Materials Today: Proceedings xxx (xxxx) xxx

Contents lists available at ScienceDirect

Materials Today: Proceedings

journal homepage: www.elsevier.com/locate/matpr

Influence of near dry wirecut electrical discharge machining parameters

on kerf width in Monel 400
N.E. Arun Kumar ⇑, N. Sathishkumar, E. Raviraj, M. Pathri Narayanan, Roderik Eugene
Department of Mechanical Engineering, St. Joseph’s College of Engineering, Old Mamallapuram Road, Chennai, Tamilnadu 600119, India

a r t i c l e i n f o a b s t r a c t

Article history: This work explains the more refined and subtle analysis on one of the significant response parameters in
Received 17 February 2020 kerf width upon machining Monel 400. Machining was done under near-dry ambiance with the support
Received in revised form 12 May 2020 of the mathematical modeling technique of Response Surface Methodology (RSM) and the more sophis-
Accepted 17 May 2020
ticated predictive computational modeling in an artificial neural network (ANN). Driving control ele-
Available online xxxx
ments considered for this WEDM process were pulse-on and off time (TON and TOFF), Voltage (V),
Wire Feed (WF), Air inlet pressure and Water flow rate. The impact on the intermittent zone that prevails
Keywords:
between the surface of the material in the spark zone and the wire during the process has not only being
WEDM
Near Dry
influenced by the control parameters but also by the dielectric medium. The implications of di-electric
RSM medium maintained at near dry conditions were given utmost importance and considerations in this
ANN investigation. The execution of the RSM and created ANN models are contrasted and trial consequences
Kerf Width of Kerf Width for their adequacy. The examination demonstrates that ANN shows more exact expectation
Air Water Mist contrasted with the RSM models.
Ó 2020 Elsevier Ltd. All rights reserved.
Selection and Peer-review under responsibility of the scientific committee of the 2nd International Con-
ference on Recent Trends in Metallurgy, Materials Science and Manufacturing.

1. Introduction
The demand for enhanced productivity without inflicting infec-
tiousness to the eco-friendliness of the ambiance brings in innova-
tive methods in existing manufacturing techniques, one such
attempt is the introduction of near dry method in the prevalent
advanced technique of wire-cut EDM. Near dry machining reduces
the unfriendly impacts on both the operator and condition. Intri-
cate profiles of machine components could be machined using a
widely accepted nonconventional Wire Electric Discharge Machin-
ing (WEDM).The challenge posed to this process is the use of
dielectric fluid. Dry-EDM was initially endeavored by Kunieda
et al. [1]. The investigators observed positive aspects of dry
machining such as significant low electrode wear proportion and
smaller release crevice length. It was found that 3D machining is
possible with a uniform high-speed air stream over the working
hole since tool electrode wear is negligible. Kao et al. [2] attempted
⇑ Corresponding author.
E-mail addresses: arunkumarne@stjosephs.ac.in, nearunkumar@gmail.com
(N.E. Arun Kumar).
mist influenced near dry EDM process using fluid gas blend as the
dielectric liquid to use the benefit of centralization of fluid and
properties of the dielectric medium to accomplish greater surface
finish. The researchers extended their work by comparing the flow
rates of near dry and wet EDM. This process improved MRR and
eradicated debris deposition. A mathematical model was proposed
for gap separate in light of dielectric quality and dynamic viscosity
of water air blend. Joshi et al. [3] done trials on dry EDM utilizing
throbbling magnetic field.The field influences major current carri-
ers in the electrode and improved the material removal rate, geo-
metric accuracy due to higher ionization and highest transfer of
thermal energy to the workpiece. Zhan Bo yu et al. [4] attempted
an examination of correlation upon the machinability attributes
while embracing the processing procedure under three working
environments of EDM on carbide. EDM milling under dry ambience
showed high machining velocity and lower tool wear proportion
compared to oil EDM milling. Sourabh K Saha et al [5] have ana-
lyzed the EDM under dry ambiance. The investigation report con-
firms that, TWR (1% of MRR) gives a minimum value in the
experimental outcomes and they are independent of the input
parameters whereas Surface finish has reliance on pulse on time

https://doi.org/10.1016/j.matpr.2020.05.471
2214-7853/Ó 2020 Elsevier Ltd. All rights reserved.
Selection and Peer-review under responsibility of the scientific committee of the 2nd International Conference on Recent Trends in Metallurgy, Materials Science and
Manufacturing.

