Int J Adv Manuf Technol (2014) 72:943–953

DOI 10.1007/s00170-014-5730-x

ORIGINAL ARTICLE

An experimental investigation on micro machining

of fine-grained graphite
Dehong Huo & Chao Lin & Kenneth Dalgarno

Received: 20 November 2013 / Accepted: 16 February 2014 / Published online: 2 March 2014
# Springer-Verlag London 2014

Abstract Industrial applications of the micro milling process conditions to demonstrate that micro milling is an ideal pro-
require sufficient experimental data from various micro tools. cess for graphite machining.
Research has been carried out on micro milling of various
engineering materials in the past two decades. However, there Keywords Micro machining . Micro milling . Moulded
is no report in the literature on micro milling of graphite. This fine-grained graphite . Design of experiments . ANOVA .
paper presents an experimental investigation on micro ma- Micro EDM electrode
chinability of micro milling of moulded fine-grained graphite.
Full immersion slot milling was conducted using diamond-
coated, TiAlN-coated and uncoated tungsten carbide micro 1 Introduction
end mills with a uniform tool diameter of 0.5 mm. The
experiments were carried out on a standard industrial preci- The need to fabricate high accuracy miniature components
sion machining centre with a high-speed micro machining with dimensions ranging from a few hundred microns to a few
spindle. Design of experiments (DoE) techniques were ap- millimetres or features ranging from a few to a few hundreds
plied to design and analysis of the machining process. Surface of microns has encouraged research on development of micro
roughness, surface topography and burrs formation under manufacturing processes. It is well recognized that micro
varying machining conditions were characterized using white manufacturing has been a key enabling technology in indus-
light interferometry, SEM and a precision surface profiler. trially producing useful micro products and processes [1, 2].
Influence of variation of cutting parameters including cutting Micro electrical discharge machining (micro EDM) is a
speeds, feedrate and axial depth of cut on surface roughness promising machining technology for the manufacture of micro
and surface damage was analysed using ANOVA method. The parts and structures over a wide range of engineering mate-
experimental results show that feedrate has the most signifi- rials. The strengths of micro EDM lie in its force-free material
cant influence on surface roughness for all types of tools, and removal process, decent attainable accuracy and surface finish
diamond tools are not sensitive to cutting speed and depth of on the workpiece; hence, it is especially suitable for micro
cut. Surface damage and burrs analysis show that the primary machining hard-to-machine materials, e.g. hardened steel,
material removal mode is still brittle fracture or partial ductile titanium alloys and electrically conductive technical ceramics,
in the experimental cutting conditions. 3D intricate micro for fabrication of micro dies and moulds, and micro tools, in a
EDM electrodes were fabricated with good dimensional ac- small batch production.
curacy and surface finishes using optimized machining Central to the micro EDM technology is the fabrication of
micro electrodes especially 3D complex shape micro electrodes
D. Huo (*) : K. Dalgarno
for micro die sink EDM. These micro EDM electrodes have
School of Mechanical and Systems Engineering, characteristic dimensions normally in the order of tens to hun-
Newcastle University, Newcastle Upon Tyne NE1 7RU, UK dreds of microns, which will allow micro machining of features
e-mail: [Link]@[Link] with a projection area of 1 mm2 or less [3]. Among several
D. Huo : C. Lin
materials available for EDM electrodes, e.g. copper, tungsten-
The State Key Laboratory of Mechanical Transmission, copper, graphite, etc., graphite is versatile and most widely used
Chongqing University, Chongqing 400044, China for conventional EDM electrodes. Due to its excellent thermal
944 Int J Adv Manuf Technol (2014) 72:943–953

