J. SHANTHI et al.: EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF THE WEDM PROCESS ...
349–356
EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF
THE WEDM PROCESS PARAMETERS TO OBTAIN THE MINIMUM
SURFACE ROUGHNESS OF THE Al 7075 ALUMINIUM ALLOY
EMPLOYED WITH A ZINC-COATED WIRE USING RSM AND GA
EKSPERIMENTALNA RAZISKAVA OPTIMIZACIJE PROCESNIH
PARAMETROV @I^NE EROZIJE S CINKOM OPLA[^ENE Cu @ICE
ZA ZAGOTOVITEV MINIMALNE POVR[INSKE HRAPAVOSTI
Al ZLITINE 7075 Z UPORABO POSTOPKOV RSM IN GA
Shanthi Jayachandran
1*
, Mohan Raman
2
, Thangavelu Ramasamy
3
1
Renovation and Modernisation Division, Mettur Thermal Power Station-I, Mettur -636406, Tamilnadu, India
2
Department of Mechanical Engineering, Sona College of Technology, Salem-636005, Tamilnadu, India
3
Renovation and Modernisation Division, Mettur Thermal Power Station-I, Mettur-636406, Tamilnadu, India
Prejem rokopisa – received: 2018-07-28; sprejem za objavo – accepted for publication: 2018-12-17
doi:10.17222/mit.2018.166
Wire Electrical Discharge Machining (WEDM) is widely used for machining conductive materials of intricate, complex and
challenging shapes in the field of aerospace, die and mould making, automobile industries and the medical field. Proper
selection of the WEDM process parameters can give good responses. Out of the various process responses, to achieve the
minimum surface roughness is very difficult, due to arcing by the electrical discharge. The present investigation has been made
to optimize the process parameters of WEDM during machining the Al 7075 aluminium alloy with zinc-coated copper wire
using the Response Surface Methodology (RSM). Four input process parameters of WEDM, i.e., Pulse On Time, Pulse Off
Time, Wire Feed and Wire Tension, were chosen as the process variables to study the process performance and to obtain the
minimum surface roughness. An analysis of variance (ANOVA) was carried out to study the effect of the process parameters on
the surface roughness and a mathematical model was developed. The model has been verified and checked for adequacy. The
best fit surface roughness (Ra) value predicted is 1.048 μm for a Pulse On (TON) value of 120 μsec, a Pulse Off (TOFF) value of 58
μsec, a Wire Feed (WF) of 3 m/min and a Wire Tension (WT) value of 9 gm. Furthermore, it is observed that with an increase in
the pulse-on time the Surface Roughness also increases, and an increase in the pulse-off time and the wire tension, the Surface
Roughness decreases. Wire Feed does not influence much on the Surface Roughness.
Keywords:WEDM (Wire Electrical Discharge Machining), optimisation, surface roughness, RSM, GA, Al 7075
@i~na erozija (WEDM; angl.: Wire Electrical Discharge Machining) se pogosto uporablja za obdelavo in rezanje kompleksno
oblikovanih izdelkov, modelov in orodij iz prevodnih materialov v letalski, orodjarski, modelarski in avtomobilski industriji ter
v zdravstvu. Pri tem postopku je zelo pomembna izbira ustreznih procesnih parametrov WEDM. Doseganje minimalne
povr{inske hrapavosti reza je zelo zahtevno zaradi nastajajo~ega talilnega obloka med razelektrevanjem. Avtorji so optimizirali
procesne parametre WEDM, ki uporablja s cinkom opla{~eno bakreno `ico med mehansko obdelavo Al zlitine 7075 z uporabo
metodologije odgovora povr{ine (RSM, angl.: Response Surface Methodology). Izbrali so {tiri vhodne procesne parametre
WEDM, in sicer: ~as vklopa (TON) in izklopa (TOFF) elektri~nega impulza, hitrost dovajanja `ice WF in napetost `ice WT kot
spremenljivke za {tudij vpliva procesa na zagotovitev minimalne hrapavosti. Z analizo variance (ANOVA) so raziskovali vpliv
procesnih parametrov na povr{insko hrapavost in na njeni osnovi razvili matemati~ni model. Model so verificirali in preizkusili
njegovo ustreznost. Najbolj{e ujemanje vrednosti za povr{insko hrapavost (Ra) je bilo 1,048 μm pri TON vrednosti 120 μsec, TOFF
vrednosti 58 μsec, hitrosti dovajanja `ice WF 3 m/min ter napetosti `ice WT 9 mN/m. Avtorji ugotavljajo, da se je z nadaljnjim
podalj{evanjem ~asa TON pove~evala povr{inska hrapavost. S podalj{evanjem ~asa TOFF in pove~evanjem napetosti `ice pa se je
povr{inska hrapavost zmanj{evala. Hitrost dovajanja `ice ni pomembno vplivala na povr{insko hrapavost.