Please cite this article as: N. E. Arun Kumar, N. Sathishkumar, E. Raviraj et al., Influence of near dry wirecut electrical discharge machining parameters on
kerf width in Monel 400, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.05.471
2 N.E. Arun Kumar et al. / Materials Today: Proceedings xxx (xxxx) xxx

and voltage. Boopathi et al. [6] performed experiments utilizing air the effects of optimized input parameters in influencing a better
mist and oxygen mist, the disclosures are that at a decent machin- kerf width only under wet and dry electrical discharge machining
ing speed a superior SF is achievable under direct air mist/oxygen ambiance, whereas very minimum attempts were made for the
mist pressure. Experiments have also been attempted in near dry studies on kerf width under near dry conditions. In this current
WEDM using air mist. The parametric investigation was completed study, a mechanism has been devised and incorporated to bring
by embracing Taguchi technique and ANOVA. The prediction for about subtle variations within the near dry domain and their
optimal cutting inputs was done by Pareto-front with the support effects on kerf width were investigated.
of an algorithm evolved out of multiple objectives (MOEA). Dhobe
et al. [7] concentrated the advanced parameters and their impacts
2. Materials and methods
on SF when cryo-treated D2 steel was machined utilizing WEDM.
Regression analysis was used to develop a mathematical model
Monel 400 is the material chosen for this work carries the
with an intention to elevate the surface quality. Ahad Gholipur
dimension of 4 mm in thickness, which finds its application exten-
et al. [8] made a comprehensive research on drilling of SPK steel
sively in Chemical and Hydrocarbon processing equipment. The
utilizing an electrical discharge drilling under dry and wet condi-
Chemical constituents of Monel alloy is presented in Table 1. The
tions to accomplish better surface integrity. Also, comparative
experiments were conducted on the CONCORD WEDM DK7725
analysis of the responses in terms of MRR and tool wear ratio
Machine. Its Working table travel is 250x320mm. Working table
was made under wet and dry machining conditions. It has been
dimension is 400x500mm with maximum machining thickness
observed that EDM in wet condition has comparatively better
300 mm. Electrical parameters namely TON, TOFF,WF and servo
MRR,TWR and surface roughness than dry EDM under a high dis-
voltage (SV) were identified as influential factors to concentrate
charge energy level, whereas near dry gave better MRR and least
their effects on kerf during near dry WEDM. Fixed Parameters for
Surface Roughness. Micrographs under SEM revealed better surface
experimentation are given in table 2. Parameter Ranges were
finish while machining under near dry condition. Iskandar et.al. [9]
picked on the introduction of machining limit of the machine
Studied and analyzed the flow representation and portrayal for
and preliminary investigations coordinated by using OVAT
enhanced MQL machining of composites. The outcomes demon-
approach as showed up in Table 3.
strated that by expanding AFR, the greatness of the pinnacle veloc-
The dielectric fluid mixture of air and water in the form of a
ity increments and the axial velocity conveyance turns out to be
mist at an appropriate proportion and at controllable pressure is
more symmetrical around the spout pivot. Di et al. [10] analysed
supplied at the time of machining with the help of a designed cir-
cutting width in micro-WEDM and devised a model for kerf width
cuit being incorporated within the workspace of the machine table,
considering the input control parameters and amplitude of wire
the machine chosen was exclusively designed so as to use only liq-
vibration, explaining the variation of cutting width. WangYua-
uids as the dielectric medium. The designed and fabricated circuit
nyuan [11] studied the kerf width varies with various micro-
attachment (Fig. 1) helps to have a suitable and proportionate mix
EDM parameters, includes wire tension and open voltage. From
of the deionized water with compressed air. The objective of the
the results, he observed that Kerf variations in miniature parts is
designed and fabricated hydro-pneumatic flow circuit is to regu-
larger than that of conventional WEDM.Tosun et al. [12] utilized
late the mass flow rate of the mist aimed at enhancing the near-
Taguchi technique to examine the various control factors on MRR
dry WEDM.
and kerf for steel.They observed that Voltage and pulse time dura-
To attain air–water mist, compressed air and deionized water
tion are dominant factors compared to wire speed and dielectric
are mixed and to be let out through a nozzle in the dielectric sup-
pressure. Pragya Shandilya et.al [13] created RSM and ANN based
ply system. It comprises of a pneumatic flow line with a properly
numerical demonstrating for foreseeing normal cutting speed of
connected air compressor, an appropriately located water storage
metal grid composite (MMC) amid WEDM. The authors considered
tank and a flow control valve coupled with an FRL unit. The release
four WEDM parameters in particular Servo voltage, TON, TOFF and
was computed by the time taken for collecting the unit volume of
WF as machining procedure parameters and also established the
water collected. Preparatory experiments conducted by varying
process model by using back propagation neural network. It was
pressure of the mist between 2 bar and 6 bar.From the experimen-
also concluded that voltage is important parameter than wire feed
tal observations, It is observed that unrestricted spark happens at
and off time for 10% SiCp/6061 Al MMC. Reza Kashiry Fard et.al
3 bar pressure with a flow rate of 5 ml/min, so These constant val-
[14] planned and led tests in light of L27 OA to concentrate the
ues have been maintained in the experimentation to attain better
impact of voltage, off and on time,release current, tool(wire) feed
material removal performance.
and tension upon surface roughness(SR) and cutting speed (CV)
of Al-SiC metal network composite by dry WEDM. The metal wire
prompts expanding in start effectiveness and electrostatic powers
Table 2
because of its higher resistance compared with copper wire. Neeraj Fixed Parameters of Setting.
Sharma et.al [15] performed tests to examine the influence of pro-
Work piece Monel 400alloy
cess factors on overcut during WEDM on HSLA steel with metal
wire. It was finished up from this work, the error rate between Tool material Brass of 0.25 mm thickness
the anticipated and trial values lies in the scope of ± 10% It is like- Shape Rectangular bar
Length of the work piece 200 mm
wise derived from this work the model created was noteworthy Width of the work piece 100 mm
and reproducibility of the outcomes was great with the estimation Thickness of the work piece 5 mm
of R2 as 0.9863. The extensive literature survey taken up has given Cutting Conditions Near-Dry
an impression that investigators showed enthusiasm in studying

