Journal of Materials Processing Technology 124 (2002) 297±304

Effect of SiC and rotation of electrode on electric discharge

machining of Al±SiC composite
B. Mohana,*, A. Rajaduraib, K.G. Satyanarayanac
a
Department of Production Technology, MIT Campus, Anna University, Chennai 600 044, India
b
Division of Manufacturing Engineering, CEG Campus, Anna University, Chennai 600 025, India
c
Regional Research Laboratory, Trivandrum, Kerala, India
Accepted 10 April 2002

Abstract

Metal matrix composites (MMCs) contain certain amount of secondary reinforcement, which makes it dif®cult to machine with
conventional cutting tools and methods. The present study analyses the effect of electric discharge machining (EDM) parameters namely
polarity, current, electrode material, pulse duration, and rotation of electrode on metal removal rate (MRR), tool wear rate (TWR) and surface
roughness (SR) value in EDM of Al±SiC MMCs with 20 and 25 vol.% SiC. Irrespective of the electrode material, polarity of the electrode and
volume percentage of SiC, the MRR increased with increase in discharge current and for a speci®c current it decreased with increase in pulse
duration. Increase in the volume percentage of SiC had an inverse effect on MRR, and positive effect on TWR and surface ®nish. Increasing
the speed of the rotating electrode resulted in a positive effect with MRR, TWR and better SR than at stationary. Optimum parameters for
EDM drilling were also developed to summarize the effect of machining characteristics such as MRR, TWR and SR.
# 2002 Elsevier Science B.V. All rights reserved.

Keywords: Al±SiC; Metal matrix composite; Electric discharge machining; Rotation of electrode

1. Introduction duction of ceramic particles, they show high hardness and

Since the 1970s, the MMCs have been successfully applied
in aerospace industries [1,2]. In the middle of the 1980s these
materials reached the automobile industry and nowadays its
use, though not very wide, is gaining importance. Initially,
structural aluminium matrix sheet, reinforced with large
®bres was developed and in the last few years, research
has been specially focused on the aluminium matrix com-
posites with discontinuous reinforcementsÐshort ®bres,
particles and whiskers of SiC and aluminium oxide [2].
Most of the parts obtained with aluminium matrix com-
posites through different manufacturing processes namely
the processes derived from the casting and bulk plastic
deformation (forging and extrusion) have different geometry
and they usually need machining operations with the required
dimensional and geometric precision as well as good surface
roughness (SR). Although the mechanical properties of the
composites have been signi®cantly improved, with the intro-
Abbreviations: MRR, material removal rate (mm3/min); TWR, tool
wear rate (mm3/min); SR, surface roughness Ra (mm)
*
Corresponding author.
E-mail address: mohan@mitindia.edu (B. Mohan).
leads to severe abrasive wear on the conventional tools [3,4].
However, advanced machining like water jet machining and
wire cut electric discharge machining (EDM) can be applied
in linear cutting. The intrinsic designation of the above
techniques limit their application to blind hole drilling
operations. Thus, EDM becomes an optimal choice in the
machining of a blind hole in the workpiece owing to its easy
control in operation [6]. A modi®ed electrode is presented in
this work to accelerate the debris removal during machining.
The issue of machining MMCs has been addressed from
various aspects. Ramulu et al. [7,8] used modi®ed EDM to
cut SiC±TiB2 and SiCw/Al with plate electrode. Soni et al.
[9±11] used rotary EDM to machine titanium alloy and die
steel with a solid electrode. Amongst these studies, little
research was found on using rotary EDM with brass and
copper electrode. The conventional EDM techniques lack
axial rotation during machining. However, the modern
technical fabrication system has compensated for the limita-
tion of EDM, the revised facility of EDM imparts a rotation
to the electrode in its vertical axis.
Experimental results con®rm that EDM of blind hole
drilling with a rotary electrode has a higher MRR, less
electrode wear and better surface ®nish than solid stationary