and electrical properties, graphite is believed to be still the first but also the research methods for the micro manufacturing
choice for electrode materials at meso/micro scale. communities, especially for micro EDM industry, to gain a
Currently, a number of micro manufacturing processes, better understanding of micro machining processes of graphite
such as micro milling, micro laser machining and micro electrodes.
waterjet machining, can be used to produce micro graphite In this paper, micro machinability for micro milling of fine-
electrodes. Micro milling is believed to be the most flexible grained graphite was investigated experimentally. Three types
and versatile micro machining process. It is capable of gener- of micro end mills, namely diamond-coated, TiAlN-coated
ating a wide variety of complex micro components and mi- and uncoated tungsten carbide end mills, were used in the
crostructures due to its strengths in achievable precision and machining experiments. Surface finish and burrs formation
surface finish in addition to its simple set-up and versatility are analysed, and optimized machining parameters were ob-
[2], and it will be a very promising process in fabrication of tained from the experiments with the aim of improving surface
3D micro graphite electrodes. finish and productivity. Complex 3D micro die sink electrodes
Graphite machining has been carried out extensively in the were fabricated using the optimal machining conditions to
industry at conventional scale, but unlike metal cutting, reli- demonstrate the potential of micro milling in graphite elec-
able and widely accepted machining conditions cannot be trode fabrication.
found in the research literatures or machining handbooks.
Only few publications are available in machining graphite
electrodes [4–6]. Micro milling, as a newly emerged micro 2 Moulded graphite and general issues on micro
machining process, has attracted lots of attentions from aca- machining graphite
demia and industries in the last two decades. The research on
micro milling have focused on various applications and in- 2.1 Moulded graphite
volved many engineering materials, including metals such as
aluminum [7–9], copper [9–12], titanium alloys [13, 14] and Moulded graphite is the most commonly used graphite mate-
steels [15, 16]; ceramics like tungsten carbides [17, 18]; rial in the industrial world. It has found applications in many
silicon [19–21]; glass [22, 23]; polymers [24, 25] and com- industrial sectors, such as aerospace, automotive, defense
posites [26]. However, to our knowledge, there are no publi- technology and electronics, to name a few. Typical products
cations available dedicated to micro machinability of graphite include EDM electrodes, carbon brushes, bearings and sliding
material in the open literature. elements, etc. Moulded graphite is a synthetic graphitic prod-
With the demand for micro parts and features increasing, uct manufactured by a compaction process from a mixture of
many cutting tools manufacturers have made commercial carbon filler and organic binder which is subsequently car-
micro tools available in the market. These commercial micro bonized and graphitized [27]. Moulded graphite is a polycrys-
tools are fabricated using various materials, such as high- talline material and the crystallites are randomly oriented. At
speed steel, tungsten carbide, cubic boron nitride (cBN) and macro scale, moulded graphite has isotropic mechanical and
single or poly-crystal diamond, and with various coating such thermal performance, and is soft and brittle. Strength of
as TiN, TiAlN and diamond to improve wear characteristics. moulded graphite is determined by the bond strength between
However, since micro milling is still at its nascent stage, crystallites. At ambient temperature, these bonds resist plastic
currently there are no reliable and widely accepted methods deformation and hence moulded graphite generally fails by
or guidelines available on how to choose cutting conditions in brittle fracture and no yield occurs. On the other hand, the
micro milling. Different tool manufacturers would recom- bond between atoms within the basal plane of a graphite
mend very different cutting conditions for same workpiece crystal is considerably stronger than the bond between the
and tool materials. For example, for micro milling graphite planes with a large anisotropy ratio, which means that
using 1-mm diameter diamond-coated tungsten carbide tools, moulded graphite also has anisotropic behaviour at micro
the recommended axial depth of cut, ap, varies from 0.02 to scale, and its mechanical and frictional properties vary in
0.25 mm, i.e., the highest ap is 12.5 times the lowest values as different crystalline orientations.
recommended by different manufacturers. Further, some rec-
ommended values are not obtainable in practice. For example, 2.2 General issues on micro machining of graphite
in order to achieve a recommended cutting speed for a small
diameter end mill, a spindle speed of 100,000 rpm or more Moulded graphite is generally considered as an easy-to-
could be required, which is not commonly used in industries. machine material in the light of its relatively low hardness
Therefore, it is much needed to conduct research on micro and strength. Very little heat is generated in the cutting zone
machinability of graphite materials to improve the process and hence less thermal distortion is presented on the work-
efficiency and the workpiece quality. The aim of the research piece; therefore, high-speed machining is possible in machin-
is to provide information not only on optimized machining data ing graphite materials. On the other hand, graphite is highly
Int J Adv Manuf Technol (2014) 72:943–953 945