Table 1
Percentage Chemical Composition of Monel Alloy.

Nickel Copper Iron Manganese Silicon Carbon Sulphur

63.96 31.539 1.816 0.990 0.171 0.109 0.007

Please cite this article as: N. E. Arun Kumar, N. Sathishkumar, E. Raviraj et al., Influence of near dry wirecut electrical discharge machining parameters on
kerf width in Monel 400, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.05.471
 N.E. Arun Kumar et al. / Materials Today: Proceedings xxx (xxxx) xxx 3

Table 3
Process parameters and their levels.

Parameters Unit Symbol Levels

Low Medium High
Pulse-ON ls A 1 3 5
Pulse-OFF ls B 6 8 10
Voltage V C 50 55 60
Wire Feed mm/min D 6 8 10

xk. This has been by and large used as a piece of showing of

Fig. 1. Near-Dry WEDM process (Courtesy: Arun Kumar et al. 2017).
machining systems. On the off chance that every one of the factors
are thought to be quantifiable, the reaction surface can be commu-
nicated. On the off chance that the normal reaction is indicated by
E(y) = g, then the surface spoken to by g = f(x1, x2, . . ...xk). The
regression is carried out with a specific end goal to depict the infor-
mation gathered whereby a watched reaction is approximated in
light of a utilitarian connection between the assessed variable. By
the experimental procedure, sufficient data has been collected
and with the help of RSM method, the required optimized outputs
need to be obtained. The values of great relevance and significance
were chosen from Table 3 as input parameters were recorded. Box-
Behnken Design was the technique adopted to design the experi-
ments using design expert 7.0 software, to study the parametric
influences on the betterment of KW. The ideal machining condi-
tions for response surface technique was chosen with reference
to the work done by Guojun Zhang et.al. [17].The input data were
collected from the experiments and fed to the RSM tool to arrive
the optimal outputs.

3. Response surface Methodology (RSM)

A numerical static strategy in RSM is embraced to make the sci-
entific model under regression and to optimize the problems. RSM
is a mix of test regression investigation and factual inductions [16].
The possibility of a reaction incorporates a dependent variable y
called the yield variable and autonomous info factors x1, x2,. . ...
4. Output responses measurement
The important response parameter of wire cut EDM is Kerf
width. It is the response parameter, which determines dimensional
accuracy of the workpiece. It is an indicator that shows the compli-
ance of the component in terms of dimensions that reflect in the

Fig. 2. Normal probability plot of residual for KW.

