MODELING AND ANALYSIS OF SURFACE ROUGHNESS AND WHITE LATER THICKNESS IN WIRE-ELECTRIC DISCHARGE TURNING PROCESS THROUGH RESPONSE SURFACE METHODOLOGY

https://iaeme.com/Home/journal/IJARET 2584 editor@iaeme.com
International Journal of Advanced Research in Engineering and Technology (IJARET)
Volume 11, Issue 11, November 2020, pp. 2584-2597 Article ID: IJARET_11_11_257
Available online at https://iaeme.com/Home/issue/IJARET?Volume=11&Issue=11
ISSN Print: 0976-6480 and ISSN Online: 0976-6499
© IAEME Publication Scopus Indexed
MODELING AND ANALYSIS OF SURFACE
ROUGHNESS AND WHITE LATER THICKNESS
IN WIRE-ELECTRIC DISCHARGE TURNING
PROCESS THROUGH RESPONSE SURFACE
METHODOLOGY
Baki Naresh
Research Scholar, Department of Mechanical Engineering,
Jawaharlal Nehru Technological University, Hyderabad, Telangana, India.
A Chennakesava Reddy
Professor, Department of Mechanical Engineering,
Jawaharlal Nehru Technological University, Hyderabad, Kukatpally, Telangana, India.
ABSTRACT
White layer thickness (WLT) formed and surface roughness in wire electric
discharge turning (WEDT) of tungsten carbide composite has been made to model
through response surface methodology (RSM). A Taguchi’s standard Design of
experiments involving five input variables with three levels has been employed to
establish a mathematical model between input parameters and responses. Percentage
of cobalt content, spindle speed, Pulse on-time, wire feed and pulse off-time were
changed during the experimental tests based on the Taguchi’s orthogonal array L27
(3^13). Analysis of variance (ANOVA) revealed that the mathematical models obtained
can adequately describe performance within the parameters of the factors considered.
There was a good agreement between the experimental and predicted values in this
study.
Key words: WEDT, Response surface methodology, white layer thickness.
Cite this Article: Baki Naresh and A Chennakesava Reddy, Modeling and Analysis of
Surface Roughness and White Later Thickness in Wire-Electric Discharge Turning
Process through Response Surface Methodology, International Journal of Advanced
Research in Engineering and Technology (IJARET), 11(11), 2020, pp. 2584-2597.
https://iaeme.com/Home/issue/IJARET?Volume=11&Issue=11
1. INTRODUCTION
In the industrial sector, tungsten carbides are widely used due to their excellent properties, such
as high hardness, high toughness, and high strength; these materials are frequently used in the
production of cutting tools, drawing or extrusion dies, water jet nozzles, and other wear-