Klju~ne besede: WEDM – `i~na erozija, optimizacija, povr{inska hrapavost, RSM, GA, Al zlitina 7075
1 INTRODUCTION
Several researchers investigated the different aspects
of WEDM, but no comprehensive research work has
been reported so far in the field of wire electrical
discharge machining of this Al 7075 alloy using
zinc-coated copper wire. Hence, an attempt was made to
explore the influences of selected variable process para-
meters.
G. E. Totten and D. S. Mackenzie
1
reported the ma-
chinability ratings of aluminium alloys span into five
groups, with ratings of A, B, C, D and E, which are
ordered in increasing order of chip length and decreasing
order of surface quality. The Al 5083 aluminium alloy
ranked D is an indicator for poor machinability. WEDM
is one of the latest machining techniques to process the
Al 5083 aluminium alloy to any complex intricate shapes
with high accuracy and precision when comparing with
diamond-based cutting tools. J. Prohaszka et al.
2
inves-
tigated the effect of electrode material coating with zinc,
tin and magnesium on machinability in the WEDM pro-
Materiali in tehnologije / Materials and technology 53 (2019) 3, 349–356 349
UDK 620.1:669.715:620.191.35 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 53(3)349(2019)
*Corresponding author e-mail:
shanthi_jc@yahoo.com (Shanthi Jayachandran)

cess. These coating materials are required for improving
the cutting efficiency because the existing wires do not
fulfill all the requirements. R. Chalisgaonkar and J.
Kumar
3
explored the characteristics of pure titanium dur-
ing rough cut operation, and after the finish cut oper-
ation, they revealed the microstructure analysis of zinc-
coated and in the uncoated wires the erosion during the
rough cut operation was found to be more than the finish
cut. B. Sivaraman et al.
4
stated that in the WEDM process
the Taguchi method is most ideal and suitable, it sim-
plifies the optimisation of multiple performance charac-
teristics by avoiding complicated mathematical compu-
tations. G. Selvakumar et al.
5
presented an optimum
input-parameter combination for the minimum Ra and
the maximum MRR was obtained by an analysis of the
signal-to-noise (S/N) ratio and the process was opti-
mized by a Pareto-optimality approach by machining the
Al 5083 aluminium alloy. D. Siva Prasad et al.
6
inves-
tigated the effect of different WEDM process parameters
on the damping behavior of the A 356.2 aluminum alloy.
The damping capacity of this alloy increases with an
increase in the frequency and increasing Pulse On.
A. Dey et al.
7
examined the machinability of the
cenosphere fly-ash reinforced Al 6061 aluminium alloys
for various combinations of the input process para-
meters. The optimal combination of process parameters
was arrived at for the maximum MRR, the minimum
Tool Wear Rate and the minimum R
a
.K.H.Hoetal.
8
carried out the WEDM process by understanding the
interrelationship between the various factors affecting
the process and identified the optimal machining con-
dition from the infinite number of combinations. The
adaptive monitoring and control systems were imple-
mented to tame the transient WEDM behaviour without
the risk of wire breakages. S. Kuriakose and M. S.
Shunmugam
9
investigated the optimal parameters of the
WEDM process for improving the cutting performance.
There is no single optimal combination of cutting
parameters, as their influences on the cutting velocity
and the surface finish are quite the opposite. In the
present work, a multiple regression model is used to
represent the relationship between the input and output
variables and a multi-objective optimization method
based on a Non-Dominated Sorting Genetic Algorithm
(NSGA) is used to optimize the WEDM process.
A. Sharma et al.
10
made an attempt to machine an Al
6063 / ZrSiO4(p) (5 %) metal-matrix composite using
WEDM. The objective was to investigate the influence of
the process parameters, i.e., T
ON
, T
OFF
, Peak Current and
SV on the Cutting Rate and found experimentally that
increasing the T
ON
and Peak current, the cutting rate
increases, whereas increasing the T
OFF
and SV decreases
the cutting rate. The higher discharge energy associated
with the increased T
ON
, Peak Current and lesser TOFF
and SV leads to more powerful explosions, which in-
crease the cutting rate. H. C. Tsai et al.
11
investigated the
electrode performance and revealed that wires generally
used are of brass or copper with a diameter of about
0.3–0.5 mm, but in recent times coated wires are widely
used, usually zinc coated over brass wire. The impacts of
coated wires were found to have a significant effect. The
productivity and surface roughness were found to be
better than uncoated wires. Several authors
12–14
also
suggested that the concentration of electrical discharges
at a certain point of the wire, which causes an increase in
the localized temperature, resulting in the breakage of
the wire. V. Chengal Reddy et al.
15
discussed the effects
of the input control parameters, such as T
ON
, T
OFF
, Cur-
rent, WT, upper flush and lower flush on the R
a
, MRR
and Kerf Width, while machining the aluminum HE 30
material and suggested the selection of the right com-
bination of input parameters by Grey Relational Analysis
(GRA).
Dain Thomas et al.
16
, developed a second-order
regression model using RSM and found that T
ON
and WT
play a major role in the surface roughness. V. R. Surya et
al.