Please cite this article as: N. E. Arun Kumar, N. Sathishkumar, E. Raviraj et al., Influence of near dry wirecut electrical discharge machining parameters on
kerf width in Monel 400, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.05.471
4 N.E. Arun Kumar et al. / Materials Today: Proceedings xxx (xxxx) xxx

Table 4
Experimental Results.

Expt.run A:T ls)

ON( B:T ls)
OFF( C:Voltage(V) D:Wire Feed (mm/min) KW(mm)
1 1 6 50 6 0.262
2 1 6 55 8 0.337
3 1 6 60 10 0.289
4 1 8 50 8 0.324
5 1 8 55 10 0.338
6 1 8 60 6 0.311
7 1 10 50 10 0.345
8 1 10 55 6 0.342
9 1 10 60 8 0.329
10 3 6 50 6 0.308
11 3 6 55 8 0.314
12 3 6 60 10 0.309
13 3 8 50 8 0.326
14 3 8 55 10 0.339
15 3 8 60 6 0.331
16 3 10 50 10 0.318
17 3 10 55 6 0.335
18 3 10 60 8 0.326
19 5 6 50 6 0.317
20 5 6 55 8 0.322
21 5 6 60 10 0.311
22 5 8 50 8 0.323
23 5 8 55 10 0.321
24 5 8 60 6 0.326
25 5 10 50 10 0.313
26 5 10 55 6 0.315
27 5 10 60 8 0.307
28 1 10 55 6 0.329
29 3 10 60 8 0.321

Table 5 5. Results and discussion

Regression statistics for KW.

Statistics KW reduced Quadratic model 5.1. Regression statistics for KW

2
R 0.9484
Adj R2 0.9003 The philosophy behind WEDM is that the distinct electric arc
Pred R2 0.6402 produced between the wire and work material to be cut are pro-
Adeq Precision (AP) 14.401 duced under a controlled and a constant flow rate of the dielectric
medium. The surface of the workpiece gets eroded by the applica-
tion of electrical sparks which consequently produces high melting
temperature. The responses from the machining based on the
experimental design were observed and recorded. Kerf width was
calculated from the experimental output data. The significant fac-
design. Kerf width can directly be measured from the slot in the tor effects are derived from the output parameters KW from the
machined work piece by using VMS 2010F as represented in Fig. 2. models to determine the impact of input parameters, following
The vision sensing device incorporating an image processing the work done by amitava mandal et al. [20]. The tests for the sig-
system has been interfaced with the electronic measurement unit. nificance of the regression model and individual model coefficient
The table 4 shows the results of the different experimental runs were performed. The ANOVA test was performed to calculate the
carried out in this study [19]. coefficient of determination, F-Value, Prob. > F- Value, an index of

Table 6
ANOVA table for reduced quadratic KW.

Source Sum of Squares DOF Mean Square F Value p-value Prob > F Significance
Model 0.014 14 9.682E-004 19.70 <0.0001 Significant
A-T ON 4.672E-005 1 4.672E-005 0.95 0.3450
B-T OFF 3.500E-003 1 3.500E-003 71.21 <0.0001 Significant
C-Voltage 5.000E-007 1 5.000E-007 0.010 0.9210
D-WF 0.000 1 0.0 0.000 1.0000
AB 4.290E-003 1 4.290E-003 87.29 <0.0001 Significant
AC 1.322E-004 1 1.322E-004 2.69 0.1217
C2 0.000 1 0.0 0.000 1.0000
D2 4.622E-004 1 4.622E-004 9.41 0.0078
Residual 0.000 15 0.0
Lack of Fit 0.000 10 0.0
Pure Error 0.000 10 0.0
Cor Total 1.743E-004 5 1.743E-004
Model 4.516E-004 29

Please cite this article as: N. E. Arun Kumar, N. Sathishkumar, E. Raviraj et al., Influence of near dry wirecut electrical discharge machining parameters on
kerf width in Monel 400, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.05.471
 N.E. Arun Kumar et al. / Materials Today: Proceedings xxx (xxxx) xxx 5

Fig. 5. Response surface of KW versus Toff and Ton.

Fig. 3. Video Measuring System (VMS 2010F).