abrasive due to its bond strength between individual carbon roughness and topography which are actually the focus of
molecules, severe tool wear occurs and leads short tool life the research. Spindle errors are believed to have significant
and low productivity, especially at high-speed machining [4, influence on the surface finishes. Since the main spindle of the
5]. Tool wear is also one of the main causes of deterioration of machine tool with higher run-out was mechanical lock
surface finish and accuracy. throughout the experiment, only the run-out and vibration of
There are some issues specifically to micro machining of the precision high-speed spindle would influence the surface
graphite electrodes. Graphite is a brittle material and material finish. This experimental set-up ensures that both the main
properties are very different from those of metals, such as spindle errors, e.g. vibration and run-out, and slideway inac-
hardness, microstructure and friction characteristics, which curacy have been eliminated or minimized. Therefore, main
leads to different machinability and hence a different method spindle and slideway errors on surface finishes will not be
to select tool and cutting parameters. Dimensional accuracy is explicitly taken into account in the analysis.
important for EDM electrodes, as die sink EDM is a copy
machining, and any errors on the electrode will be copied onto 3.2 Workpiece material
the workpiece. Unlike conventional machining, where some
machining errors can be corrected by subsequent machining Moulded graphite can be considered as homogenous and
operations, for micro parts manufacture these subsequent ma- isotropic materials at macro scale. However, this assumption
chining operations are usually not possible or too costly to may not hold true because of the well-known grain size effect.
perform. Therefore, precision fabrication of electrodes directly Commercial moulded graphite has grain sizes ranging from a
by micro machining is the key for the success of micro EDM few microns to tens of microns, and the grain sizes determine
micro parts and features. Surface roughness is another important its material properties. Fine-grained graphite is used in the
factor of graphite electrodes and other precision graphite com- experiment, since it has high strength and hardness, lower
ponents machining because a better surface finish on the elec- porosity and better achievable surface roughness. Table 1 lists
trode can lower electrode wear during EDM. Moulded graphite the workpiece material properties of the fine-grained graphite
is a porous material, so it is difficult to obtain good surface finish used in the experiment.
as metals. Therefore, a study on how the process parameters
affect the surface finish is much desirable to improve surface 3.3 Micro end mills
quality and productivity in micro milling of graphite.
Three types of micro flat end mills, namely diamond-coated
tungsten carbide (DC), TiAlN-coated tungsten carbide
3 Experimental set-up (TiAlN) and uncoated tungsten carbide tools (WC), were used
in the experiments. Three tools have a same nominal diameter
3.1 Machine tool of 0.5 mm. A uniform tool shank diameter of 3 mm is used for
all micro tools to fit 3 mm ultra-precision spindle collet. A
Experiments for this work were carried out on a precision relative large tool diameter was selected in the experiments to
CNC machining centre (Hurco VM10). The use of standard avoid premature tool breakage in those aggressive cuts, and
CNC machine tools is to ensure that the results from the hence ensure that all tools can last without excess tool
experiments are more industrial feasible. The main spindle chipping and wear. Table 2 lists the geometries of micro end
of the CNC machining centre was locked during the experi- mills used in the experiments.
ments, and a high-speed spindle (NSK-HES810) with electric The cutting edge radius was measured on the cross-
drive and ceramic bearings was mounted onto the main spin- sectional SEM images by fitting a circle, as illustrated in
dle. The high-speed spindle offers a continuous power output
of 350 W and an output torque of 3 cNm over the speed range Table 1 Material properties of the fine grained graphite
of 20,000 to 80,000 rpm. The high-speed spindle allows a
Properties Values
reasonable higher cutting velocity even small diameter cutters
are used. Ultra-precision collets were used to clamp the micro Density (kg/m3) 1,830
end mills and the spindle run-out was controlled within 1 μm. Porosity (%) 10
Although the precision machining centre offers single axis Modulus of elasticity (GPa) 13.5
positioning accuracy of 5 μm, as it will be detailed in Compressive strength (MPa) 90
Section 3.4 that only single axis feeding was used in the Hardness (Shore, °S) 60
optimization experiments, i.e. along y-axis. Therefore, the Thermal expansion coefficient (10−6/K) 4.8
slideway positioning errors will only influence dimensional Thermal conductivity (W/m K) 105
accuracy in Y direction which is not of interests in the opti- Average grain size (μm) 5
mization experiments, and will not influence the surface
946 Int J Adv Manuf Technol (2014) 72:943–953

Table 2 Geometries of micro tools used in the experiments changes to reflect this size effects. Feedrate range in the
Diamond TiAlN Uncoated experiments were selected to be same order as cutting edge
coated coated radius in order to investigate if there exists a critical minimum
chip thickness in micro machining of graphite.
Nominal diameter (mm) 0.5 0.5 0.5
No. of flute 2 2 2 3.4 Experimental procedure
Helix angle (deg) 25 20 20
Flute length (mm) 1 1.5 1.5 The experiments in this work include full immersion slot
Rake angle (deg) −5 −2 0 milling. The experimental set-up is illustrated in Fig. 2. For
Clearance angle (deg) 10 20 17 each test, a 5-mm-long and 0.5-mm-wide micro slot was
Nominal corner radius (mm) 0.05 0 0 milled along Y direction. Before experiments, the top surface
Tool edge radius (μm) 9 4 3 of the workpiece was prepared using a 12-mm end-mill. A
small axial depth of cut of 5 μm at each pass was used for the
surface preparation to eliminate surface damage. A machine
Fig. 1. The estimated values are 9, 4 and 3 μm for diamond vision system (InfiniStix™ microscope) is integrated into the
coated, TiAlN coated and uncoated tools, respectively. machine tool providing 3 μm resolution in Z direction to assist
Section 4 will discuss if tool sharpness would affect surface positioning the cutter to the workpiece surface.
finish. In previous research on micro milling of metals, a Dry machining was performed to avoid contamination, and
critical minimum chip thickness was observed. The critical ventilation system was used to control and collect powdery
minimum chip thickness determines the transition between dust. Three controlled quantitative factors used in the experi-
cutting and plowing, i.e. if the uncut chip thickness is less ments are cutting speed (m/min), feedrate (μm/tooth) and
than a critical value, there will be no chip formation, and axial depth of cut (μm). Because there are only three con-
corresponding cutting forces and surface roughness values trolled factors considered for each type of micro tools in the
experiments, full factorial designs were implemented to cap-
a ture all of the main effects and their interactions. Two levels of
cutting speed and depth of cut were selected, albeit four levels
of feedrate were selected. The 4×2×2 mixed level full facto-
rial design matrix is presented in Table 3. Each set of cutting
parameter was repeated once to reduce machining errors and
separate effect due to interactions from measurement noise, so
R= 9µm total of 32 slots were machined for each cutter. In order to
reduce the influence of tool wear on the surface and burr
formation, brand new micro tools were used for the experi-
b ments. Extra slots (exp #17, see Table 3) were milled at the
end of experiments using the same cutting conditions as exp
R= 4µm
#1 (see Table 3). This validation slot was then compared with
the first slot (exp #1).