17
predicted the machining characteristics of an Al
7075-TiB2 composite using ANN for the maximum
MRR, minimum Dimensional Error (DE) and better
surface finish. In this study the control factors considered
were T
ON
, T
OFF
, Current and Bed Speed based on
Taguchi’s L 27 orthogonal array. S. Prashantha et al.
18
investigated the Al 6061 aluminium alloy reinforced with
SiC particles by varying the percentage of SiC from 3 %,
6%and9%byweight. T
ON
, T
OFF
, Current (I) and Bed
speed (BS) are varied to find their effects on the MRR.
From the analysis, the average MRR for an unreinforced
Al 6061 aluminium alloy is 9.2 mm
3
/min and the average
MRR is 9.15 mm
3
/min, 9.13 mm
3
/min and 9 mm
3
/min,
respectively for Al 6061 aluminium alloy MMCs with
3 %, 6 % and 9 % SiC, i.e., the MRR decreases with an
increase of the silicon carbide particles. S. Prasad Ari-
katla et al.
19
executed a study on a titanium alloy using
RSM in WEDM and registered the quality of the
machined surface by T
ON
& Input Power and WT&SV .
In the I case Ra increases while in the II case Ra de-
creases. G. Amitesh and K. Jatinder
20
investigated the
influence of machining parameters on the material
removal rate and the cutting speed for the machining of
Nimonic 80 A with Brass wire as the electrode in wire
electrical discharge machining. From the observation,
the cutting speed (CS) and MRR both increase with an
increase in T
ON
and the Peak Current (IP). Furthermore,
it decreases with an increase in T
OFF
and the spark gap set
voltage. M. T. Antar et al.
21
explored the role of coated
wire in the production and surface integrity, while
machining aerospace alloys in WEDM. Coated wires are
stated to protect the core from thermal shock and also
from wire rupture. Its other effects were found on vibra-
tion, damping effect, heat transfer and resistance, which
ultimately increased the machining speed. N. Kinoshita
et al.
22
observed that wire breaks due to a rapid rise in the
pulse frequency of the gap voltage. They developed a
monitoring and control system that switches off the pulse
generator and the servo system, preventing the wire from
breaking, but it affects the machining efficiency.
The above literature review indicates that most of the
researchers have considered the influence of a limited
J. SHANTHI et al.: EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF THE WEDM PROCESS ...
350 Materiali in tehnologije / Materials and technology 53 (2019) 3, 349–356

number of control parameters on the performance
measures of Wire Electric Discharge Machined parts.
Furthermore, from the literature review it is understand-
able that the impact of coated wires was found to
improve the productivity and surface finish, better than
uncoated wires. Hence, it is intended to investigate the
effect of the selected variable process parameters on the
surface roughness, while machining with zinc-coated
copper wire of diameter 0.25 mm. So far, no such inves-
tigation was carried out on the Al 7075 aluminium alloy.
In this study an attempt has been made to optimize the
various WEDM process parameters such as T
ON
, T
OFF
,
WF and WT to find out the best fit to obtain minimum R
a
.
2 EXPERIMENTAL PART
2.1 Work material
As the Al 7075 Aluminium alloy is a lightweight ma-
terial, zinc-based alloy and possesses excellent corrosion
resistant, it is widely used in marine, aerospace applica-
tions and in the medical field. Aluminium alloys have
very good mechanical properties such as a high tensile
strength, a very high yield strength, good fatigue
strength, and superior corrosion resistance, and average
machinability. However, these alloys were very difficult
to fabricate as they are not ductile and have a low frac-
ture toughness at room temperature. An unconventional
machining process like WEDM is used intensively for a
better process The response is to machine an aluminium
alloy, due to its exceptional strength properties, whereas
it is very difficult to machine in a conventional method.
The analysis of the consequence of different process
parameters is essential.
2.2 Machine tool
WEDM requires thin, single-strand conducting metal
wire that is used as the electrode. There are several elec-
trode materials available, but brass wire is commercially
used. The wire can either be coated wire (zinc, brass, tin,
magnesium etc.) or uncoated wire (copper, brass, molyb-
denum). This conducting wire is passed through the
pre-drilled hole in the metal piece to be machined. The
wire is fed from a spool and held between upper and
lower guides tha are made of diamond, and in turn it is
finally controlled by CNC and moved in the X-Y plane.
This allows the wire-cut EDM to be programmed to cut
very intricate, delicate and complex shapes. By using a
pump, dielectric fluid (Deionised Water) is continuously
passed over the work piece to remove, clear and flush out
the debris that is cut from the work piece. When a D.C.
supply is attached to the circuit, thousands of spark dis-
charges occur across the gap between the wire and the
work piece, which increases the temperature and causes
the melting of material, erosion and even vaporizing and
thus removing the metal from the work piece. The
removed fine material particles are carried away by the
dielectric fluid circulating around it.