Modeling and Analysis of Surface Roughness and White Later Thickness in Wire-Electric
Discharge Turning Process through Response Surface Methodology
resistant parts, etc. [1,2]. In spite of the excellent properties of tungsten carbides, they are
difficult to process using traditional cutting methods. In WEDM, the material from the
workpiece is melted and vaporized by means of heat erosion. Since materials with electric
conduction can be machined by EDM, regardless of their mechanical properties, it is
appropriate to use EDM to finish tungsten carbide. In recent research, it has been shown that
surface integrity is not only a key factor in tools made from tool steel, but it is also a factor in
tools made from cemented carbides. In fact, since breakdowns are often caused by existing rim
zone defects, and high investment costs have resulted in a lot of research in order to find ways
to increase the tool's life, particularly for molds and forming dies used in cold forming. It has
been found that defects can either occur during the primary production process or during the
secondary tool manufacturing process [3, 4, 5]. As a result of the high hardness of cemented
carbides, it has become almost uneconomical to machine them using conventional cutting
technologies, although there has been a lot of progress in this area [6,7]. In the tool making
industry, electrical discharge machining (EDM) has become one of the most important
technologies. The major advantage of this manufacturing technology is the inherent ability to
machine materials independent of their mechanical properties as well as the high rate of removal
from the material being machined. It is for this reason, as well as the need for high precision,
that electrical discharge machining has become one of the most important technologies used to
manufacture dies and molds [8,9]. In today's EDM industry, improving surface integrity by
reducing surface defects and minimizing surface fine finishing processes is the main challenge.
Wire electric discharge turning (WEDT) is one of the most important variant of wire electric
discharge machining (WEDM). The working principle of WEDT is same as WEDM. WEDT
utilizes rapid, repetitive spark discharges from direct current supply between the workpiece and
the tool submerged into a dielectric liquid [10]. The work material melts or evaporates during
each discharge, resulting in intense heat. When plasma flushing efficiency decreases, some
molten material may be sucked and severely boiled and removed from the molten crater due to
intense suction [11].
WEDT considers the surface quality to be the most important performance parameter. The
surface quality is expressed through the surface roughness and the WLT. It is evident that WLT
plays a vital role in providing operational characteristics for the part (e.g. fatigue, corrosion,
creep life, etc. fracture resistance, surface friction and coating ability). A white layer represents
a layer that has been heated to melting point but has not yet been sufficiently heated to be
ejected into the gap between the wire electrode and the workpiece and ejected. During the
flushing process, the un-ejected molten metal is rapidly cooled by the dielectric fluid and
resolidified in the cavity, resulting in a change in the metallurgical structure and characteristics
of the layer. Some of the particles that have been expelled have solidified and have been
redeposited on the surface before they are flushed out of the gap. There is a dense infiltration
of carbon in the white layer, which makes its structure distinct from that of the base material.
As the electrode and dielectric fluid melt, the hydrocarbons of the electrode and dielectric fluid
leak into the white layer, enriching carbon during the WEDM process.
According to Levi and Maggi [12], WEDM produces a thin heat-affected zone thickness
varying from 1 µm at 5 µJ to 25 µm at high spark energy. Moreover, the white layer will be
produced by annealing the zone under the machined surface. During WEDM machining, the
WLT is proportional to the discharge energy. 50 μm for finish machining to around 200 μm for
high cutting speed [13,14]. For low WLT values, the part must be machined more than once.
As a result, the desired WLT is usually specified, and the appropriate processes are selected to
reach the desired quality [13]. Prediction systems or post-process inspections can be used to
monitor the actual WLT. However, post-process inspection can't prevent parts from being
processed before defective batches are discovered. Further, WEDM is a complex process

Baki Naresh and A Chennakesava Reddy
controlled by many variables. Variations in a process parameter can impact WLT. In order to
determine the best machining strategy, it is necessary to identify the various factors that affect
the WEDM process and find a variety of approaches to obtain the best possible machining
conditions and performance [15,16]. Finally, the WLT prediction system can be used to
determine the best machining conditions as well as the indirect calculation of the WLT using
the results from the prediction [17,18].
In order to achieve an efficient machining process, a predictive model between the input
WEDM parameters and the output performance characteristics must be available. There is a
paucity of research that has been done on the modeling and analysis of the WEDT process with
a focus on white layer thickness (WLT) and the surface roughness (Ra) as the machining
performance parameters. According to this study, mathematical models have been developed
using response surface methodology in order to predict the Ra and WLT. For the purpose of
ensuring the validity of a model, the analysis of variance (ANOVA) is used.
2. RESPONSE SURFACE METHODOLOGY
Using the response surface methodology, we can model the response variables as well as
optimize models that include quantitative independent variables in order to model and optimize
the response variable models [19]. The modeling of the desired response to several independent
input variables can be obtained by utilizing a design of experiments and applying regression
analysis to the modeling of the response to several independent input variables. A response
surface for a set of measurable variables can be expressed as follows [20]:
𝑦 = 𝑓(𝑥1, 𝑥2, 𝑥3 … … . 𝑥𝑘) + 𝜀 (1)
Where y is the answer of the system, f is the response function (or response surface),
𝑥1, 𝑥2, 𝑥3, … … . . 𝑥𝑘 are the independent variables and 𝜀 is the fitting error. In this study, for five
variables under consideration, a second order polynomial regression model, which is called
quadratic model, is proposed. The quadratic model of f can be written as follows [21]:
𝑓 = 𝑎𝑜 + ∑ 𝑎𝑖
3
𝑖=1 𝑥𝑖 + ∑ 𝑎𝑖𝑥𝑖
3
𝑖=1 + ∑ 𝑎𝑖𝑗𝑥𝑖𝑥𝑗
3
𝑖<𝑗 + ε (2)
Statistical software, Design Expert 13.0 can be used to determine the coefficients of
regression models from the experimental results.
3. EXPERIMENTAL PROCEDURE
Experiments are planed and executed based on the Taguchi’s Design of Experiments. Among
the most important DOE's, Taguchi's method provides a simple, systematic, and efficient
approach to determining the optimal process parameters. Using Taguchi's method, a large
number of control factors can be selected with a reduced number of experiments using an
orthogonal array of DOEs. The Taguchi L27 orthogonal array was selected because of the five
machining variables with three levels as shown in Table 1. The machining parameters spindle
speed (x2), pulse-on time (x3), wire feed (x4) and pulse-off time (x5) were chosen in this study
to investigate the effect on machining performance including WLT. The other machining
parameters were kept constant as a fixed value during experiments as recommended by the
machine maker to optimize the process as shown in Table 2.