0924-0136/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 4 - 0 1 3 6 ( 0 2 ) 0 0 2 0 2 - 9
298 B. Mohan et al. / Journal of Materials Processing Technology 124 (2002) 297±304

electrode. Hence, this technique is a viable one to drill
MMC.
2. Objectives of the work
The objectives of the present study on EDM of Al±SiC
MMCs are as follows:
(1) To investigate the effect of polarity of the electrode,
discharge current, electrode material, volume percen-
tage of SiC particle, and pulse duration on material
removal rate (MRR), tool wear rate (TWR) and surface
roughness (SR).
(2) To study the effect of rotation of electrode on MRR,
TWR and SR.
(3) To optimize the cutting condition for maximizing the
MRR.
pressure of 3.5 kg/cm2 by lateral ¯ushing. MRR is propor-
tional to the product of the energy transferred per pulse and
the pulse frequency. Changing the pulse current at a constant
frequency varies the energy of the pulse. Hence all the
experiments were performed with pulse current, pulse dura-
tion and volume percentage of SiC as variables. The pulse
currents selected for this study were 2, 5, 8 and 11 A. The
selected pulse durations were 88, 176, and 264 ms.
Blind hole drilling operations were carried out for a depth
of 10 mm on aluminium alloy (LM25) reinforced with 20
and 25 vol.% of SiC particle. The composition of the alloy is
given in Table 1. The MRR and TWR were calculated by
measuring the average amount of material removal. The
MRR and TWR were measured by using an electronic
balance of sensitivity of 0.1 mg. The SR of the machined
surfaces were measured with the help of Surtronic 3‡. The
Ra values are used to quantify the SR. The cutoff length for
each measurement was 0.8 mm.

3.1. Rotating tool assembly

3. Experimental details
One of the primary objectives was to study the effect of
The experiments were carried out using an Electronica-
rotation of tool. Therefore, a mechanism to rotate the tool
M-T-3822 machine, which was equipped with transistor
was developed. The electrode was rotated and sunk simul-
switched power supply. The electrode was fed downwards
taneously to machine a hole in the workpiece. An electric
under servo control into the workpiece. Fig. 1 depicts the
motor was used to rotate the electrode (tool). A V belt was
experimental setup. Copper and brass cylindrical electrodes
used to transmit the power from the motor to the electrode.
of 12 mm diameter were used as tool. Kerosene was used as
The speed of the rotating electrode was controlled with the
a dielectric ¯uid. The dielectric ¯uid was circulated with a
help of a regulated power supply. A mechanical tachometer
was used to measure the speed of the rotating electrode.
Experiments were conducted to identify the variables that
are likely to have signi®cant in¯uence on the responses. In
this study, MRR, TWR and SR were considered as
responses. These values for the stationary and rotating
electrode were compared.

4. Results and discussion

The effect of EDM variables namely polarity, current,

electrode material, volume percentage of SiC, pulse duration
and rotation of electrode on MRR, TWR and SR of Al±20%
SiC and Al±25% SiC composites were studied and the
results are as follows.

4.1. Material removal rate

Fig. 2a and b represents the variation of MRR with

Fig. 1. Experimental setup. discharge current for copper and brass electrodes while

Table 1
Composition of aluminium alloy

Si (%) Fe (%) Cu (%) Mn (%) Mg (%) Zn (%) Ti (%) Al

6.25±7.5 0.2 0.2 0.1 0.2±0.4 0.1 0.2 Balance

 B. Mohan et al. / Journal of Materials Processing Technology 124 (2002) 297±304 299

Fig. 2. Variation of MRR with discharge current: (a) Al±20% SiC; (b) Al±25% SiC.