J. SHANTHI et al.: EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF THE WEDM PROCESS ...
Materiali in tehnologije / Materials and technology 53 (2019) 3, 349–356 351
Table 1: Al 7075 alloy composition in percentage by weight
Al Cr Cu Fe Mg Mn Si Ti Zn others
87.1–91.4 0.18–0.28 1.2–2.0 Max 0.5 2.1–2.9 Max 0.3 Max 0.4 Max 0.2 5.1–6.1 Max 0.15
Figure 1: WEDM machine set up
Table 2: Properties of Al 7075 Aluminium Alloy
Sl. No. Property Value
1 Density 2.81 g/cc
2 Ultimate Tensile strength 572 MPa
3 Tensile Yield strength 503 Mpa
4 Elongation at break 11 %
5 Modulus of Elasticity 71.7 GPa
6 Poisson Ratio 0.33
7 Hardness Brinell 150
8 Fatigue strength 159 MPa
9 Shear Modulus 26.9 GPa
10 Machinability 70 %
11 Thermal Conductivity 130 W/m k
12 Melting Point 477–635 °C
13 Electrical Resistivity 5.15 e
–006
 cm

2.3 Specimen
An Al 7075 aluminum alloy rectangular block of
length 200 mm, width 16 mm and thickness 10 mm was
taken as the work material. The experiment was per-
formed on a WEDM machine with zinc-coated copper
wire (tool) of 0.25 mm diameter and deionized water is
used as the dielectric fluid.
2.4 Parameters considered in this experiment
Table 3: Fixed parameters
Sl. No. Parameter Unit value
1 Input Power V 230
2
Dielectric fluid
pressure
kg/cm
2
1 machine unit
(low)
3 Pulse Peak Voltage V 2 machine unit
4 Servo Voltage V 20
5 Servo Frequency cycles/s 2100
Table 4: Variable parameters and range
Sl. No. Parameter Unit From To
1 Pulse On (T
ON
) μs 120 128
2 Pulse Off (T
OFF
)μ s5 05 8
3 Wire Feed (WF) m/min 1 3
4 Wire Tension (WT)N59
2.5 Experimental values
Table 5: Experimental results of R
a
using L27 orthogonal array matrix
Variable Process
Parameters
Response R
a
Error
Re-
marks
Sl. no.
Pulse
on
Pulse
off
Wire
Feed
Wire
Ten-
sion
Exper-
imen-
tal
RSM
pre-
dicted
μs μs m/min N μm μm %
1 128 54 2 5 1.84 2.08 11.62
2 120 54 1 7 1.80 1.92 6.44
3 124 58 2 9 1.36 1.39 1.88 WB
4 124 58 1 7 1.91 1.95 1.95
5 124 50 2 5 1.79 2.19 18.12
l6 124 54 2 7 1.94 2.15 9.94
7 120 58 2 7 1.62 1.48 9.31
8 128 54 3 7 1.91 2.22 13.96
9 128 54 1 7 1.93 2.23 13.38
10 124 54 3 5 1.87 2.02 7.61
11 124 54 3 9 1.74 1.94 10.12
12 120 54 2 5 1.73 1.69 2.37
13 120 54 2 9 1.35 1.55 13.13
14 124 54 2 7 1.93 2.15 10.40
15 128 54 2 9 1.39 1.88 26.14 WB
16 128 50 2 7 1.98 2.47 19.97
17 124 50 1 7 1.86 2.27 17.99
18 124 50 3 7 2.11 2.52 16.14
19 120 54 3 7 1.67 1.80 7.43
20 128 58 2 7 1.62 1.62 0.12
21 124 54 2 7 1.98 2.15 8.08
22 124 54 1 5 2.02 2.17 6.83
23 120 50 2 7 1.55 1.89 17.99
24 124 54 1 9 1.73 1.92 9.90
25 124 50 2 9 1.65 2.11 21.95
26 124 58 2 5 1.70 1.65 3.03 WB
27 124 58 3 7 1.54 1.57 2.04
2.6 Surface roughness
Roughness is a measure of the texture (quality) of the
surface. It is quantified by the vertical deviations of the
actual surface from its ideal form. If these deviations are
large, the surface is rough; if they are small, the surface
is smooth. The surface roughness R
a
was measured with
a Mitutoyo Surftest 211 Surface Roughness tester and
the values are tabulated in Table 5.