Table 1 Control variables and their levels
S.
No.
Control
variable
Notation Units
Levels
-1 0 1
1
Cobalt
percentage
A Wt.% Co 8 10 12
2 Rotational speed B rpm 10 20 30
3 Pulse-on time C μs 7 10 13
4 Wire Feed D m/min 6 9 12
5 Pulse-off time E μs 30 35 40
Table 2 Machining Conditions
Wire: uncoated brass wire of diameter 0.25 mm
Work piece material: tungsten carbide
Work piece dimensions: 10 mm dia. × 25mm L, 10.2 dia. × 25mm L
Length of cut: 10 mm
Angle of cut: vertical
Dielectric fluid: distilled water
Wire tension: 9 Kg
Servo Voltage: 30 V
Arc ON: 10 μs
Arc OFF: 35 μs
Length of cut: 10 mm
Depth of cut: 0.15mm
Table 3 Experimental results for surface roughness and white layer thickness values based on
Taguchi’s L27 array.
Exp.
No.
Control variable Ra
(µm)
WLT
(µm)
A B C D E
1 -1 -1 -1 -1 -1 2.77 9.6
2 -1 -1 0 0 0 3.87 9.3
3 -1 -1 1 1 1 5.20 46.9
4 -1 0 -1 0 0 3.02 31.4
5 -1 0 0 1 1 3.39 38.5
6 -1 0 1 -1 -1 3.86 35.7
7 -1 1 -1 1 1 2.34 27.4
8 -1 1 0 -1 -1 1.83 15.9
9 -1 1 1 0 0 3.91 25.8
10 0 -1 -1 0 1 4.42 40.4
11 0 -1 0 1 -1 5.19 28.7
12 0 -1 1 -1 0 4.02 31.0
13 0 0 -1 1 -1 4.34 25.8
14 0 0 0 -1 0 2.19 11.2
15 0 0 1 0 1 5.40 39.7
16 0 1 -1 -1 0 1.14 15.9
17 0 1 0 0 1 3.48 13.3
18 0 1 1 1 -1 5.23 18.2
19 1 -1 -1 1 0 5.28 42.7
20 1 -1 0 -1 1 3.53 23.5
21 1 -1 1 0 -1 5.21 30.1
22 1 0 -1 -1 1 2.68 19.5
23 1 0 0 0 -1 3.38 7.9
24 1 0 1 1 0 6.37 40.3
25 1 1 -1 0 -1 2.33 21.5
26 1 1 0 1 0 4.34 37.1
27 1 1 1 1 1 3.55 34.3