Fig. 6. Response surface of KW versus WF and Toff.

Fig. 4. Response surface of KW versus Toff and Ton.

variation used to measure the degree of fit. The regression model is

better to fit if R2 = 1. A reduced quadratic model for the KW was
given in Table 6 in accordance with the Adj R2 is usually used to
calculate a number of variations from the mean. The statistics
related to the response variable KW gives a satisfactory compar-
ison between the values of R2 which has been predicted and
adjusted, as shown in Table 5. The signal adequacy expected is
better since the ratio obtained is 14.401 which is more than the
desirable value of 4 [18].
The developed Regression model is critical, since the
F-estimation results providing the value of 19.70 provides the
justification. For this case, C and AB are critical influencing terms.
Values more noteworthy than 0.1000 demonstrates that the model
terms are significant (Table 6). Normal residuals plot of the Fig. 7. VMS image of low discharge energy with kerf (0.331 mm).

Please cite this article as: N. E. Arun Kumar, N. Sathishkumar, E. Raviraj et al., Influence of near dry wirecut electrical discharge machining parameters on
kerf width in Monel 400, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.05.471
6 N.E. Arun Kumar et al. / Materials Today: Proceedings xxx (xxxx) xxx

Fig. 6 portrays the interactive effect of WF and TOFF on kerf. The

Fig. 8. VMS image of low discharge energy with kerf (0.262 mm).
regression model to KW is shown in Fig. 3. It can be realistic that
the residuals follow the normal distribution and the assumptions
of normality are valid. It clearly shows that two outliers were pre-
sent among the 29 experimental runs. It is happened due to the
randomised distribution of residues by the effect of process param-
eter setting. All other experimental runs are within the range. A
mathematical model for KW has been developed as shown in
equation (1).
KW ¼ ð1:42619Þ þ ð0:061671  AÞ þ ð0:10162  BÞ
þ ð0:045684  C Þ  ð5:19298E  0:003  DÞ
 ð4:09375E  0:003  A  BÞ
 ð2:87500E  0:004  A  C Þ
 ð5:37500E  0:004  B  C Þ  ð2:05044E  0:003  A2Þ
 ð3:30044E  0:003  B2Þ  ð3:68070E  0:004  C2Þ
þ ð3:24561E  0:0042  D2Þ
kerf increases initially and after attaining the maximum value, kerf
decreases with the increase of WF and TOFF. The kerf width during
discharge increases when there is increase in energy per pulse and
decrease in the duration given for discharge there are changes in
Pulse off and kerf width. The former increases and the later
decreases. The frequent successive sparks is possible at greater
feed values as exemplified by Gupta et al. [22]. The sparks are
struck with much reduced time intervals at the time of erosion
process with lower setting values of wire feed rate, as experi-
mented by garg et al. [24].Minimum kerf (0.262 mm) is obtained
with the parametric combination minimum pulse-on (1 lm), min-
imum pulse-off (6 lm), lower wire feed (6 mm/min) and low volt-
age (50 V). At high discharge energy, the greater ionization effect
that happens because of the collision of the quantum of negatively
charged electrons with the neutral particles produces larger kerf
width (0.331 mm). The intensity of collision between the quantum
of electrons and ions in the cutting space makes the kerf to expand
(Fig. 7) [25].The accuracy is more in the image of the kerf width
(0.262 mm) with low discharge energy shown in Fig. 8 when com-
pared to high discharge energy, because the time taken to transfer
heat to the surrounding medium is less from the concentrated
small plasma zone as illustrated by Somashekar et al. [25].
5.3. ANN modeling
ANN methods are utilized for displaying the machining process
because of their capacity to rough the nonlinear and complex con-
nection between different data and yield. In the Artificial neural
network, networks are trained continuously to find the optimum
condition. The correlation coefficient value has to go nearer to
1.000, after the learning rate (ƞ) of 0.05 and the constant of
momentum (a) as 0.9 being taken to train the network of neurons.
A hyperbolic tangent of the function of hidden and output neurons
was taken for the investigation. The quantity of neurons and hid-
den layers of these neurons depends on the model to be subjected
Table 7
Comparison of Experimental and ANN results.