2.7 Response Surface Methodology (RSM)
This is a collection of statistical and mathematical
techniques useful for developing, improving and opti-
mizing processes. RSM consists of an experimental
approach to investigate the independent variables in the
process, the experimental statistical model developed for
an appropriate similar relationship between the yield and
the process variables. Optimization methods for finding
values of the process variables that produce desirable
values of the responses. In order to investigate the effects
of the WEDM parameters on the above-mentioned ma-
chining criteria, second-order polynomial response sur-
face mathematical models can be developed. In the gene-
ral case, the response surface is described by an equation
of the form:
Y =  0
+ 
j
x
j
+ 
jj
x
2
j
+ 
ij
x
i
x
j
(1)
where Y is the response, in current research surface
roughness, whereas the terms  0
,  j
,  ij
are second-order
regression coefficients. The second term under the sum-
mation sign of this polynomial equation is attributable
to a linear effect, whereas the third term corresponds to
the higher-order effects and the fourth term of the
equation includes the interactive effects of the process
parameters. The above equation can be rewritten as:
Y =  0
+  1
X
1
+  2
X
2
+  3
X
3
+  4
X
4
+  11
X
12
+  22
X
22
+
 33
X
32
+  44
X
42
+….. (2)
J. SHANTHI et al.: EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF THE WEDM PROCESS ...
352 Materiali in tehnologije / Materials and technology 53 (2019) 3, 349–356
Figure 2: a) Work piece before machining and b) after machining

The value of  , the regression coefficient, will be
determined by the least-squares method.
2.8 Wire breakage
A wide variety of the control strategies preventing the
wire from breaking are based on a knowledge of the
characteristics of the wire. The breaking of the wire can
be due to the excessive thermal load producing unwar-
ranted heat on the wire. Most of the thermal energy
generated during the WEDM process is transferred to the
wire, while the rest is lost to the flushing fluid. However,
when the instantaneous energy rate exceeds a certain
limit depending on the thermal properties of the wire
material, the wire will break. The WB in the Table 5
indicates the occurrence of wire breakage while conduct-
ing the experiment, hence the corresponding set of
parameters to be avoided for better performance.
3 RESULTS
3.1 Response Surface Methodology (RSM)
The relationship between the selected variable pro-
cess parameters and the response R
a
was obtained by a
quadratic regression equation using RSM in Minitab.
This regression equation is useful in predicting the
response surface roughness with respect to the variable
process parameters Pulse on Time (A), Pulse off Time
(B), Wire Feed (C) and Wire Tension (D).
R
a
= –231.114 + 2.903 A + 1.773 B + 0.866 C + 1.162
D – 0.01 A
2
– 0.008 B
2
+ 0.05 C
2
– 0.048 D
2
– 0.007
A·B + 0.007 A·C – 0.002 A·D – 0.039 B·C – 0.006
B·D – 0.02 C·D
Figure 3 shows the difference between the experi-
mental and predicted R
a
. The error percentage was cal-
culated based on the input variable process parameters
and the predicted value, and tabulated in Table 5, and
found to be reasonable.
3.2 Analysis of variance for R
a
Table 6: ANOVA result for R
a
Source DF Seq SS Adj MS F
% Contri-
bution
Regression 14 1.02098 0.072927 15.96
Linear 4 0.45663 0.057484 12.58
Pulse on 1 0.07521 0.175001 38.31 6.99
Pulse off 1 0.11801 0.132032 28.90 10.97
Wire Feed 1 0.01401 0.002568 0.56 1.30
Wire Tension 1 0.24941 0.018231 3.99 23.18
Square 4 0.40138 0.100344 21.97
Pulse on*
Pulse on
1 0.07993 0.140833 30.83 7.43
Pulse off*
Pulse off
1 0.05184 0.083333 18.24 4.82
Wire Feed*
Wire Feed
1 0.07707 0.133333 2.92 7.16
Wire Tension*
Wire Tension
1 0.19253 0.192533 42.15 17.90
Interaction 6 0.16297 0.027162 5.95
Pulse on*
Pulse off
1 0.04623 0.027162 10.12 4.30
Pulse on*
Wire Feed
1 0.00302 0.046225 0.66 0.28
Pulse on*
Wire Tension
1 0.00122 0.003025 0.27 0.11
Pulse off*
Wire Feed
1 0.09610 0.001225 21.04 8.93
Pulse off*
Wire Tension
1 0.01000 0.096100 2.19 0.93
Wire Feed*
Wire Tension
1 0.00640 0.010000 1.40 0.59
Residual Error 12 0.05482 0.006400
Lack-of-Fit 10 0.05342 0.004568 7.63
Pure Error 2 0.00140 0.005342
Total 26 1.07580 0.000700
3.3 Contribution of the process variable parameter
during machining
Table 6 shows the results of the ANOVA for a 95 %
confidence level of R
a
. It is observed from the table that
foremost variable that affects R
a
is linear WT with a con-
tribution of 23.18 %. The second important factor is
squared WT with a contribution of 17.90 %, then linear
T
OFF
with 10.97 %
,
interaction of T
OFF
× WF, squared T
ON
,
squared WF, linear T
ON,
squared T
OFF
, interaction of T
ON
×
T
OF F
and linear WF, with a contribution of (8.93, 7.43,
7.16, 6.99, 4.82, 4.30 and 1.30) % respectively. In addi-
tion to the above, it is found that interaction of T
OFF
×
WT, WF × WT, T
ON
× WF, T
ON
× WT contribution values
are less than 1 %, which states that these interactions do
not affect R
a
. Further, from Table 5, it is observed that
the predicted values from RSM are close to the experi-
mental values of R
a
. The higher correlation coefficient
(R
2
) of more than 90 % confirms the fitness of the model.