4. EXPERIMENTAL SET-UP
The specimens that are used in this investigation was a tungsten carbide composite rod with a
diameter of ϕ10.2 mm and ϕ10.5 mm for a length of 35 mm. The Wire electric discharge turning
(WEDT) experiments have been carried out on CNC wire-cut machine model JOEMARS WTT
655 (Taiwan make) using a brass wire with a diameter of ϕ0.25 mm. An in-house fabricated
rotary axis spindle is attached to WEDM which can be submergible in the dielectric during
machining. Figure 1 shows the experimental setup used in the present work.
Figure 1 Experimental setup on WEDM with an attached rotary spindle system for turning
The surface roughness of the machined surface after each experiment was measured using
a MITUTOYO make Talysurf model, with a cut off length of 0.8mm (according to DIN EN
ISO 3274:1998). Fig. 2 show the measurement of surface roughness. For every experiment, the
surface roughness on the machine workpiece surface at 6 different locations were taken and
was recorded their average as the final response value.
Figure 2 Measurement of surface roughness of the machined workpieces

4.1. Sample Preparation
Figure 3 Sample preparation
After machining, cross-sections were polished as shown in Fig. 3 and etched with
Murakami’s etchant (KOH (10g/100ml water) + K3[Fe(CN)6] (10g/100ml water)) for WC-Co
to get a better image of the different WC-Co phases. As a next step, the samples were examined
using an SEM with secondary electrons (SE) and backscattered electrons (BSE) to assess white
layers, cracks, and pores. The purpose of such images is to determine the thickness of the white
layer, to measure the crack length, and to identify pores in the rim zone. The spindle speed,
pulse-on time, wire feed and pulse-off time were independent variables studied to predict Ra
and WLT responses.
5. RESULTS AND DISCUSSION
The Cobalt percentage (wt.%Co), spindle speed, Pulse-on time (Pon), Wire feed (Wfeed) and
pulse-off time (Poff) were the independent variables studied to predict responses (Ra and
WLT). The independent variables and their levels used in this study are shown in Table 1. Eq.
2 can be rewritten according to five variables in the coded form as:
𝑦 = 𝑎0 + 𝑎1𝑥1 + 𝑎2𝑥2 + 𝑎3𝑥3 + 𝑎3𝑥3 + 𝑎4𝑥4 + 𝑎5𝑥5 + 𝑎11𝑥1
2
+ 𝑎22𝑥2
2
+ 𝑎33𝑥3
2
+ 𝑎44𝑥4
2
+
𝑎55𝑥5
2
+ 𝑎12𝑥1𝑥2 + 𝑎13𝑥1𝑥3 + 𝑎14𝑥1𝑥4 + 𝑎15𝑥1𝑥5 + 𝑎23𝑥2𝑥3 + 𝑎24𝑥2𝑥4 + 𝑎25𝑥2𝑥5 + 𝑎34𝑥3𝑥4 +
𝑎35𝑥3𝑥5 + 𝑎45𝑥4𝑥5 (3)
𝑅𝑎 = 4.07 + 0.2344 ∗ 𝐴 − 0.2456 ∗ 𝐵 + 0.7178 ∗ 𝐶 − 0.3967 ∗ 𝐷 + 0.8444 ∗ 𝐸 + 0.0600 ∗ 𝐴 ∗
𝐵 − 0.0808 ∗ 𝐴𝐶 − 0.0650 ∗ 𝐵𝐶 − 0.1267 ∗ 𝐵𝐷 + 0.1050 ∗ 𝐶𝐷 − 0.2675 ∗ 𝐶𝐸 − 0.1600 ∗ 𝐵2
+
0.3600 ∗ 𝐶2
(4)
𝑊𝐿𝑇 = 22.88 − 0.7201 ∗ 𝐴 − 6.77 ∗ 𝐵 + 0.9673 ∗ 𝐶 + 5.82 ∗ 𝐷 + 4.30 ∗ 𝐸 + 4.40 ∗ 𝐴𝐷 −
2.69 ∗ 𝐵𝐸 + 3.53 ∗ 𝐷𝐸 + 6.33 ∗ 𝐶2
(5)
5.1. Checking the Adequacy of the Developed Models
The adequacies of the models are checked by using the analysis of variance (ANOVA)
technique. The ANOVA tables for Ra and WLT are presented in Tables 5 and 6, respectively.
The p-values in Table 5 and Table 6 for models are less than 0.05, which indicate model terms
are significant. In case of surface roughness model, A, B, C, D, E, CE, C² are significant model
terms. Main effects plot as shown in Fig. 8 (a) reveals the most influencing parameters on
surface roughness are Pulse-off time followed by pulse-on time, wire feed, spindle speed and
wt. %Co. The Interaction of Pulse-on time and Pulse off time have significant effect on surface
roughness.