3.5 Metrology issues

The average surface roughness of the bottom surface of the

micro milled slots was measured using a white light interfer-
c ometer (Zygo NewView 5000) along the centreline of the
slots. To reduce the measurement uncertainty and assess re-
peatability, four measurements on different area were conduct-
R= 3µm ed for each slot and an average value of surface roughness
(Ra) is used for analysis. An optical microscope and surface
profiler (Talysurf Serial-2) were used to assess burr
formation. Besides, the surface morphology was
analysed on the basis of SEM micrographs. Analysis on the
response–surface roughness will be presented in Section 4.
Fig. 1 Cutting edge radius measurement by SEM: a diamond-coated Mean values of measured surface roughness, Ra, are listed in
tool; b TiAlN-coated tool; c uncoated tool Table 3.
Int J Adv Manuf Technol (2014) 72:943–953 947

a b
Z n = 20k-80k rpm

X High speed spindle

Y

Micro endmill

Feedrate, f

Fixture and workpiece

Axial depth of cut, ap
5 mm
ae = 0.5mm

Fig. 2 Experimental set-up for micro milling of graphite a schematic diagram; b machining set-up photo

4 Results and discussions and 6, which corresponds to a confidence level of 95 % in

evaluation of process parameters.
ANOVA was performed on surface roughness data to isolate The ANOVA tables show that feedrate has the most signif-
the main effects of each source of variations and determine icant influence on surface roughness for all three types of
effects of interactions. Tables 4, 5 and 6 summarize the results micro end mills, and it contributes to 85.2, 80.6 and 70.7 %
of ANOVA on surface roughness (Ra values) for three types overall responses for diamond-coated, TiAlN-coated and un-
of micro end mills. All tabulated F values (α=0.05) except the coated micro end mills, respectively. Cutting speed also has
interaction of cutting speed and depth of cut in Tables 5 and 6 influence on surface roughness for TiAlN-coated and uncoat-
are smaller than the calculated values as shown in Tables 4, 5 ed tools, whereas cutting speed has almost no influence on

Table 3 Experimental matrix

of micro milling of graphite Exp no. Experimental matrix—controlled factorsa Responses
and mean measured surface
roughness, Ra Feedrate Cutting speed/spindle Depth of cut Measured surface roughness, Rab
speed (μm)

fz (μm/tooth) Vc (m/min) n (rpm) ap (μm) Diamond TiAlN Uncoated

coated coated

1 0.5 31.4 20,000 20 0.29 0.35 0.47

2 2 31.4 20,000 20 0.4 0.48 0.75
3 5 31.4 20,000 20 0.54 0.65 0.78
4 10 31.4 20,000 20 1 1.2 1.13
5 0.5 78.5 50,000 20 0.29 0.49 0.79
6 2 78.5 50,000 20 0.52 0.61 0.9
7 5 78.5 50,000 20 0.76 0.95 0.84
8 10 78.5 50,000 20 0.87 1.45 1.31
9 0.5 31.4 20,000 60 0.3 0.4 0.47
a 10 2 31.4 20,000 60 0.45 0.54 0.61
Uncontrolled factors: width of
cut, ae =0.5 mm; cutter diameter, 11 5 31.4 20,000 60 0.66 0.62 0.69
D=0.5 mm; no of flutes, Z=2;
12 10 31.4 20,000 60 0.78 1.04 1.21
slot length=5 mm
b 13 0.5 78.5 50,000 60 0.39 0.42 0.6
Surface roughness values
presented in the table are the 14 2 78.5 50,000 60 0.48 0.45 0.7
arithmetic mean of measured 15 5 78.5 50,000 60 0.75 0.97 1.08
Ra values 16 10 78.5 50,000 60 0.82 1.05 1.09
c
Validation slot, same machining 17c 0.5 31.4 20,000 20 0.30 0.33 0.55
condition as exp #1
948 Int J Adv Manuf Technol (2014) 72:943–953