Figure 4a shows the fitness of the experimental R
a
obtained to the predicted R
a
. It is observed that there is
not much deviation found between the experimental
value and the predicted value obtained from RSM. Fig-
J. SHANTHI et al.: EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF THE WEDM PROCESS ...
Materiali in tehnologije / Materials and technology 53 (2019) 3, 349–356 353
Figure 3: Experimental vs. predicted values of R
a

ure 4b shows the effect plot for the SN ratios of the data
mean of R
a
, R
a
is the minimum for the minimum T
ON
,
maximum T
OFF
, moderate WF & maximum WT. R
a
is the
maximum for mid-value T
ON
, minimum T
OFF
, minimum
WF and WT.
3.4 Contour and surface plots of RSM
4 DISCUSSION
4.1 Discussion of contour plot and surface plot ob-
tained from RSM
Figure 5a depicts the effect of T
ON
and T
OFF
on R
a
when WF and WT are held constant, for the T
ON
value
ranging from 123.5 μs to 128 μs and T
OFF
rate from 50 μs
J. SHANTHI et al.: EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF THE WEDM PROCESS ...
354 Materiali in tehnologije / Materials and technology 53 (2019) 3, 349–356
Figure 5: Contour plot and surface plot of surface roughness: a) R
a
versus T
OFF
, T
ON
,b)R
a
versus WF, T
ON
,c)R
a
versus WT, T
ON
,d)R
a
versus
WF, T
OFF
,e)R
a
versus WT, T
OFF
,f)R
a
versus WT, WF
Figure 4: a) Fitness of the experiment, b) effect of SN ratios

to 54.5 μs, the R
a
value is a maximum (>1.90 μm). There
is a gradual increase in the R
a
for the intermediate T
OFF
value and lower for higher and lower T
OFF
value, as the
T
ON
value decreases the R
a
value also decreases. And for
the T
ON
value ranging from 120 μs and T
OFF
value 50 μs,
the R
a
value is a minimum (<1.60 μm).
Figure 5b depicts the effect of T
ON
and WF on R
a
when T
OFF
and WT are held constant, for the T
ON
value
ranging from 122.5 μs to 126.5 μs and for the WF value
ranging between 1 m/min and 1.25 m/min, the R
a
value is
a maximum (>2.00 μm), hence this condition should be
avoided. As the T
ON
value decreases and WF increases
simultaneously the R
a
value decreases. For the T
ON
value
of 120 μs and the WF range from 2.5 m/min to
3.0 m/min, the R
a
value is a minimum (<1.7 μm). This is
the ideal condition.
Figure 5c depicts the effect of T
ON
and WT on R
a
when T
OFF
and WF are held constant, for the value of T
ON
ranging from 122.5 μs to 128 μs and for WTvalue5Nto
7N,theR
a
value is a maximum (>1.9 μm), for the T
ON
rate 120 μs to 121.5 μs and WT values from 8.5 N to 9 N,
the R
a
value is a minimum (<1.4 μm). As the T
ON
value
decreases and WT value increases simultaneously, the R
a
values decreases further for selected low T
ON
value and
high WT value the R
a
value is found to be a minimum.
Figure 5d depicts the effect of T
OFF
and WF on R
a
when T
ON
and WT are held constant, the R
a
value is a
maximum (>2.0 μm ) for a T
OFF
value of 50 μs to 53 μs
and WF value above 2.5 m/min, i.e., in general it is clear
that for minimum T
OFF
value and maximum WF value the
R
a
is a maximum. For a T
OFF
of more than 58 μs and WF
value of 3.0 m/min, the R
a
is minimum (<1.6 μm) and for
the other values of T
OFF
&WF, the R
a
value is medium.
Figure 5e depicts the effect of T
OFF
and WT on R
a
when T
ON
and WF are held constant, it is observed that
for the T
OFF
value ranging from 50 μs to 56 μs and WT
value ranging from 5.5 N to 7.5 N, the R
a
value is a
maximum (>1.9 μm), the Surface Roughness will be in
the extreme condition, and for T
OFF
value exceeding
57.5 μs and WT value exceeding 8.7 N, the R
a
value is a
minimum (<1.4 μm). As the T
OFF
and WT increases
simultaneously the R
a
value decreases.
Figure 5f depicts the effect of WF and WT on R
a
when T
ON
and T
OFF
are held constant. When the WF
ranges from 1 m/min to 1.75 m/min and WT values
ranges from5Nto7.3N,theR
a
value is a maximum
(>2.0 μm) the Surface Roughness will be in the ex-
cessive, which is an adverse condition. For the all values
of WF and WT the value exceeding 8.8 N R
a
is a mini-
mum (<1.7 μm), as WT increases R
a
decreases.