machining Al±(SiC)p. Fig. 2a depicts the MRR with dis- machining gap consists of products of dielectric decomposi-
charge current for 20 vol.% SiC particulate. According to tion and eroded metallic particles. The conducting particles
this ®gure the MRR was more when the electrodes were remain suspended in the gap and exert signi®cant in¯uence
at positive polarity than at negative. Yan et al. [12] and on the dielectric breakdown characteristics. The ignition
Wang and Yan [13] observed a similar phenomenon on time delay as a function of the particle concentration and the
conforming that using positive electrode polarity in EDM relationship was given by
caused a higher MRR under higher discharge energy. This  
Ncr
phenomenon might be attributable with the transfer of Td ˆ C3 ln (1)
energy during the discharging process [13]. It was inferred Na
from the above ®gures that the MRR with brass electrode where C3 is a constant, Ncr the critical particle concentration
was more than the copper electrode with increasing pulse and Na the average particle concentration.
current irrespective of the electrode polarity and volume Effective spark discharges are characterized by a ®nite
percentage of SiC. This might be due to the high speci®c time delay before the pulse current reaches the nominal
resistance with the brass electrodes than the copper electrode value. When the Ncr is equal to Na, then the plasma channel
which increases the spark intensity and the electrode preferentially takes the same path as that of the previous
gap [13].
Fig. 3 shows the effect of pulse duration on MRR on
different electrodes. The MRR decreased with the increase
in the pulse duration. A short pulse duration caused less
surface vapourization, whereas a long pulse duration may
cause the plasma channel to expand and to decrease the
energy density for the workpiece. In EDM, the dielectric
¯uid acts as an insulator. The electrode (tool) and the
workpiece convect away a small amount of heat generated
by the discharges and ¯ushes off the discharge byproducts
from the electrode gap. As machining proceeds the con-
centration of the particle in the gap increased rapidly [13].
The effect of rotation of electrode in comparison with the
stationary electrode is shown in Fig. 4. For any given current
the MRR was more for the rotary electrode than stationary.
This increase in MRR was due to the effective ¯ushing of the
rotary electrode. The rotation of the electrode imparts a
whirl and effectively ¯ush the gap (resulting in increased
MRR) and the machined surface was better than that
obtained with a stationary electrode. The debris in the Fig. 3. MRR vs. pulse duration.
300 B. Mohan et al. / Journal of Materials Processing Technology 124 (2002) 297±304
Fig. 4. Effect of electrode speed on MRR.
pulse, which was not completely localized. This leads to
process instability due to inadequate ¯ushing in the case of
stationary electrodes. When the cylindrical electrode rotates,
due to the centrifugal action, a new layer of dielectric ¯uid
was thrown into the machining gap. This induces a con-
ducive atmosphere for effective discharge and encourages
process stability. The enhanced discharge increased the
MRR and ef®ciency [5].
The rate of debris formation is increased at higher
pulse current, whereas in the case of rotary electrode, a
Fig. 5. TWR vs. discharge current: (a) Al±20% SiC; (b) Al±25% SiC.
small whirl imparted to the electrode brings about a sig-
ni®cant increase in MRR. Insuf®cient ejection force causes
part of the molten metal to remain as a recast layer. The
recast layer reduces the MRR in the case of stationary
electrode [5].
4.2. TWR
The TWR of electrode was obtained from the weight
difference of electrode before and after the performance
trial:
Wi Wf
TWR ˆ (2)
…density of electrode  t†  1000
Wi is the initial weight of electrode in g, Wf the ®nal weight
of electrode in g and t the period of trial in min.
Fig. 5a and b shows the curves of TWR vs. discharge
current for different polarities of the copper and brass
electrode. The TWR on the negatively connected electrode
was less than the positively connected electrode irrespective
of the electrode material and the volume percentage of SiC.
This is due to the increase in MRR with the positively
connected electrode. It was also observed from the above
®gures that as the current increased, the TWR also increased
irrespective of the electrode material and polarity. A high
current generates an increased discharging energy and
causes a high electrode wear. The increase in MRR increases
the debris at the gap, which increases the wear on the
electrode [5,13].
 B. Mohan et al. / Journal of Materials Processing Technology 124 (2002) 297±304
Fig. 6. Progress of TWR with pulse duration.
On machining Al±SiC with different volume percentages
of SiC the TWR was more as the volume percentage of SiC
is increased. This increase in wear was due to the SiC
particle's abrasive wear nature and decreased conductivity
of the material. The TWR was more in brass electrode than
in the copper electrode for a particular current and pulse
duration since the MRR was more with brass electrode.
Fig. 6 shows the relationship between TWR with pulse
duration. The TWR decreases in an inverse relation with
pulse duration as similarly observed by Wang and Yan [13].
This phenomenon is attributable to the brass electrode which
has a good thermal conductivity. Thus heat generated during
the machining was easily removed. The heat removal facil-
itates a reduction of the temperature around the surface of
the electrode for a long pulse duration, which reduces the
TWR.
Fig. 7. Effect of electrode speed on TWR.
301
The TWR on machining different volume percentages of
SiC with brass electrode in rotating condition is shown in
Fig. 7. The TWR on machining with rotary electrode was less
in comparison with stationary electrode. In a conventional
electrode the loosened SiC deposition occurred in a localized
area, which inhibited high electrode wear. Frequent arcing
during static EDM was also found to add carbide deposits on
the electrode surface. In a rotary electrode, the carbide
deposition was spread over a larger area on the circumfer-
ence. The rotation of the electrode also contributed better
heat transfer from the electrode thus bringing down the
electrode surface temperature. It was also probable that
the layer breaks down during the course of rotation before
re-depositing in the successive cycles. Also the electrode
wear was uniformly distributed over a larger area on the
circumference of the electrode, whereas it was con®ned to a
local area in a conventional stationary electrode [5].
4.3. Surface roughness
SR was an important process response, which dictates the
conditions with which the component has to be machined. If
surface ®nish was the criterion then the material must be
machined with low MRR. When the electrodes are con-
nected at positive polarity, the craters on the workpiece
surface largely have an irregular pro®le, whereas for a
negative polarity, the craters are ¯at.
The variation in SR with different pulse current levels and
for different polarities of the electrodes were shown in
Fig. 8a and b. It was also observed from the above ®gures
that the increase in discharge current resulted in an increase
in Ra value irrespective of the electrode and volume percen-
tage of SiC. This event is due to the increase in discharge
energy which subsequently causes a larger crater on the
surface of the body. When machining with copper electrode
the SR values was less than with brass electrode. The
increase in volume percentage of SiC effected a positive
effect on the roughness value which was possibly due to the
voids left on the surfaces by the SiC particles debonding
[5,13].
Fig. 9 presents the SR vs. pulse duration under peak
current. Increase in pulse duration resulted in less SR.
The long pulse duration in the machining process expands
the plasma channel, decreases the energy density and there-
fore induces a shallow crater on the surface of the workpiece
[13].
The variation in surface ®nish with discharge current for
various electrode rotational speeds is given in Fig. 10. The
roughness value decreases with increasing speed of the
electrode at constant pulse current. The phenomenon of
arcing, which occurs frequently with a stationary electrode
impairs the work surface. The energy contained in a pro®le
pulse is given by
Z tp
Eˆ I…t†V…t† dt (3)
td
302 B. Mohan et al. / Journal of Materials Processing Technology 124 (2002) 297±304