Table 4 ANOVA for micro milling of graphite experiment using diamond-coated tools

Source of variation Sum of squares Degrees of Mean square F (calculated) F (tabulated, α=0.05) P value Contribution
freedom ratio %

Feedrate (A) 2.753 3 0.918 235.765 2.80 1.02E−28 85.232

Cutting speed (B) 0.0225 1 0.0225 5.78 4.04 0.020118 0.697
Depth of Cut (C) 5.63E−05 1 5.63E−05 0.0145 4.04 0.90482 0.002
AB 0.101 3 0.0337 8.656 2.80 0.000107 3.127
AC 0.078 3 0.026 6.673 2.80 0.000741 2.415
BC 0.004 1 0.004 0.13 4.04 0.71996 0.016
ABC 0.0873 3 0.029 7.474 2.80 0.000334 2.703
Error 0.187 48 3.89E−03 235.765 5.789
Total 3.23 63

surface roughness when diamond-coated tools were used. choice for micro milling of graphite if optimization of surface
Depth of cut has positive influence on surface roughness when quality is the main concern. Within the cutting parameter range
TiAlN-coated and uncoated tools were used, i.e. an increase in used in the experiments, minimum average surface roughness
axial depth of cut will improve surface roughness. value of around 0.3 μm was measured at low feedrate of
Both the interaction between feedrate and cutting speed 0.5 μm/tooth and under both low and high depth of cut when
and the interaction between feedrate and depth of cut have using diamond-coated tools. Maximum surface roughness
effects on the measured surface roughness, although these mean value of around 1.4 μm was measured at high feedrate
interaction effects are not significant as single factor main of 10 μm/tooth under high depth of cut of 60 μm when using
effects. Higher order interactions were not significant; there- either TiAlN-coated or uncoated tools.
fore, it does not seem worthwhile to examine the statistical It is observed from Figs. 3a, 4a and 5a that an increase in
significance of these effects. feedrate from 0.5 to 10 μm/tooth for all three micro tools
Main effects were plotted in Figs. 3, 4 and 5 to quantitatively results in increase of surface roughness values and this ten-
assess the effect of each process parameters on micro machin- dency is very linear in the machining conditions used in the
ing process. Tool sharpness in terms of cutting edge radius is experiments. This tendency is also similar to that for micro
believed to have a positive influence on surface roughness at milling of metallic materials.
the order of a few microns, i.e. smaller tool edge radius is When cutting speed increases from 31.4 to 78.5 m/min,
expected to result in better surface finish. However, it can be there is around 20 % increase in surface roughness value for
seen from Table 3 and Figs. 3, 4 and 5 that under same cutting TiAlN-coated and uncoated tools, whilst only 5 % increase in
conditions diamond-coated tools give better surface roughness, surface roughness value was observed for diamond tools, as
although the diamond-coated tools are the bluntest among the shown in Figs. 3b, 4b and 5b. This suggests that higher cutting
three types of micro tools. There is no clear difference between speed hence higher machining efficiency can be achieved by
TiAlN-coated tools and uncoated tools in terms of achievable using diamond-coated tools without significantly compromis-
surface quality, indicating TiAlN-coated tools are not a good ing surface quality.

Table 5 ANOVA for micro milling of graphite experiment using TiAlN-coated tools

Source of variation Sum of squares Degrees Mean square F (calculated) F (tabulated, α=0.05) P value Contribution
of freedom ratio %

Feedrate (A) 5.712 3 1.904 211.731 2.80 1.13E−27 80.6

Cutting speed (B) 0.304 1 0.304 33.793 4.04 4.83E−07 4.3
Depth of cut (C) 0.113 1 0.113 12.574 4.04 0.000885 1.6
AB 0.203 3 0.068 7.528 2.80 0.000316 2.9
AC 0.202 3 0.067 7.471 2.80 0.000334 2.8
BC 0.068 1 0.068 7.590 4.04 0.008268 1
ABC 0.057 3 0.038 2.946 2.80 0.13471 0.7
Error 0.432 48 0.009
Total 7.086 63
Int J Adv Manuf Technol (2014) 72:943–953 949

Table 6 ANOVA for micro milling of graphite experiment using uncoated coated tools

Source of variation Sum of squares Degrees Mean square F (calculated) F (tabulated, α=0.05) P value Contribution
of freedom ratio %