In the case of White layer thickness, B, D, E, AD, BE, DE, C^2 are significant model terms.
Table 4 ANOVA table for response surface Reduced Quadratic model (response: surface roughness)
Sum of Squares df Mean Square F-value p-value
Model 29.01 13 2.23 46.70 < 0.0001 significant
A-wt. %Co 0.9894 1 0.9894 20.71 0.0005
B-Speed 1.09 1 1.09 22.71 0.0004
C-Pon 9.27 1 9.27 194.08 < 0.0001
D-Wfeed 2.83 1 2.83 59.27 < 0.0001
E-Poff 12.84 1 12.84 268.63 < 0.0001
A*B 0.0289 1 0.0289 0.6054 0.4504
A*C 0.0731 1 0.0731 1.53 0.2382
B*C 0.0507 1 0.0507 1.06 0.3218
B*D 0.1641 1 0.1641 3.43 0.0867
C*D 0.1210 1 0.1210 2.53 0.1355
C*E 0.5725 1 0.5725 11.98 0.0042
B² 0.1536 1 0.1536 3.21 0.0963
C² 0.7776 1 0.7776 16.27 0.0014
Residual 0.6212 13 0.0478
Cor Total 29.63 26
Std. Dev. 0.2186 R² 0.9790
Mean 4.20 Adjusted R² 0.9581
C.V. % 5.20 Predicted R² 0.9313
Adeq Precision 24.0625
Table 5 ANOVA table for response surface Reduced Quadratic model (response: white layer
thickness)
Source Sum of Squares df Mean Square F-value p-value
Model 2117.75 9 235.31 13.38 < 0.0001 significant
A-wt.%Co 5.94 1 5.94 0.3379 0.5687
B-Speed 781.91 1 781.91 44.45 < 0.0001
C-Pon 15.98 1 15.98 0.9086 0.3538
D- Wfeed 558.89 1 558.89 31.78 < 0.0001
E-Poff 212.08 1 212.08 12.06 0.0029
A*D 125.67 1 125.67 7.14 0.0161
B*E 80.38 1 80.38 4.57 0.0474
D*E 80.55 1 80.55 4.58 0.0471
C² 236.47 1 236.47 13.44 0.0019
Residual 299.01 17 17.59
Cor Total 2416.76 26
Std. Dev. 4.19 R² 0.8763
Mean 28.12 Adjusted R² 0.8108
C.V. % 14.92 Predicted R² 0.6770
Adeq Precision 13.2953
Further, using the models, we plotted the experimental and predicted data in Fig. 4 and Fig.
6 for Ra and WLT, respectively. These figures and ANOVA analysis for Ra and WLT indicated
that the models Eqs. (4) and (5) were highly significant and adequate to represent the actual
relationship between the variables and responses, with very small P values (<0.05) and high
values of coefficient of determination (R2
=0.07 for Ra and R2
= 0.87 for WLT).
In addition, the developed response surface models for Ra and WLT have been checked by
using residual analysis. The residual plots for response parameters Ra and WLT are shown in