Expt. No. Expt. Results ANN Based Results Absolute Error in%

5.2. Response surface analysis 1 0.262 0.2640742 0.7917

2 0.337 0.3368585 0.0420
3 0.289 0.2861984 0.9694
Fig. 4 speaks to the consolidated impact of the on and off time 4 0.324 0.3255894 0.4905
on kerf (mm). From Fig. 3, it is conceivable to peruse that the kerf 5 0.338 0.3373371 0.1961
width increments with an expansion of pulse off time and the 6 0.311 0.3136607 0.8555
7 0.345 0.3435437 0.4221
increment of the TON. Due to the increase in the time duration
8 0.342 0.3410822 0.2684
for the supply voltage the uniform spark erupts with the propor- 9 0.329 0.3295494 0.1670
tional increase of the discharge energy, eventually decreasing the 10 0.308 0.3030106 1.6199
kerf width [21]. The reason for the attainment of Minimum kerf 11 0.314 0.3168051 0.8934
during on and off time lies in the lower setting level. This is due 12 0.309 0.3091159 0.0375
13 0.326 0.3274068 0.4315
to the impingement of spark on the workpiece for a lesser duration 14 0.339 0.3376414 0.4008
of time. It leads to the low release of spark energy. The low level of 15 0.331 0.3309352 0.0196
spark energy causes the small crater formation on the workpiece 16 0.318 0.3160389 0.6167
surface as reported by garg et al. [22]. Fig. 5 signifies the combina- 17 0.335 0.3350371 0.0111
18 0.326 0.3253499 0.1994
torial effect of voltage and Wire feed on kerf. From Fig. 5, unmis-
19 0.317 0.3141299 0.9054
takably the kerf width diminishes with the expansion in voltage 20 0.322 0.3231249 0.3494
at the lower setting level of wire feed. The higher spark energy pro- 21 0.311 0.3098263 0.3774
duced due to higher peak current releases debris in large volume. 22 0.323 0.3247606 0.5451
The debris formation is not advisable for machining since they 23 0.321 0.3192195 0.5547
24 0.326 0.3269603 0.2946
erode the tool material and consumes the spark energy in sparking 25 0.313 0.3137081 0.2262
the debris [21].The pinnacle esteem is achieved when wire feed 26 0.315 0.3170243 0.6427
and voltage are in the higher setting level. Because of increment 27 0.307 0.3089198 0.6254
in voltage, normal release crevice augmented bringing about 28 0.329 0.3299288 0.2823
29 0.321 0.3211631 0.0508
higher kerf [23].