Feedrate (A) 3.133 3 1.044 120.887 2.80 2.24E−22 70.7

Cutting speed (B) 0.353 1 0.353 40.845 4.04 6.32E−08 8.0
Depth of cut (C) 0.064 1 0.064 7.351 4.04 0.009271 1.4
AB 0.105 3 0.035 4.050 2.80 0.012052 2.4
AC 0.131 3 0.044 5.056 2.80 0.004006 3.0
BC 0.011 1 0.011 1.276 4.04 0.26421 0.2
ABC 0.218 3 0.073 8.396 2.80 0.000137 4.9
Error 0.415 48 0.009 9.4
Total 4.428 63

The surface roughness decreases by 10 % when depth of In micro milling of metals, the cutting edge radius and
cut was increased from 20 to 60 μm when TiAlN-coated and minimum chip thickness size effect is the dominant factor
uncoated tools were used as shown in Figs. 4c and 5c. This for material removal mechanism and chip generation physics.
trend is different from conventional machining of graphite In previous micro milling of metallic materials, numerous
materials and can be explained by grain size effect. The researchers have investigated that this size effect and the
average graphite grain size in the workpiece was measured critical minimum chip thickness was reported to be around
by the manufacturer as 5 μm. When the grain size is close to or 0.1–0.4 of cutting edge radius for various materials [10, 15,
smaller than depth of cut, brittle fracture may occur within the 30, 31]. Micro tools used in this work have a cutting edge
grain, so that the material cannot be considered as homoge- radius in the range of 3–9 μm from SEM measurement results
nous and isotropic. The grain size effect will affect the micro and the ratio of uncut chip thickness to cutting edge radius is
cutting processes, including mechanism of chip formation and in the range of 0.05 to 1; however, no explicit cutting edge
surface generation [28, 29]. Local material strength within the radius size effects were observed in the experiments for all
grain is higher than the bulk value, more specific cutting three types of cutters. Further examinations on surface defects
energy is required to remove material, and therefore the sur- show that micro cracking occurred in the bottom and side
face finish is deteriorated. When the depth of cut increases, the surface of the slots. Figure 6 shows typical surface damages in
material becomes more statistically homogenous and the ma- the form of fractured craters and residual flake chips. Different
terial property tends to be similar to that of bulk materials, and to micro milling of metals, no tools marks were observed on
therefore higher feedrate can promote good surface finish machined surfaces from TiAlN-coated and uncoated tools,
within a certain range. However, this effect has not been whilst indistinct tool marks were found by close examination
clearly observed when diamond-coated tools were used on surfaces machined by diamond-coated tools. Surface dam-
(Fig. 3c). ages and absence of tool marks on the machined surfaces

a b c
1 1 1
Mean Surface Roughness, Ra (micron)

Mean Sur face Roughness, Ra (micron)

Mean Sur face Roughness, Ra (micron)

0.9 0.9 0.9

0.8 0.8 0.8

0.7 0.7 0.7

0.6 0.6 0.6

0.5 0.5 0.5

0.4 0.4 0.4

0.3 0.3 0.3

0.2 0.2 0.2

0 2 4 6 8 10 30 40 50 60 70 80 20 30 40 50 60
Feedrate (micron/tooth) Cutting Speed (m/min) Depth of Cut,ap (micron)

Fig. 3 Main effect plots for influences of single factors on surface roughness with diamond-coated micro end mills: a feedrate, fz; b cutting speed, Vc; c
axial depth of cut, ap
950 Int J Adv Manuf Technol (2014) 72:943–953

a b c
1.4 1.4 1.4
Mean Surface Roughness, Ra (micron)

Mean Surface Roughness, Ra (micron)

Mean Surface Roughness, Ra (micron)
1.2 1.2 1.2

1 1 1

0.8 0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0 2 4 6 8 10 30 40 50 60 70 80 20 30 40 50 60
Feedrate (micron/tooth) Cutting Speed (m/min) Depth of Cut,ap (micron)

Fig. 4 Main effect plots for influences of single factors on surface roughness with TiAlN-coated micro end mills: a feedrate, fz; b cutting speed, Vc; c
axial depth of cut, ap

indicate that the primary cutting mode was still brittle fracture image of three 3D intricate micro machined electrodes. Fine-
or partial ductile under the cutting conditions in the experi- grained tungsten carbide end mills with diameter of 0.2 mm
ments. Another evidence of graphite materials being cut pri- were used to machine these samples. An average surface
marily at brittle fracture mode is that no obvious burrs were roughness of 0.8 μm was measured on the bottom
observed for all the slots machined, which is similar to con- surfaces, and slot width deviations were controlled well
ventional machining of most brittle materials. within 10 μm based on surface profiler measurements.
For three types of micro tools, it is observed that there is no Good dimensional accuracy and surface roughness obtained
significant difference between first set of slots (exp #1) and from the experiments demonstrated that micro milling is a
validation slots (exp #17) in terms of surface roughness and promising process for fabrication of 3D micro EDM
micro surface morphology. Therefore, the tool wear effect is electrodes.
regarded negligible under this certain cutting distance in this
work.
6 Conclusions