Fig. 5 (a-c) and Fig. 7 (a-c) respectively. In normal probability plots, the data are distributed
approximately in a straight line, which show a good correlation
Table 7 Observed and predicted values of surface roughness and white layer thickness
Run
Factors Ra (μm) WLT (μm)
A B C D E Observed Predicted Observed Predicted
1 -1 -1 -1 -1 -1 2.77 2.74 29.6 30.85
2 -1 -1 0 0 0 3.83 3.98 26.4 30.36
3 -1 -1 1 1 1 5.72 5.62 46.9 49.60
4 -1 0 -1 0 0 3.24 3.40 31.4 28.96
5 -1 0 0 1 1 4.25 4.28 38.5 32.84
6 -1 0 1 -1 -1 4.68 4.71 32.7 28.71
7 -1 1 -1 1 1 3.41 3.48 27.4 28.75
8 -1 1 0 -1 -1 2.95 3.05 16.9 17.33
9 -1 1 1 0 0 4.53 4.46 25.8 24.13
10 0 -1 -1 0 1 4.91 4.85 40.4 42.00
11 0 -1 0 1 -1 2.95 3.04 28.7 24.94
12 0 -1 1 -1 0 5.62 5.46 31.1 31.13
13 0 0 -1 1 -1 2.36 2.10 25.8 26.23
14 0 0 0 -1 0 4.85 4.47 12.9 17.06
15 0 0 1 0 1 5.36 5.73 39.7 34.48
16 0 1 -1 -1 0 4.11 4.00 15.9 15.66
17 0 1 0 0 1 4.73 4.51 12.5 17.72
18 0 1 1 1 -1 3.65 3.68 18.2 24.09
19 1 -1 -1 1 0 3.63 3.61 42.7 44.51
20 1 -1 0 -1 1 5.35 5.45 23.5 22.17
21 1 -1 1 0 -1 4.79 4.82 30.1 29.23
22 1 0 -1 -1 1 5.56 5.64 19.5 18.08
23 1 0 0 0 -1 3.32 3.46 13.9 17.86
24 1 0 1 1 0 5.18 5.01 40.3 39.68
25 1 1 -1 0 -1 2.47 2.64 21.5 19.15
26 1 1 0 1 0 3.45 3.44 32.6 25.61
27 1 1 1 1 1 5.85 5.89 34.3 38.05
between experimental and predicted values for both responses as shown in Fig. 5(a) and 7
(a). Fig. 5 (b) and 7(b) indicate the residual versus predicted values, which show only the
minimal variation between observed and fitted values. Fig 5(d) and 7(d) show the residuals
calculated against the order of experimentation. As a whole analysis of residual plots for both
responses, the models do not reveal inadequacy.
Figure 4 Comparison of experimental and predicted surface roughness values

5.2. Effect of Variables on Responses
Figure 8 shows the plots of the main effects of five variables on Ra and WLT for each of the
five variables. Using this figure, it would be possible to analyze data in a designed experiment
with respect to the important factors, where the factors are multi-level and have two or more
levels of significance each. Based on an approximate measurement of graphs, namely the
difference between two points for each factor, it was found that pulse-off time is the most
significant factor when it comes to Ra and WLT, while spindle speed seems to be the least
important factor. As a result of a pulse-off time, there is a decrease in the rates of heat energy,
which is subjected to both electrodes by the electricity, as well as a decrease in the rates of
melting and evaporation due to the decreased rate of heat energy. Consequently, as the pulse
off time increases, less heat is transferred into the workpiece, and the dielectric is unable to
clear away the molten material, so the molten material accumulates on top of the parent
material, causing it to become hotter. In terms of the volume of the molten material, the
thickness of the white layer varies depending on the amount of the molten material that is
resolidified during the pulse off-time.
a. Normal probability plot of the residuals
b. Residuals versus the fitted values c. Residuals versus the order of the data
Figure 5 Plot of residuals for response surface roughness