Fig. 8. Variation of SR with discharge current: (a) Al±20% SiC; (b) Al±25% SiC.

where td is the ignition time delay, tp the pulse on time, I(t)
the gap current, and V(t) the working voltage. With increased
peripheral speed of the electrode, the ignition time delay
increases, thus bringing down the energy transferred through
the individual discharges for material removal. This
diminishes the crater dimensions to give a better roughness
value. Under conditions of effective ¯ushing, the adherence
of resolidi®ed eroded particles on the work surface is
reduced and the resulting surface presents a better ®nish
than that of the one obtained with a stationary electrode. In
the later case, an increase in MRR is possible only at the
expense of surface ®nish. On the contrary, in a rotatory
electrode the increase in MRR is accompanied by the
production of a better surface.
5. Mathematical modelling and optimization
The mathematical models relating to the responses with
the process variables are to enable optimization of machin-
ing characteristics of MMC in EDM.
The mathematical models commonly used was
Y ˆ f …I; T†, where y was the response, I, T are the process
parameters such as current and pulse duration.
From the experimental observation, a basic power model
was assumed as follows:
Y ˆ XI a T b (4)
where X is a constant and a and b are exponents of I and T,
respectively. From the experimental results and by using

Fig. 9. SR vs. pulse duration. Fig. 10. Effect of electrode speed on SR.
 B. Mohan et al. / Journal of Materials Processing Technology 124 (2002) 297±304 303

Table 2 algorithm programme by treating the problem as a linear

Constants of regression equation programming problem. The objective of the problem was to
Material Electrode Response Polarity ln X a b maximize MRR with minimum TWR and SR. The optimum
parameters for 20 and 25 vol.% of SiC are given in Table 3.
Al±20% SiC Copper MRR Negative 4.08 0.53 0.33
Positive 4.88 0.426 0.415
TWR Negative 2.92 2.63 1.52
Positive 6.85 0.76 1.35 6. Conclusion
Ra Negative 0.756 0.4 0.07
Positive 1.76 0.47 0.049
This work evaluates the feasibility of machining Al±SiC
Brass MRR Negative 1.77 1.46 0.35 composites by EDM. Based on the experimental results the
Positive 3.83 0.613 0.263 following can be concluded:
TWR Negative 7.63 0.79 1.12
Positive 8.53 0.351 1.39 1. The MRR was more with positive polarity and increased
Ra Negative 0.341 0.64 0.055
with increase in current and it was more with brass
Positive 3.38 0.57 0.122
electrode in comparison with copper electrode.
Al±25% SiC Copper MRR Negative 4.32 0.34 0.33 2. The increase of either pulse duration or the volume
Positive 4.65 0.292 0.387 percentage of SiC resulted in less MRR and it was more
TWR Negative 7.69 1.4 1.71 with increase in rotational speed.
Positive 2.34 0.90 0.093
3. The TWR was less when volume percentage of SiC
Ra Negative 0.398 0.53 0.06
Positive 0.934 0.383 0.12 particle was less and increased with the increase in
current.
Brass MRR Negative 3.68 0.65 0.34
Positive 4.86 0.414 0.469
4. The SR value (Ra) decreased with decrease in pulse
TWR Negative 8.92 0.82 1.26 current and increased with increase in volume percen-
Positive 7.6 0.82 1.19 tage of SiC.
Ra Negative 1.46 0.63 0.16 5. The optimum process parameters for different volume
Positive 1.67 0.612 0.134 percentages of SiC and electrodes for maximum
material removal rate were obtained.
regression analysis procedure, the constant X and the expo-
nents a and b were estimated for different volume percen-
tages of silicon carbide and for different electrodes with Acknowledgements
positive and negative polarity. Table 2 lists the constants of
the regression equation. The authors are grateful to the Regional Research Labora-
tory, Trivandrum, Kerala, for the supply of the composite
5.1. Optimization materials.