5 Examples of micro die sink EDM electrodes This paper presents an experimental investigation on micro
milling of moulded fine-grained graphite. Diamond-coated,
Micro EDM electrodes with high aspect ratio and complex TiAlN-coated and uncoated tungsten carbide micro end mills
shapes were fabricated using the optimal cutting conditions with 0.5 mm diameter were used to create full width slots.
obtained through this research. Figure 7 shows the SEM Various cutting conditions including feedrate, cutting speed

a b c
1.4 1.4 1.4
Mean Surface Roughness, Ra (micron)

Mean Surface Roughness, Ra (micron)

Mean Surface Roughness, Ra (micron)

1.2 1.2 1.2

1 1 1

0.8 0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0 2 4 6 8 10 30 40 50 60 70 80 20 30 40 50 60
Feedrate (micron/tooth) Cutting Speed (m/min) Depth of Cut,ap (micron)

Fig. 5 Main effect plots for influences of single factors on surface roughness with uncoated micro end mills: a feedrate, fz; b cutting speed, Vc; c axial
depth of cut, ap
Int J Adv Manuf Technol (2014) 72:943–953 951

ƒFig. 6 SEM micrographs of the micro machined surface morphology

(cutting conditions: Vc=31.4 m/min; fz=2 μm/tooth, axial depth of cut=
20 μm) by a diamond-coated tool; b TiAlN-coated tool; c uncoated tool

and axial depth of cut were experimentally investigated based

on design of experiments techniques, and their influences on
surface finish were assessed using ANOVA method. The
following conclusions can be drawn from this work:

& Feedrate has the most significant influence on surface

roughness for all three types of micro tools. Higher
feedrate results in higher surface roughness values and

500µm

b
500µm

c
1mm

Fig. 7 Examples of micro machined EDM electrodes a micro thin-

walled structure; b micro pillar array; c micro mixing device
952 Int J Adv Manuf Technol (2014) 72:943–953