Figure 6 Comparison of experimental and
predicted white layer thickness values
a. Normal probability plot of the residuals
b. Residuals versus the fitted values c. Residuals versus the order of the data
Figure 7 Plot of residuals for white layer thickness
Three-dimensional (3D) response surface plots were also formed based on the quadratic
model to assess the change of response surface as shown in Figs. 9 and 10. The relationship
between the variables and responses can be also further understood by these plots. The effects
of pulse on time and pulse off time and cobalt %, and pulse on time and wire feed on responses,
while keeping the other parameter at center level, are shown in Figs. 9 and 10
a) Surface roughness

b. White layer thickness
Figure 8 of main effects on responses
a. Effect of pulse-on time and pulse-off time on surface
roughness
b. Effect of percentage of cobalt and pulse-on time on
surface roughness
c. Effect of spindle speed and pulse-on time on surface roughness
Figure 9 3D response surface plots showing the effects of two variables on surface roughness.

a. Effect of wt.%Co and wire feed rate on white layer
thickness
b. Effect of spindle speed and pulse-off time on white
layer thickness
c. Effect of wire feed and pulse-off time on white layer thickness
Figure 10 3D response surface plots showing the effects of two variables on white layer thickness.
Fig. 10 (a) shows %Co has significant effect on WLT at lower wire feed rate and Fig. 10
(b) Pulse off time has significant effect combined with spindle speed. Fig.10 (c) reveals that at
lower levels of pulse-off and wire feed will lead to lower WLT formation.
6. CONCLUSIONS
The purpose of this study was to model and analyze the Ra and WLT during the wire electric
discharge turning process via response surface methodology (RSM). A Taguchi's orthogonal
array L27 in RSM is made up of five variables, in order to compute the training data. Wt. This
experimental study has been carried out utilizing %Cobalt, spindle speed, pulse-on time, wire
feed, and pulse-off time as variables in order to conduct the experiment. In order to analyze the
data, an analysis of variance (ANOVA) was used. In summary, we can draw the following
conclusions based on the main features of the study:
• The predicted values of the surface roughness and the WLT are reasonably close to the
experimental values, with the R2 for surface roughness being 0.97 and the R2 for WLT
being 0.87.
• Pulse-off time was found to be the most important factor affecting the Ra, while wire
feed was found to be the most influential factor affecting the WLT, while weight.%Co
had no significant effect on either of the responses.

• There is a greater understanding of individual variable effects and interactions between
variables if we use the main effect and 3D response surface plots for each variable.
• There is no doubt that the RSM model can be successfully used to model some
machining parameters for Ra and WLT, and it is a very economical method of obtaining
information on any system with the least number of experiments necessary.
REFERENCES
[1] M’Saoubi R, Axinte D, Soo SL et al (2015) High performance cutting of advanced aerospace
alloys and composite materials. CIRP Ann 64(2):557–580.
https://doi.org/10.1016/j.cirp.2015.05.002
[2] Andreas K (2015) Einfluss der Oberflächenbeschaffenheit auf das Werkzeugeinsatzverhalten
beim Kaltfließpressen. Dissertation, Friedrich-Alexander-Universität Erlangen-Nürnberg
[3] Andreas K (2015) Einfluss der Oberflächenbeschaffenheit auf das Werkzeugeinsatzverhalten
beim Kaltfließpressen. Dissertation, Friedrich-Alexander-Universität Erlangen-Nürnberg
[4] Llanes L (2010) Surface modification approaches for enhancing fracture and fatigue resistance
of electrical discharge machined hardmetals: manuscript refereed by Dr Leonid Frayman.
General Carbide Corporation, Barcelona
[5] Behrens B-A, Yilkiran T (2012) Influence of the fabrication method on the wear resistance of
hot forging dies. Prod Eng Res Dev 6(3):267–276. https://doi.org/10.1007/s11740-012-0373-y
[6] Ottersbach M, Zhao W (2016) Experimental investigations on the machinability of tungsten
carbides in orthogonal cutting with diamond-coated tools. Procedia CIRP 46:416–419.
https://doi. org/10.1016/j.procir.2016.04.008
[7] Hintze W, Steinbach S, Susemihl C et al (2018) HPC-milling of WC–Co cemented carbides
with PCD. Int J Refract Met Hard Mater 72:126–134.
https://doi.org/10.1016/j.ijrmhm.2017.12.019
[8] Jahan MP, Rahman M, Wong YS (2011) A review on the conventional and micro-
electrodischarge machining of tungsten carbide. Int J Mach Tools Manuf 51(12):837–858.
https://doi. org/10.1016/j.ijmachtools.2011.08.016
[9] Boccadoro M, Klünsner T (2016) Senk- und Drahterodieren von höchstlegierten
Werkzeugstählen hoher Härte und von WC-CoHartmetallen. In: Kolaska H, Danninger H,
Biermann D (eds) Zerspanung von und mit pulvermetallurgischen Werkstoffen: [35. Hagener
Symposium Pulvermetallurgie mit Fachausstellung, 24./25. November 2016]. Heimdall-Verl.,
Hagen, pp 43–58
[10] Abdullah and M. R. Shabgard, Effect of ultrasonic vibration of tool on electrical discharge
machining of cemented tungsten carbide (WC-Co), The International Journal of Advanced
Manufacturing Technology, 38 (11-12) (2008) 1137- 1147.
[11] D. D. Dibitoto, P. T. Eubank, M. R. Patel and M. A. Barrufet, Theoretical models of the
electrical discharge machining process. I. A simple cathode erosion model, Journal of applied
physics, 66 (9) (1989) 4095-4103.
[12] N. Levy and F. Maggi: ‘WED machinability comparison of different steel grades’, CIRPAnn. –
Manuf. Technol., 1990, 39, (1), 183–185.