Optimization of machining characteristics of Al/(SiC)p

composites in EDM is to ®nd an optimal solution for a given References
objective subjecting it to the constraints. If the objective is to
maximize the MRR for different volume fractions of com- [1] M.J. Koczac, et al., MMC for ground vehicle, aerospace and
posites and the constraints involved are TWR and SR, then industrial application, in: Suresh, et al. (Eds.), Fundamentals of Metal
linear programming is a strong tool to optimize. In this study, Matrix Composites, Butterworths, Guildford, UK, 1993.
the mathematical models were optimized using genetic [2] S.S. Joshi, N. Ramakrishnan, D. Sarathy, P. Ramakrishnan,
Development of technology for discontinuously reinforced alumi-
nium (DRA) composites in integrated design and process technology,
Table 3 in: Proceedings of the First World ConferenceÐIntegrated Design
Optimum process parameters and Process Technology, Vol. 11, 1995, pp. 492±497.
Workpiece Tool Polarity Discharge Pulse MRR [3] E.L. Gallab, M. Skud, Machining Al/SiC particulate metal matrix
current duration (mm3/min) composites. Part II. Workpiece surface integrity, J. Mater. Process.
(A) (ms) Technol. 83 (1998) 277±285.
[4] E.L. Gallab, M. Skud, Machining Al/SiC particulate metal matrix
20% SiC Copper Negative 8.38 168 29.87 composites. Part III. Comprehensive tool wear models, J. Mater.
Positive 6.22 118 34.52 Process. Technol. 101 (2000) 10±20.
Brass Negative 7.61 144 31.64 [5] P. Koshy, V.K. Jain, G.K. Lal, Experimental investigation into
Positive 6.07 92 43.27 electrical discharge machining with a rotating disk electrode, J. Prec.
Eng. 15 (1993) 6±15.
25% SiC Copper Negative 7.34 112 25.69
[6] P. Arivazhaghan, A. Rajadurai, Electronic discharge machining of
Positive 6.12 108 30.19
Al/SiC carbide MMCs, in: Proceedings of the International
Brass Negative 6.56 136 27.12
Conference on Advanced Mechanics and Industrial Engineering,
Positive 5.91 96 33.54
University of Roorkee, 1997, p. 709.
304 B. Mohan et al. / Journal of Materials Processing Technology 124 (2002) 297±304
[7] M. Ramulu, EDM sinker cutting of ceramic particulate composites
Si±TiB2, Adv. Ceram. Mater. 3 (4) (1988) 324±327.
[8] M. Ramulu, M. Taya, EDM machinability of SiCw/Al composites, J.
Mater. Sci. 24 (3) (1989) 1103±1108.
[9] J.S. Soni, Micro-analysis of debris formed during rotary EDM of
titanium alloy (Ti±6Al±4V) and die steel (T215±Cr12), Wear 177 (1)
(1994) 71±79.
[10] J.S. Soni, G. Chakravarthi, Surface characteristics of titanium alloy
with rotary EDM, Bull. Mater. Sci. 16 (3) (1993) 213±227.
[11] J.S. Soni, G. Chakravarthi, Machining characteristics of titanium alloy
with rotary electric discharge machining, Wear 171 (1994) 51±58.
[12] B.H. Yan, H.C. Tsai, Y.C. Lin, Study on EDM characteristics of
cemented carbides, in: Proceedings of the 14th National Conference
on Mechanical Engineering, The Society of Mechanical Engineers,
1997, pp. 157±164.
[13] C.C. Wang, B.H. Yan, Blind hole drilling of Al2O3/6061 Al
composites using rotary EDM, J. Mater. Process. Technol. 102
(2000) 90±102.