Please cite this article as: N. E. Arun Kumar, N. Sathishkumar, E. Raviraj et al., Influence of near dry wirecut electrical discharge machining parameters on
kerf width in Monel 400, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.05.471
 N.E. Arun Kumar et al. / Materials Today: Proceedings xxx (xxxx) xxx
Fig. 9. Comparison of Expt. and ANN results.
Fig. 10. Correlation graph of experimental and ANN results for kerf width.
to the trials. These selections are to be made carefully before a start
of the neural network formation and the commencements of the
trials [13].
5.4. Genetic algorithm
The Genetic method is the development of maximal condition
where there is no refinement in the solution further. The logic hid-
den in this process of generating offspring chromosomes involves
the techniques like the crossover and mutation. Genetically alter-
ations made in the selected chromosomes refer to the mutation
process, whereas the crossover involves the reproduction of new
chromosomes by merging the two parent chromosomes. The chro-
mosomes develop amid a few emphases or eras. The chromosomes
are then surveyed by using a particular wellbeing criteria and the
best ones are kept while the others are discarded. This strategy is
reiterated until the end model is met or one chromosome has the
best health and in this way is taken as the best response for the
issue. In this manner, a perfect setting for the Near-dry WEDM
methodology was perceived using the inherited computation of
GA.
5.5. ANN prediction of kerf width
Preparing of the neural system model was performed on 29 test
information focuses as clarified in above segment. Table 7 demon-
strates the exploratory and anticipated qualities for kerf width for
the 29 preparing sets. Examination of Experimental and ANN
results appear in Fig. 9.For each info mix, the anticipated estima-
Please cite this article as: N. E. Arun Kumar, N. Sathishkumar, E. Raviraj et al., Influence of near dry wirecut electrical discharge machining parameters on
kerf width in Monel 400, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.05.471
7
tions of reactions were contrasted and the individual exploratory
qualities and the rate total mistake is processed as takes after:
Y jexpt  Y jpred
AbsoluteErrorð%Þ ¼ ð2Þ
Y jexpt
Where, Yjexpt and Yjpred is the experimental value and predicted
value of the KW respectively for the jth trial.
Table 7 demonstrates the supreme rate mistake for each info
mix. As per this table, the most extreme error is 1.6199. In this
manner it can be reasoned that forecasts are in great concurrence
with the test comes about on the grounds that the most extreme
rate total mistake of the anticipated an incentive regarding the ten-
tatively watched an incentive for KW is nearer to experimental
value. The connection between’s the objectives and the anticipated
estimations of ANN for the testing examples are spoken to as
scramble graphs, as shown in Fig. 10 for the kerf. The estimation
of relationship coefficient is 0.9939. From the measurable perspec-
tive, the nearer estimation of correlation coefficients (R) to 1, the
all the more intense the system in corresponding the information
space to yield space.
6. Conclusion
In this paper, the effects of control factors on response variable,
kerf width is carried out using Response Surface Methodology fol-
lowed by comprehensive statistical analysis (ANOVA) to identify
the predominant factors influencing kerf width.
Following lines exemplify the concluding points from the study
 The input parameters for the experiments designed based on
RSM were optimized to machine Monel 400 under a very close
dryness condition dictated to the dielectric medium and fur-
ther, Mathematical models for KW were developed, for further
analysis.
 Pulse-off plays a vital role in attaining minimum kerf width.
Minimum spark gap (0.262 mm) is obtained with the paramet-
ric combination minimum pulse-on (1 lm), minimum pulse-off
(6 lm), lower wire feed (6 mm/min) and low voltage (50 V).
 ANN modeling predicts values that are in resemblance to the
values obtained from experimental observation.
 The Value of error calculated by ANN predictions and the error
evaluated through experimentation has a maximum difference
of only 1.6199 in absolute terms.
 The Experimental results appreciate the accuracy with which
the ANN predicts the value of kerf. The kerf value is not at all
high and it lies pretty much closer to ANN Prediction rather it
is not higher.
8 N.E. Arun Kumar et al. / Materials Today: Proceedings xxx (xxxx) xxx
CRediT authorship contribution statement
N.E. Arun Kumar: Conceptualization, Methodology, Investiga-
tion, Writing - original draft, Writing - review & editing. N.
Sathishkumar: Resources, Validation, Formal analysis, Supervi-
sion. E. Raviraj: Visualization, Data curation. M. Pathri Naraya-
nan: Project administration. Roderik Eugene: Investigation.
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
