131 No 10 (Ott 1987) pp 70-71 tool wear)’ JSME Bulletin 26 No 218 (1983) pp 1420-1427
21 Kinknid, R. ‘Quality holes in composites with PCD cutting tools’ 35 Brown, C.A. and Surappa, M.K. ‘Machinability of a cast aluminum
Cutting Tool Engineering 40 No 3 (1988) pp 50-52 alloy-graphite particle composite’ Muter Sci Eng A: Structural
22 Cronjager, L. and Meister, D. ‘Drilling of fibre and particle Materials: Properties, Microstructure and Processing Al02 No 1
reinforced aluminum’, in Composite Material Technology 1991, (1988) pp 31-37
Proc of 14th Annual Energy Sources Technology Conf and 36 Rohatgi, P.K., Murali, N., Shetty, H.R. and Chandrashekhar, R.
Exhibition, Houston, TX, Jan 20-23, 1991, Vol 37, pp 185-189 ‘Improved machinability of graphite particle-aluminum alloy
23 Bak, D.J. ‘Drill senses hardness changes as it passes through composites’ Mater Sci Eng 26 (1976) pp 115-122
stackups’ Design News 45 (1989) pp 182-183 37 Gibson, P.R., Clegg, A.J. and Das, A.A. ‘Production and
24 Doran, J.H. and Maikish, C.R. ‘Machining boron composites’ evaluation of squeeze-cast graphitic Al-Si alloys’ Muter Sci
Composites Materials in Engineering Design, B.R. Noton, Ed., Technoll (1985) pp 559-567
ASM, 1973, pp 242-250 38 Andersson, C.H., Stahl, J.E. and Andersson, M. ‘Plastic
25 Ho-Cheng, H. and Dahran, C.K.H. ‘Delamination during drilling deformation and machining properties of some low volume
in composite 1aminates’J Engfor Industry 112 (1990) pp 236-239- fraction short-fibre reinforced metal matrix composite materials’,
26 Sadat. A.B.. Chan. W.S. and Wane. B.P. ‘Delamination of in Mechanical and Physical Behaviour of Metallic and Ceramic
graphite/epoxy laminate during drilling operation’ Composite Composites, Proc 9th Riso Int Symp on Metallurgy and Mater
Material Technology 1990, ASME PD. Vol32, D. Hui, T.J. Kozik, Sci Sept 5-9 (1988), S.I. Andersen et al. eds, pp 263-268
Eds pp 85-88 (1990) 39 Ever&e, G.C. and Rogers, T.G. ‘A theory of machining of
27 Sadat, A.B., Chan, W.S. and Wang, B.P. ‘Prediction of edge fibre-reinforced materials’ J Coma Mater 5 (1971) DD 74-106
delamination load in a sawing operation of graphite/epoxy 40 Sadat, A.B., ‘Machining of graphite/epoxy dompd& materials’
laminate’, in Composite Material Technology 1991, Proc of 14th SAMPE Q 19 No 2 (Jan 1988) pp 104
Annual Energy Sources Technology Conf and Exhibition, 41 Sadat, A.B. ‘The effect of cutting speed and width of cut on
Houston, TX, Jan 20-23, 1991, pp 113-116 interlaminar tensile strength of graphite/epoxy composite
28 Koplev, A. ‘Cutting of CFRP with single edge tool’ Proc 3rd Int material’ ASTM J Composite Technology and Research 10 No 4
Conf Comp Mater, ICCM3, Paris, 1980, pp 1597-1605 (1988) pp 173-176
29 Koplev, A., Lystrap, A. and Vorm, T. ‘The cutting process, chips, 42 Friend, C.A., Clyne, R.W. and Valentine, G.G. ‘Machining
and cutting forces in machining CFRP’ Composites 14 No 4 (1983) graphite composite materials’, Composite Materials in Engineering
pp 371-376 Design. B.R. Noton. Ed. (ASM 1073) DD 217-224
30 Takeyama, H. and Iijima, N. ‘Machinability of glass fiber 43 Inouk, H. and Kawaguchi; I. ‘Study on ihe grinding mechanism
reinforced plastics and application of ultrasonic machining’ CIRP of glass fiber reinforced plastics’ J Eng Mater Technolll2 (1990)
Annals 37 No 1 (1988) pp 93-96 pp 341-345
30 Takeyama, H. and Iijima, N. ‘Machinability of glass fiber 44 Antonsson, A.B. and Runmark, S. ‘Airborne fibrous glass and dust
reinforced plastics and application of ultrasonic machining’ CIRP originating from worked reinforced plastics’ Am Indust Hygiene
Annuls 37 No 1 (1988) pp 93-96 Assoc J 48 No 8 (1987) pp 684-687
31 Ramulu, M., Faridnia, M., Garb& J.L. and Jorgensen, J.E.
‘Machining of graphite/epoxy composite materials with poly-
crystalline diamone (PCD) tools’J Eng Mater Technolll3 (1991)
pp 430-436 AUTHORS
32 Sakuma, K. and Seto, M. ‘Tool wear in cutting glass fibre
reinforced plastics (the relationship between cutting temperature The authors are with the Department of Mechanical and
and tool wear)’ JSME Btdietin 24 No 190 (1981) pp 748-755 Aerospace Engineering and Engineering Mechanics,
33 Santhanakrishnan, G., Krishnamurthy, R. and Malhotra, S.K.
University of Missouri-Rolla, Rolla, MO 65401, USA.
‘High speed steel tool wear studies in machining of glass-fibre
reinforced plastics’ 132 No 2 (1989) pp 327-336 (Received 2 October 1991; accepted in revised form
34 Sakuma, K. and Seto, M. ‘Tool wear in cutting glass fibre 6 April 1992)

Composites Manufacturing No. 2 1992 83