4.2 Simulation results obtained from the genetic algo-
rithm
The equation obtained from RSM, i.e., Equation (3),
is used for the minimization of R
a
by using GA in
MATLAB software to find the optimal selected variable
process parameters for the best fit R
a
values. Where T
ON
,
T
OFF
, WF and WT are 4 variables, the population size
selected for this simulation is 100, the Rank is set as the
Fitness Scaling Function, Stochastic uniform is set as the
Selection Function, the Reproduction elite count is 2, the
Cross over probability is 0.8, the Cross over function is
scattered, the Mutation probability is 0.05, the Initial pe-
nalty is 10, the Iteration is 100, the Number of genera-
tions is 100, and the Stopping criteria is Best Fitness.
The optimization function is formulated as
Minimize R
a
(T
ON
, T
OFF
, WF, WT)
Subject to the following condition:
120 μs < T
ON
> 128 μs
50 μs < T
OFF
>58 μs
1 m/min < WF > 3 m/min
5N<WT >9N
Figure 6 depicts the best optimal solution for the
Surface Roughness obtained in the simulation, for the
input value of T
ON
– 120 μs, T
OFF
-58 μs, WF-3 m/min,
WT-9 N the best fit Predicted R
a
Value is 1.048 μm.
4.3 Confirmation experiment
Confirmatory experiments carried out for the input
value of T
ON
120 μs, T
OFF
58 μs, WF 3 m/min, WT9N
and the experimental R
a
value obtained is 1.08 μm and
the error percentage between the predicted and experi-
mental value is calculated as 3.05, which is less than
5 %. It confirms the excellent reproducibility of the
results. Table 7 indicates the error obtained.
Table 7: Comparison of the Optimized Process Parameter and its best
fit R
a
value
Sl. no. Parameter Values
1 Pulse On (TON
) 120 μs
2 Pulse Off (TOFF
)5 8 μ s
3 Wire Feed (WF) 3 m/min
4 Wire Tension (WT)9 N
Best Fit R a
Value
(Predicted) μm
1.048
Actual R a Value
(Experimental) μm
1.08
Error % 3.05
J. SHANTHI et al.: EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF THE WEDM PROCESS ...
Materiali in tehnologije / Materials and technology 53 (2019) 3, 349–356 355
Figure 6: Predicted optimal solution of R
a
using GA

5 CONCLUSIONS
The present work elucidates the effect of variable
WEDM process parameters while machining an Al 7075
alloy is investigated by using zinc-coated copper wire, a
quadratic model for the Surface Roughness was deve-
loped, using RSM, to correlate the effects of the process
parameters and the same is used in GA to establish the
best fit R
a
. Based on the results obtained the following
conclusions were furnished for machining of an Al 7075
alloy using zinc-coated copper wire.
The best fit Surface Roughness value predicted is
1.048 μm for Pulse On (T
ON
) value 120 μs, Pulse Off
(T
OFF
) value 58 μs, Wire Feed (WF) 3 m/min and for
Wire Tension (WT) value 9 gm.
It is observed that for the increase in the Pulse On
time, the Surface Roughness also increases.
The Surface Roughness decreases as the Pulse Off
time increases.
The increase in the Wire Tension leads to a decrease
of the Surface Roughness.
Wire Feed does not influence much on the Surface
Roughness.
For the maximum value of Pulse Off (T
OFF
), Wire
Feed (WF), the Wire Tension (WT) and minimum Pulse
On (T
ON
) the Surface Roughness predicted is a minimum,
which is the ideal condition for improving the quality of
machined parts.
Furthermore, WEDM can be employed for ma-
chining an Al 7075 alloy with other coated and uncoated
wires in order to compare the machined surfaces.