[13] H. El-Hofy: ‘Advanced machining processes’, 2005, New York, McGraw-Hill.
[14] J. A. McGeough: ‘Electrodischarge machining’ in ‘Advanced methods of machining’, 1988,
New York, Springer.
[15] R. Garg: ‘Effect of process parameters on performance measures of wire electrical discharge
machining’, Ph.D. thesis, National Institute of Technology, Kurukshetra, 2010.
[16] J. T. Huang, Y. S. Liao and W. J. Hsue: ‘Determination of finish-cutting operation number and
machining-parameters setting in wire electrical discharge machining’, J. Mater. Proc. Technol.,
1999, 87, (1–3), 69–81.
[17] I. Maher, L. H. Ling, A. A. D. Sarhan and M. Hamdi: ‘Improve wire EDM performance at
different machining parameters – ANFIS modeling’, IFAC-PapersOnLine, 2015, 48, (1), 105–
110.
[18] M. Shabgard, S. Oliaei, M. Seyedzavvar and A. Najadebrahimi: ‘Experimental investigation
and 3D finite element prediction of the white layer thickness, heat affected zone, and surface
roughness in EDM process’, J. Mech. Sci. Technol., 2011, 25, (12), 3173– 3183.
[19] Lin BT, Jean MD, Chou JH (2007) Using response surface methodology for optimizing
deposited partially stabilized zirconia in plasma spraying. Appl Surf Sci 253:3254–3262
[20] Aslan N (2007) Application of response surface methodology and central composite rotatable
design for modelling the influence some operating parameters of a Multi-Gravity Separator for
coal cleaning. Fuel 86:769–776
[21] Chiang KT, Chang FP, Tsai DC (2007) Modeling and analysis of the resolidified layer of SG
cast iron in the EDM process through response surface methodology. J Mater Process Technol
182:525– 533

MODELING AND ANALYSIS OF SURFACE ROUGHNESS AND WHITE LATER THICKNESS IN WIRE-ELECTRIC DISCHARGE TURNING PROCESS THROUGH RESPONSE SURFACE METHODOLOGY

Recommended

More Related Content

Similar to MODELING AND ANALYSIS OF SURFACE ROUGHNESS AND WHITE LATER THICKNESS IN WIRE-ELECTRIC DISCHARGE TURNING PROCESS THROUGH RESPONSE SURFACE METHODOLOGY (20)

More from IAEME Publication (20)

Recently uploaded (20)

MODELING AND ANALYSIS OF SURFACE ROUGHNESS AND WHITE LATER THICKNESS IN WIRE-ELECTRIC DISCHARGE TURNING PROCESS THROUGH RESPONSE SURFACE METHODOLOGY