Available online at www.sciencedirect.

com

ScienceDirect
Materials Today: Proceedings 2 (2015) 2548 – 2554

4th International Conference on Materials Processing and Characterization

Multiple Response Optimization of Wire EDM on Aluminium

HE30 by using Grey Relational Analysis
V. Chengal Reddya, N. Deepthib, , N.Jayakrishnac
a,b,c
Assistant Professor, Dept. of Mechanical Engineering, AITS-Rajampet-516126, Andhra Pradesh, India.

Abstract

Wire Electrical Discharge Machining (WEDM) allowed success in the production of newer materials, especially for the aerospace
and medical industries. Using WEDM technology, complicated cuts can be made through difficult-to-machine electrically
conductive components. The high degree of the obtainable accuracy and the fine surface quality make WEDM valuable. WEDM
is so complex in nature that the selection of appropriate input parameters is not possible by the trial-and-error method. The selection
of machining parameters in any machining process significantly affects production rate, product quality and production cost of a
finished component. WEDM process involves a large number of variables that affect its performance. In this paper, an attempt is
made to study the effect of various process parameters such as pulse on time, pulse off time, wire tension, current, upper flush and
lower flush for Aluminium HE30. The experimentation has been completed with the help of Taguchi grey relational analysis.
Taguchi grey relational analysis is used to optimize the process parameters on multiple performance characteristics such as Material
removal rate, surface finish. The experimental result analysis showed that the combination of pulse on time, pulse off time, wire
tension, lower flush, wire tension and upper flush is essential to achieve maximization of material removal rate and minimization
of surface roughness and kerf width.
© 2014 The Authors. Elsevier Ltd. All rights reserved.
© 2015 Published by Elsevier Ltd.
Selection andpeer-review
Selection and peer-review under
under responsibility
responsibility of theofconference
the conference committee
committee membersmembers
of the 4thof the 4th International
International conference
conference on Materials on
Materials Processing
Processing and Characterization.
and Characterization.

Keywords: Wire EDM; Aluminium HE30; Taguchi method; surface roughness; MRR; GRA;

1. Introduction

Wire electrical discharge machining (WEDM) has become an important non-traditional machining process, is mainly used to cut intricate
shapes and designs into hard metals, which are otherwise difficult to form, mold or manipulate. It is most useful in the electronics and aerospace
sectors for prototyping and manufacturing various parts. Most often, steel and titanium are processed with help of wire electrical-discharge
machining. It is a thermo-
* Corresponding author. Tel.: 8019956112;
E-mail address: vcr.312@gmail.com

2214-7853 © 2015 Published by Elsevier Ltd.

Selection and peer-review under responsibility of the conference committee members of the 4th International conference on Materials Processing
and Characterization.
doi:10.1016/j.matpr.2015.07.201
 V. Chengal Reddy et al. / Materials Today: Proceedings 2 (2015) 2548 – 2554 2549

electrical process in which material is removed by generating a series of discrete sparks between electrode and work
piece immersed in a liquid dielectric medium. A thin electrically conductive wire is used as the electrode. Normally,
the wire is held by a pin guide at the upper and lower parts of the work piece as shown in fig 1. In most cases, the wire
will be discarded once used. The major advantage of EDM is that this machining process enables us to obtain
components with desired shape and closer Typical materials, used as EDM electrodes, include copper, graphite,
tungsten, brass, steel, copper–tungsten, copper– chromium alloy, haste alloy etc. WEDM finds extensive applications
in various fields like aerospace, automobile, electronics industries. Electrical discharge machining (EDM) actually is
a process of utilizing the removal phenomenon of electrical-discharge in dielectric. Therefore, the electrode plays an
important role, which affects the material removal rate and the tool wear rate.

Fig.1. Schematic diagram of WEDM

The selection of machining parameters in a machining process significantly affects production rate and quality of
machined components. The selection of these parameters in WEDM is primarily dependent on the operator’s experience
and machining parameter tables provided by the machine-tool manufacturers. However, such criterion does guarantee
neither high production rate nor good surface quality. Hence there is a need to develop a methodology to find the optimal
process parameters. The present work highlights the development of mathematical models for correlating the inter-
relationships of various WEDM machining parameters of Aluminum HE30 material such as pulse on time, pulse off
time, wire tension, current, upper flush and lower flush on the Material removal rate (MRR), Kerf Width and surface
roughness. This work has been established based on the response surface methodology (RSM) approach. Mathematical
models fitted to the experimental data will contribute towards the selection of the optimum process conditions.

2. METHODOLOGY

Grey Relational Analysis

In the grey relation analysis, experiment data, i.e., measured responses are first normalised in the range of 0 to 1.
This process is called normalization or grey relation generation. Based on this data, grey relation coefficients are
calculated to represent the correlation between the ideal (best) and the actual normalized experimental data. Overall,
grey relation grade is then determined by averaging the grey relation coefficient corresponding to selected responses.
The overall quality characteristics of the multi-response process depend on the calculated grey relation grade.