References
[1] Masonori Kunieda, Chika Furudate, High Precision Finish Cutting by Dry
WEDM, Annals of the CIRP 50 (1) (2001) 121–124.
[2] C.C. Kao, Jia Tao, Albert J. Shih, Near dry electrical discharge machining, Int. J.
Mach. Tools Manuf 47 (2007) 2273–2281.
[3] S. Joshi, P. Govindan, A. Malshe, K. Rajurkar, Experimental characterization of
dry EDM performed in a pulsating magnetic field, CIRP Annals – Manufacturing
Technology 60 (2011) 239–242.
[4] Yu. Zhan Bo, Takahashi Jun, Kunieda Masonri, Dry electrical discharge
machining of cemented carbide, J. Mater. Process. Technol. 149 (2004) 353–
357.
[5] Sourabh K. Saha, S.K. Choudhury, Experimental investigation and empirical
modeling of the dry electric discharge machining process, Int. J. Mach. Tools
Manuf 49 (2009) 297–308.
[6] S. Boopathi, K. Sivakumar, Experimental Comparative Study of Near-Dry Wire-
Cut Electrical Discharge machining (WEDM), European Journal of Scientific
Research 75 (4) (2012) 472–481.
[7] M.M. Dhobe, I.K. Chode, C.L. Gogte, Optimization of Wire Electro Discharge
Machining Parameters for improving Surface Finish of Cryo-Treated Steel
Using DOE, Mater. Manuf. Processes 29 (11–12) (2014) 1381–1386.
[8] Wuyi Ming, Zhen Zhang, Yu. Guojun Zhang, JianwenGuo Huang, Yuan Chen,
Multi-Objective Optimization of 3D-Surface Topography of Machining YG15 in
WEDM, Mater. Manuf. Processes 29 (5) (2014) 514–525.
[9] Ah.ad. Gholipoor, Hamid Baseri, Mohammad Reza Shabgard, Investigation of
near dry EDM compared with wet and dry EDM processes, J. Mech. Sci.
Technol. 29 (5) (2015) 2213–2218.
[10] S. Di, X. Chu, D. Wei, Z. Wang, G. Chi, Y. Liu, Analysis of kerf width in micro-
WEDM, Int. J. Mach. Tools Manuf 49 (2009) 788–792.
Please cite this article as: N. E. Arun Kumar, N. Sathishkumar, E. Raviraj et al., Influence of near dry wirecut electrical discharge machining parameters on
kerf width in Monel 400, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.05.471
[11] Wang Yuanyuan, Machining precision analysis on Micro-WEDM, Harbin
Institute of Technology, 2007.
[12] N. Tosun, C. Cogun, G. Tosun, A study on kerf and material removal rate in wire
electrical discharge machining based on Taguchi method, Journal of Material
Processing Technology 152 (2004) 316–322.
[13] Pragya Shandilya Jain P K and Jain.N K, RSM and ANN Modeling Approaches
For Predicting Average Cutting Speed During WEDM of SiCp/6061 Al MMC
Procedia Engineering 64 2013 767 774
[14] Reza Kashiry Fard, Reza Azar Afza, Reza Teimouri, Experimental investigation,
intelligent modeling and multi-characteristics optimization of dry WEDM
process of Al–SiC metal matrix composite, J. Manuf. Processes 15 (2013) 483–
494.
[15] Neeraj Sharma Renu Sharma AjitSingha Deepak. Modelling the WEDM Process
Parameters for Cryogenic Treated D-2 Tool Steel by integrated RSM and GA
Procedia Engineering 97 2014 1609 1617
[16] Marimuthu K JeyakumarS, T. Ramachandran, Optimization of machining
parameters of Al6061 composite to minimize the surface roughness –
modelling using RSM and ANN, Indian journal of Engineering & Materials
Sciences 22 (2015) 29–37.
[17] D.C. Montgomery, Design and analysis of experiments, 6th edition., Wiley,
New York, 2005.
[18] Guojun Zhang Zhen Zhang Jianwen Guo Wuyi Ming Yu. Mingzhen Li Huang.
Modeling and Optimization of Medium-Speed WEDM Process Parameters for
machining SKD11 Materials and Manufacturing Processes 28 10 2013 1124
1132
[19] N.E. Arun Kumar, Babu A. Suresh, Influence of Input Parameters on Near Dry
WEDM of Monel Alloy, Mater. Manuf. Processes 2017 (2017), https://doi.org/
10.1080/10426914.2017.1279297.
[20] Amitava Mandal, Amit Rai Dixit, Alok Kumar Das, Niladri Mandal, Modeling
and Optimization of machining C-263 Superalloy using Multicut Strategy in
WEDM, Mater. Manuf. Processes 31 (7) (2016) 860–868.
[21] Palani Sivaprakasam, P. Hariharan, S. Gowri, Optimization of Micro-WEDM
Process of Aluminium Matrix Composites (A413–B4C): A Response Surface
Approach, Mater. Manuf. Processes 31 (7) (2013) 1340–1347.
[22] Gupta Pardeep Rajesh Khanna Rahul Dev Gupta Neeraj Sharma. Effect of
Process Parameters on Kerf Width in WEDM for HSLA Using Response Surface
Methodology. Journal of Engineering and Technology 2 1 2017 1 6
[23] Sanjeev Kr, Alakesh Manna Garg, Ajai Jain, Multi-Objective Optimization of
Wire Electro- Discharge Machining of Al/ZrO2 (P) Metal Matrix Composite by
Response Surface Methodology, Journal of Machining and Forming
Technologies 4 (3–4) (2012) 225–240.
[24] Sanjeev Kumar Garg, Alakesh Manna, Ajai Jain.(2015) Experimental
Investigation of Spark Gap and Material Removal Rate of Al/ZrO2(P) –MMC
Machined with Wire EDM Journal of Brazilian Society of Mechanical Sciences
& Engineering , doi: 10.1007/s40430-015-0394-5,2015.
[25] K.P. Somashekhar, N. Ramachandran, J. Mathew, Material removal
characteristics of microslot (kerf) geometry in l-WEDM of aluminium,
International Journal of Advanced Manufucturing Technology 51 (2010)
611–626.