6 REFERENCES
1
G. E. Totten, D. S. Mackenzie, Handbook of Aluminium [M]//Vol 1,
Physical metallurgy and processes, Marcel Dekker, New York and
Basel 2003
2
J. Prohaszka, A. G. Mamalis, N. M. Vaxevanidis, The effect of elec-
trode material on machinability in wire electro-discharge machining,
J Mater Process Technol, 69 (1997) 233–237, doi:10.1007/s00170-
016-9598-9
3
R. Chalisgaonkar, J. Kumar, Investigation of the machining parame-
ters and integrity of the work and wire surfaces after finish cut
WEDM of commercially pure titanium, Brazilian Society Mecha-
nical Science and Engineering (2015), doi:10.1007/s40430-
015-0335-3
4
B. Sivaraman, C. Eswaramoorthy, E. P. Shanmugham, Optimal con-
trol parameters of machining in CNC Wire-Cut EDM for Titanium,
Int. Journal of Applied Sciences and Engineering Research, 4 (2015)
1, doi:10.6088/ijaser.04011
5
G. Selvakumar, G. Sornalatha, S. Sarkar, S. Mitra, Experimental
investigation and multi-objective optimization of wire electrical
discharge machining (WEDM) of 5083 aluminum Alloy, Transaction
of Nonferrous Metals Society of China 24 (2014) 373–379,
doi:org/10.1016/S1003-6326(14)63071-5
6
D. Siva Prasad, C. Shoba, K. Rahul Varma, A. Khurshid, Influence of
wire EDM parameters on the damping behaviour of A 356.2 alumi-
num Alloy, Journal of Alloys and Compounds (2015),
doi:10.1016/j.jallcom.2015.06.050
7
A. Dey, S. Debnath, K. M. Pandey, Optimization of electrical dis-
charge machining process parameters for Al6061/Cenosphere
composite using grey-based hybrid approach, Transaction of Non-
ferrous Metals Society of China, 27 (2017), 998–1010,
doi:10.1016/S1003-6326(17)60117-1
8
K. H. Ho, S. T. Newman, S. Rahimifard, R. D. Allen, State of the art
in wire electrical discharge machining (WEDM), International
Journal of Machine Tools & Manufacture, 44 (2004), 1247–1259,
doi:10.1016/j.ijmachtools.2004.04.017
9
Shajan Kuriakose, M. S. Shunmugam, Multi-objective optimization
of wire-electro discharge machining process by Non-Dominated
Sorting Genetic Algorithm, Journal of Materials Processing Tech-
nology, 170 (2005), 133–141, doi:10.1016/j.jmatprotec.2005.04.105
10
A. Sharma, M. Pal Garg, K. Kumar Goyal, Prediction of optimal
conditions for WEDM of Al 6063/ ZrSiO4 (p) MMC, Procedia
Materials Science, 6 (2014), 1024–1033, doi:10.1016/j.mspro.2014.
07.173
11
H. C. Tsai, B. H. Yan, F. H. Huang, EDM performance of Cr/Cu
based composite electrodes, International Journal of Machine Tools
& Manufacturing, 43 (2003) 3, 245–252, doi:10.1016/j.matpr.
2015.07.328
12
K. Shoda, Y. Kaneko, H. Nishimura, M. Kunieda, M. X. Fan, Adap-
tive control of WEDM with on-line detection of spark locations,
Proc. of the Tenth International Symp. for Electro-Machining
(ISEM-10), Magdeburg 1992
13
M. Kunieda, H. Kojima, N. Kinoshita, On-line detection of EDM
spark location by multiple connection of branched electric wires,
Ann. CIRP 39 (1) (1990) 171–174, doi:10.1016/S0007-8506(07)
61028-2
14
T. Tanimura, C. J. Heuvelman, P. C. Vennstra, The properties of the
servo gap sensor with wire spark-erosion machining, Ann. CIRP 26
(1) (1977) 59–63
15
V. Chengal Reddy, N. Deepthi, N. Jayakrishna, Multiple response
optimization of wire EDM on aluminium HE30 by using grey
relational analysis, Materials Today: Proc. 2, 2015, 2548–2554,
doi:10.1016/j.matpr.2015.07.201
16
D. Thomas, R. Kumar, G. K. Singh, P. Sinha, S. Mishra, Modelling
of surface roughness in coated wire electric discharge machining
through response surface methodology, Materials Today: Proc. 2,
2015, 3520–3526, doi:10.1016/j.matpr.2015.07.091
17
V. R. Surya, K. M. Vinay Kumar, R. Keshavamurthy, G. Ugrasen, H.
V. Ravindra, Prediction of machining characteristics using artificial
neural network in wire EDM of Al 7075 based in-situ composite,
Materials Today: Proc. 4 (2017), 203–212, doi:10.1016/j.matpr.
2017.01.014
18
S. Prashantha, R. B. Veeresha, S. M. Shashidhara, U. S. Mallikarjun,
A. G. Shivasiddaramaiah, A study on machining characteristics of Al
6061-Sic metal matrix composite through wire – cut electro dis-
charge machining, Materials Today: Proc. 4 (2017), 10779–10785,
doi:10.1016/j.matpr.2017.08.027
19
S. Prasad Arikatla, K. TamilMannan, Arkanti Krishnaiah, Parametric
optimization in wire electrical discharge machining of titanium alloy
using response surface methodology, Materials Today: Proc. 4,
(2017), 1434–1441, doi:10.1016/j matpr 2241-7853
20
G. Amitesh & K. Jatinder, An investigation into the machining cha-
racteristics of NIMONIC 80A using CNC wire EDM, International
Journal of Advanced Engineering Technology, 3 (2012) 1, 170–174
21
M. T. Antar, S. L. Loo, D. K. Aspinwall, Productivity and work piece
surface integrity when WEDM aerospace alloys using coated wires,
Journal of Procedia engineering (2011)
22
N. Kinoshita, M. Fukui, G. Gamo, Control of wire-EDM preventing
electrode from breaking, Ann. CIRP 31 1 (1982), 111–114,
doi:10.20965/ijat.2014.p0468
J. SHANTHI et al.: EXPERIMENTAL INVESTIGATION FOR THE OPTIMIZATION OF THE WEDM PROCESS ...
356 Materiali in tehnologije / Materials and technology 53 (2019) 3, 349–356