2.1. Normalization

Normalization of the signal to noise ratio is performed to prepare raw data for the analysis where the original
sequence is transformed to a comparable sequence. Linear normalization is usually required since the range and unit in
one data sequence may differ from the others.
2550 V. Chengal Reddy et al. / Materials Today: Proceedings 2 (2015) 2548 – 2554

There are three different types of data normalization according to the requirement of Lower the Better (LB), Higher
the Better (HB), or Nominal the Best (NB) criteria.
If the target value of original sequence is infinite, then it has a characteristic of the “higher is better”. The original
sequence can be normalized as follows:
௫೔೚ ሺ୩ሻି୫୧୬௫೔೚ ሺ୩ሻ
‫ݔ‬௜‫ כ‬ൌ  (1)
௠௔௫௫೔೚ ሺ୩ሻି୫୧୬௫೔೚ ሺ୩ሻ
When the “Smaller is better” is a characteristic of the original sequence, then the original sequence should be
normalized as follows:
௠௔௫௫೔೚ ሺ୩ሻି௫೔೚ ሺ୩ሻ
‫ݔ‬௜‫ כ‬ൌ  (2)
௠௔௫௫೔೚ ሺ୩ሻି୫୧୬௫೔೚ ሺ୩ሻ
However, if there is a definite target value (desired value) to be achieved, the original sequence will be normalized
in form:
ห௫೔೚ ሺ௞ሻି௫ ೚ ห
‫ݔ‬௜‫ כ‬ൌ ͳ െ  (3)
௫೔೚ ሺ௞ሻି௫ ೚
Or, the original sequence can be simply normalized by the most basic methodology, i.e. let the values of original
sequence be divided by the first value of the sequence:
௫೔೚ ሺ୩ሻ
‫ݔ‬௜‫ כ‬ൌ  (4)
௫೔೚ ሺଵሻ
Where i = 1 . . . m; k = 1. . . n. m is the number of experimental data items and n is the number of parameters.š୧୭ (k)
denotes the original sequence,š୧‫ כ‬the sequence after the data pre-processing,ƒšš୧୭ (k) the largest value ofš୧୭ (k),
minš୧୭ (k) the smallest value of š୧୭ (k)ƒ†š ୭ is the desired value.
Determination of deviation sequences Δ0 i(k) :
The deviation sequence, Δ0i(k) is the absolute difference between the reference sequence x0*(k) and the
comparability sequence xi*(k) after normalization. It is determined using equation:
ΔͲ‹ሺሻ ൌ  ȁš଴‫ כ‬ሺሻ െ š୧‫ כ‬ሺሻȁ (5)
2.2 Calculation of grey relational coefficient (GRC)
GRC for all the sequences expresses the relationship between the ideal (best) and actual normalized S/N ratio. If
the two sequences agree at all points, then their grey relational coefficient is 1. The grey relational coefficient ξi(k) for
the kth performance characteristics in the ith experiment can be expressed as :
௫೔೚ ሺ୩Δ୫୧୬ା఍Δ୫ୟ୶ሻ
ߦ௜ ሺ݇ሻ  ൌ  (6)
Δை೔ ሺ௞ሻା఍Δ୫ୟ୶
Where ΔOi is the deviation sequence of the reference sequence andš଴‫ כ‬ሺሻ
is the comparability sequence. ζ is
distinguishing or identification coefficient: ζ ‫[ א‬0, 1] (the value may be adjusted based on the actual system
requirements). A value of ζ is the smaller and the distinguished ability is the larger. ζ = 0.5 is generally used. Grey
relational coefficient for 27 comparability sequences.
2.3 Calculation of grey relational grade (GRG)
After the grey relational coefficient is derived, it is usual to take the average value of the grey relational coefficients
as the grey relational grade. The grey relational grade is defined as follows:
ଵ
γ୧  ൌ  σ୬୩ୀଵ Ɍ‹ሺሻ (7)
୬
However, in a real engineering system, the importance of various factors varies. In the real condition of unequal
weight being carried by the various factors, the grey relational grade was extended and defined as above.
The grey relational grade γi represents the level of correlation between the reference sequence and the comparability
sequence. If the two sequences are identical, then the value of grey relational grade is equal to 1. The grey relational
grade also indicates the degree of influence that the comparability sequence could exert over the reference sequence.
Therefore, if a particular comparability sequence is more important than the other comparability sequences to the
reference sequence, then the grey relational grade for that comparability sequence and reference sequence will be higher
than other grey relational grades.