this linear tendency is similar to that of micro milling of 5. Zhou L, Wang CY, Qin Z (2009) Tool wear characteristics in
high-speed milling of graphite using a coated carbide micro
metallic materials.
endmill. Proc Inst Mech Eng B J Eng Manuf 223:267–277
& The diamond-coated tools give better surface roughness 6. Yang YK, Chuang MT, Lin SS (2009) Optimization of dry
although the diamond-coated tools have much larger tool machining parameters for high-purity graphite in end milling
edge radius than other two types of micro tools used in the process via design of experiments methods. J Mater Process Technol
209:4395–4400
experiments. High-speed machining with diamond-coated
7. Newby G, Venkatachalam S, Liang SY (2007) Empirical analysis of
is more effective in improving machining productivity and cutting force constants in micro-end-milling operations. J Mater
achieving good surface finish. Process Technol 192–193:41–47
& The surface roughness decreases when depth of cut is 8. Ng CK, Melkote SN, Rahman M, Senthil KA (2006) Experimental
study of micro- and nano-scale cutting of aluminum 7075-T6. Int J
increased when TiAlN-coated and uncoated tools were
Mach Tools Manuf 46:29–936
used, and this phenomenon can be explained by grain size 9. Lee K, Dornfeld DA (2002) An experimental study on burr formation
effect. Higher depth of cut is therefore recommended in micro milling aluminum and copper. Trans N Am Manuf Res Inst
under certain machining conditions from optimizing ma- 30:255–262
10. Filiz S, Conley CM, Wasserman MB, Ozdoganlar OB (2007) An
chining productivity and surface finish point of view.
experimental investigation of micro-machinability of copper 101
Different to micro milling of ductile metallic materials, using tungsten carbide micro-endmills. Int J Mach Tools Manuf 47:
no cutting edge size effect and minimal chip thickness 1088–1100
were observed for all three types of micro tools. 11. Huo D, Cheng K (2010) Experimental investigation on
micromilling of oxygen-free, high-conductivity copper using
& No obvious burrs and clear tool marks were observed for
tungsten carbide, chemistry vapour deposition, and single-
all machined slots, whilst surface damages were found in crystal diamond micro tools. Proc Inst Mech Eng B J Eng Manuf
the forms of fractured craters, residual flake chips and 224:995–1003
micro cracks, which indicate that the primary cutting 12. Rahman M, Kumar SA, Prakash JRS (2001) Micro milling of pure
mode is still brittle fracture or partial ductile under the copper. J Mater Process Technol 116:39–43
13. Ozel T, Thepsonthi T, Ulutan D, Kaftanoglu B (2011) Experiments
cutting conditions in the experiments. and finite element simulations on micro-milling of Ti–6Al–4 V. Ann
CIRP 60:85–88
Micro EDM electrodes with high aspect ratio and complex 14. Schueler GM, Engmann J, Marx T, Haberland R, Aurich JC
shapes were fabricated using the optimal cutting conditions (2010) Burr formation and surface characteristics in micro-
endmilling of titanium alloys. In: Aurich JC, Dornfeld D
obtained through this research. The results are also useful for (eds) Burrs—analysis, control and removal. Springer, Berlin,
micro milling of other brittle materials, such as glass and pp 129–138
silicon. Although tool wear was found negligible in the ex- 15. Weule H, Hüntrup V, Tritschler H (2001) Micro-cutting of
perimental conditions in this research, future efforts will be steel to meet new requirements in miniaturization. Ann CIRP
50:61–64
made on characterization of tool wear in micro milling of 16. Bissacco G, Hansen HN, Chiffre LD (2005) Micromilling of hard-
graphite. ened tool steel for mould making applications. J Mater Process
Technol 167:201–207
17. Nakamoto K, Katahira K, Ohmori H, Yamazaki K, Aoyama T
Acknowledgments The authors are grateful for the support of
(2012) A study on the quality of micro-machined surfaces on
Newcastle University and Visiting Scholar Foundation of the State Key
tungsten carbide generated by PCD micro end-milling. Ann CIRP 61:
Lab. of Mechanical Transmission in Chongqing University. The authors
567–570
are also grateful to Mr Jamie Hodgson and Ben Walker at Newcastle
18. Fanuc. Website: [Link]
University for their assistance in the work.
htm. Accessed on 25 August 2013
19. Rusnaldy, Jo Ko T, Sool Kim H (2007) Micro-end-milling of single-
crystal silicon. International Journal of Machine Tools &
References Manufacture 47:2111–2119
20. Rusnaldy, Jo Ko T, Sool Kim H (2008) An experimental
study on microcutting of silicon using a micromilling machine.
1. Ehmann KF, Bourell D, Culpepper ML, Hodgson TJ, Kurfess TR, International Journal of Advanced Manufacturing Technology 39:
Madou M, Rajurkar K, De Vor RE (2005) International assessment of 85–91
research and development in micromanufacturing. World 21. Arif M, Rahman M, San WK (2012) An experimental inves-
Technology Evaluation Center, Baltimore tigation into micro ball end-milling of silicon. J Manuf Process 14:
2. Huo D, Cheng K, Wardle F (2010) Design of a five-axis ultra- 52–61
precision micro-milling machine-UltraMill. Part 1: holistic design 22. Arif M, Rahman M, San WY, Doshi N (2011) An experimental
approach, design considerations and specifications. Int J Adv approach to study the capability of end-milling for microcutting of
Manuf Technol 47:867–877 glass. Int J Adv Manuf Technol 53:1063–1073
3. Maradia U, Boccadoro M, Stirnimann J, Beltrami I, Kuster F, 23. Foy K, Wei Z, Matsumura T, Huang Y (2009) Effect of tilt angle on
Wegener K (2006) Die-sink EDM in meso-micro machining. cutting regime transition in glass micromilling. Int J Mach Tools
Procedia CIRP 1:166–171 Manuf 49:315–324
4. Schroeter RB, Kratochvil R, Gomes JO (2006) High-speed finishing 24. Friedrich CR, Vasile MJ (1996) Development of the micromilling
milling of industrial graphite electrodes. J Mater Process Technol process for high-aspect-ratio microstructures. J Microelectromech
179:128–132 Syst 5:33–38
Int J Adv Manuf Technol (2014) 72:943–953 953

25. Ogilvie IRG, Sieben VJ, Floquet CFA, Zmijan R, Mowlem MC, 28. Cheung C, To S, Lee W (2002) Anisotropy of surface roughness in
Morgan H (2010) Reduction of surface roughness for optical quality diamond turning of brittle single crystals. Mater Manuf Process 17:
microfluidic devices in PMMA and COC. J Micromech Microeng 251–267
20:1–8 29. Min S, Dornfeld D, Inasaki I, Ohmori H, Lee D, Deichmueller M,
26. Cho MW, Kim DW, Cho WS (2007) Analysis of micro-machining Yasuda T, Niwa K (2006) Variation in machinability of single crystal
characteristics of Si3N4-hBN composites. J Eur Ceram Soc 27:1259– materials in micromachining. Ann CIRP 55:103–106
1265 30. Ikawa N, Shimada S, Tanaka H (1992) Minimum thickness of cut in
27. Pierson HO (1993) Handbook of carbon, graphite, diamonds and micro machining. Nanotechnology 3:6–9
fullerenes: processing, properties and applications. Noyes 31. Cheng K, Huo D (2013) Micro cutting: fundamentals and applica-
Publications, New Jersey tions. Wiley, Chichester