3. Experimental Details

A detailed survey has been carried out to find out how the machining parameters affect surface roughness, MRR,
kerf width of AL HE30. Based on this, the six machining parameters pulse-on time, pulse-off time, wire tension, upper
 V. Chengal Reddy et al. / Materials Today: Proceedings 2 (2015) 2548 – 2554 2551

flush, lower flush, current were selected. Taguchi’s L27 orthogonal array in the design of experiments (DOE) technique
has been implemented to conduct the experiments. The six parameters with 3 levels each and then
3x3x3x3x3x3=729runs were required in the experiments for six independent variables. But using Taguchi’s orthogonal
array, the number of experiments reduced to 27 experiments from 729 experiments. All the experiments were conducted
on CNC-WEDM and the tool material used for the present study was Brass. The machining parameters used and their
levels chosen are presented in Table 1.

Table 1. Machining parameters used and their levels

levels
Control parameters Notation Units
1 2 3
Pulse on time A μsec 110 111 112
Pulse off time B μsec 61 62 63
Current C Amp 11 12 13
Wire tension D gm 660 780 900
Upper flush E kg/cm2 8 9 10
lower flush F kg/cm2 7 8 9

Table 2. Experimental design using L27 orthogonal array matrix

S. NO Pulse on time Pulse off time Current Wire tension Upper flush Lower flush
(μsec) (μsec) (Amp) (gm) (kg/cm2) (kg/cm2)
1 110 61 11 660 8 7
2 110 61 11 660 9 8
3 110 61 11 660 10 9
4 110 62 12 780 8 7
5 110 62 12 780 9 8
6 110 62 12 780 10 9
7 110 63 13 900 8 7
8 110 63 13 900 9 8
9 110 63 13 900 10 9
10 111 61 12 900 8 8
11 111 61 12 900 9 9
12 111 61 12 900 10 7
13 111 62 13 660 8 8
14 111 62 13 660 9 9
15 111 62 13 660 10 7
16 111 63 11 780 8 8
17 111 63 11 780 9 9
18 111 63 11 780 10 7
19 112 61 13 780 8 9
20 112 61 13 780 9 7
21 112 61 13 780 10 8
22 112 62 11 900 8 9
23 112 62 11 900 9 7
24 112 62 11 900 10 8
25 112 63 12 660 8 9
26 112 63 12 660 9 7
2552 V. Chengal Reddy et al. / Materials Today: Proceedings 2 (2015) 2548 – 2554

27 112 63 12 660 10 8

4. Analysis and discussion of experimental results

In the present study, Surface roughness, MRR and Kerf width of in different parameters and experimental runs are
listed in table no. typically lower values of surface roughness and kerf width, higher value of MRR as the target values
are desirable. Therefore the data sequences have the smaller the better characteristics for surface roughness and Kerf
width, Higher-the-better characteristics for MRR. The values of average surface roughness, kerf width and MRR are set
to be the reference sequence. More over the results of experiments were the compatibility sequencesš୧‫ כ‬ሺሻ, i = 1 - 27,
k = 1 – 3.

Table 3. Sequences after data preprocessing.

S.No Kerf Width (mm) Surface Roughness (μm) MRR (gm/min)
1 3.261 0.365 0.0522
2 3.101 0.38 0.0513
3 2.884 0.35 0.04432
4 3.178 0.29 0.04517
5 3.375 0.255 0.03814
6 2.358 0.32 0.04492
7 3.885 0.47 0.0799
8 3.592 0.51 0.1349
9 3.86 0.38 0.0974
10 3.168 0.335 0.0568
11 2.858 0.39 0.0625
12 3.185 0.41 0.07018
13 3.682 0.47 0.1192
14 3.419 0.625 0.1687
15 3.677 0.355 0.0955
16 2.823 0.365 0.0495
17 3.172 0.3 0.0396
18 2.96 0.33 0.0428
19 3.78 0.415 0.1131
20 3.889 0.37 0.0966
21 3.896 0.305 0.08267
22 3.06 0.29 0.0471
23 2.97 0.355 0.056
24 3.407 0.32 0.0519
25 3.323 0.325 0.0558
26 3.187 0.375 0.0648
27 3.064 0.3 0.0509

Table 3 list all of the sequences following data pre processing using equation1, 2.
The distinguishing coefficients Ɍ can be substituted for the grey relation coefficient in equation 6. If all the process
have equal weighing Ɍ is 0.5.
 V. Chengal Reddy et al. / Materials Today: Proceedings 2 (2015) 2548 – 2554 2553

Table 4. Grey Relation Coefficients and Grey Relation Grades.

S.No Kerf Width (mm) Surface Roughness (μm) MRR (gm/min) Weighted grey relational grade
1 0.4545 0.5868 0.38809 0.47646
2 0.4735 0.5249 0.38458 0.46099
3 0.4359 0.4555 0.3574 0.41626
4 0.3684 0.5534 0.36085 0.42755
5 0.3333 0.6383 0.3333 0.43496
6 0.4012 0.3333 0.35980 0.36496
7 0.6119 0.9958 0.4988 0.70216
8 0.6877 0.7589 0.7688 0.73846
9 0.4734 0.9708 0.5747 0.67296
10 0.4180 0.5496 0.4059 0.45783
11 0.4873 0.4481 0.428 0.45446
12 0.5149 0.5561 0.4589 0.50996
13 0.6112 0.8207 0.6816 0.70450
14 1 0.6600 1 0.88660
15 0.4420 0.8168 0.5666 0.60846
16 0.4545 0.4385 0.3776 0.42353
17 0.379 0.5510 0.3391 0.42300
18 0.4122 0.4781 0.3516 0.41390
19 0.5224 0.8978 0.6503 0.69010
20 0.4068 1 0.5715 0.65940
21 0.3843 0.4590 0.5102 0.45110
22 0.3684 0.5100 0.3683 0.41550
23 0.4420 0.4813 0.4028 0.44200
24 0.4009 0.6541 0.3849 0.47990
25 0.4065 0.6139 0.4020 0.47410
26 0.4672 0.5568 0.4374 0.48710
27 0.379 0.5119 0.38305 0.42460

The weighted grey relational grade calculated for each sequence is taken as a response for the further analysis. The
larger-the-better quality characteristic was used for analyzing the GRG, since a larger value indicates the better
performance of the process. The number of repeated test is one, since only one relational grade was acquired in each
group for this particular calculation of S/N.
Table 4. Shows the grey calculation results. The grey relation grades are now analysed with Taguchi in Minitab
software. This result shows that the best processing condition is the (A3, B1, C1, D2, E3, F2).

Fig. 2. Mean Effects Plot for S/N ratios

2554 V. Chengal Reddy et al. / Materials Today: Proceedings 2 (2015) 2548 – 2554

The best optimum condition is the i.e., at a feed of 3 mm/min, with lubricant distilled water and work piece material
AL HE30 is the optimum condition.

Table5: Confirmation experiment results

Control parameters MRR Surface Kerf width
(gm/min) roughness (mm)
(μm)
Before A1, B1, C1, D1, E1, F1 0.0522 3.261 0.365
optimization
After A3, B1, C1, D2, E3, F2 0.153 2.861 0.257
optimization

5. CONCLUSION

The selection of right combination of input parameters in WEDM is difficult as the process involves a large number
of control variables. The effects of input parameters pulse-on time, pulse-off time, current, wire tension, upper flush and
lower flush on surface roughness, metal removal rate and kerf width while machining the Aluminium HE30 material
were analyzed with the experimental data obtained after conducting the experiments as per the Design of Experiments.
Grey Relational Analysis (GRA), for finding the optimal parameters affecting MRR, Surface Roughness and Kerf
Width are found to be Pulse on time = 112μs, Pulse off time=61 μs, Current=11Amp, Wire tension=780gm, Upper
flush=10 kg/cm2, lower flush=8 kg/cm2 for higher MRR value of 0.153mm³/min, lower Surface Roughness value of
2.861 μm and lower Kerf Width value of 0.257mm. Further study Artificial intelligent system such as the fuzzy logic
system, simulated annealing, genetic algorithms might be used to enhance the ability of the prediction system.

REFERENCES

[1] D. V. S. S. S. V. Prasad, A. Gopala Krishna, Analysis of performance characteristics of wire electrical discharge machining, Journal of
manufacturing, 2009, Vol.4, Issue.3, pp 205-211.
[2] Prasad D.V.S.S.S.V., Gopala Krishna A “Empirical modeling and optimization of wire Electric Discharge Machining”, International Journal of
Advanced Manufacturing Technology, 2009.
[3] Gokler, M.I.; Ozanozgu, A.M. : “Experimental investigation of effects of cutting parameters on surface roughness in the WEDM process”,
International Journal of Machine Tools and Manufacture, 40, pp. 1831-1848, 2000.
[4] Hargovind Soni, M.K.Pardhan,” Multi reponse optimization of EDM using Grey- PCA Approach ,” ICEMT 2013.
[5] Saurav Datta, Siba Sankar Mahapatra, 2010, Modeling, Simulation and Parametric Optimization of wire EDM Process using response surface
methodology coupled with grey-Taguchi technique, International Journal of Engineering Science & Technology vol.2(5), p.162.
[6] M. Durairaj, S. Gowri, M.H. Gauthamkumar, M. Ashok kumar, R. Aishwarya, 2012, Optimization of Wire Cut Electrical Discharge Machining
of Inconel 800 using Grey Relational Analysis vol.576, p.552.