Parametric optimization for cutting speed – a statistical regression modeling for wedm

INTERNATIONAL JOURNALEngineering and TechnologyRESEARCH IN
International Journal of Advanced Research in OF ADVANCED (IJARET), ISSN 0976 –
6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 1, January - February (2013), © IAEME
ENGINEERING AND TECHNOLOGY (IJARET)
ISSN 0976 - 6480 (Print)
ISSN 0976 - 6499 (Online)
IJARET
Volume 4, Issue 1, January- February (2013), pp. 142-150
© IAEME: www.iaeme.com/ijaret.asp ©IAEME
Journal Impact Factor (2012): 2.7078 (Calculated by GISI)
www.jifactor.com

PARAMETRIC OPTIMIZATION FOR CUTTING SPEED – A
STATISTICAL REGRESSION MODELING FOR WEDM

S V Subrahmanyam1, M. M. M. Sarcar2
1
Asst Professor, Dept of Mechanical Engg, GVP College of Engineering, Vizag, A.P. India
2
Professor and HOD of Mechanical Engg., A.U. College of Engineering, Vizag, A.P. India

ABSTRACT

Better finish, low tolerance, higher production rate, miniaturization, complex shapes
and profiles of the harder, newer, latest materials like hardened steel, titanium, high strength
temperature resistant alloy, fiber-reinforced composites and ceramics is the present demand
of the manufacturing industries such as Aerospace, nuclear, missile, turbine, automobile, tool
and die making. To satisfy these needs a different class of modern machining techniques,
unconventional in nature, like Wire Electrical discharge Machining (WEDM) emerged. In
WEDM the material removal takes place due to thermal erosion. In this process there is no
contact between the tool and work. In WEDM rough machining produces lesser accuracy and
surface finish, while finish machining produces less surface roughness with less speed. To get
optimum process parameters for higher cutting efficiency and accuracy is very difficult.
Hence, the objective of this paper is, to improve the Cutting Speed and to optimize the effects
of eight input process parameters on cutting speed during the machining of hot die steel (EN-
31) using Taguchi L27(38) orthogonal array (OA) as design of experiments (DOE).

Keywords: EN31, Cutting Speed, Orthogonal array, WEDM

I. INTRODUCTION

In WEDM the process of Metal erosion effect takes place when electric sparks are
generated between the work piece and a wire electrode flushed or immersed in the dielectric
fluid. The WEDM machining plays a major role in manufacturing sectors especially
industries like aerospace, ordinance, automobile and general engineering etc. WEDM
machining process parameters can be optimized by using taguchi method. Taguch method is
based on Orthogonal Array, which provides a set of experiments which are well balanced and
reduces variance for control parameters during the experimentation. Nihat Tosun et al [1] find

142

International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN 0976 –

on the effect and optimization of machining parameters on the notch and material removal
rate (MRR) in wire electrical discharge machining (WEDM) operations. Can Cogun [2] the
settings of machining parameters were determined by using taguchi experimental design
method. The level of importance of the machining parameters on the kerf and the MRR is
determined by using ANOVA. Amar Patnaik et al [3] Introducing zinc coated copper as
electrode tool with the process parameters of discharge current, pulse duration, pulse
frequency, wire speed, wire tension, dielectric flow rate. By using factors, maximization of
MRR and minimization of surface roughness is done in WEDM process using taguchi
method. H.Singh et al [4] analyze the effects of various input process parameters like pulse
on time, pulse off time, gap voltage, peak current , wire feed and wire tension have been
investigated and impact on MRR is obtained. Finally they reported MRR increase with
increase in pulse on time and peak current. MRR decrease with increase in pulse off time and
servo voltage. Wire feed and wire tension has no effect on MRR. Sarkar et al. [5] performed
experiments using +-titanium aluminide alloy as work material and then formulated
mathematical models to predict the cutting speed, SF and dimensional deviation as the
function of different control parameters. In WEDM operations, material removal rate (MRR)
determines the economics of machining and rate of production. In setting the machining
parameters, the main goal is the maximum MRR. The main purpose of this paper is to
investigate effects of machining parameters on the material removal rate of wire EDMed
En31 alloy steel. Hewidy et al [6] developed mathematical models correlating the various
WEDM machining parameters (peak current, duty factor, wire tension and water pressure)
with metal removal rate, wear ratio and surface roughness based on the response surface
methodology. A.K.M. Nurul Amin et al [7] Conducted experiments on cutting of tungsten
carbide ceramic using electro-discharge machining (EDM) with a graphite electrode by using
taguchi methodology

II. EXPERIMENTAL SET UP AND DATA COLLECTION

The experimental setup, design of experiment based on Taguchi Orthogonal Array
and the method of conducting experiments are discussed in this section.

2.1. Work Material and tool/cutting tool material

The experiments were conducted on EN 31 alloy steel material as a work piece. The
work piece material chemical composition of the is shown in Table 1. Brass wire of 0.25 mm
diameter was used as tool electrode in the experimental set up. This is a diffused wire of brass
of type ELECTRA_Duracut. 0.25 mm diameter stratified wire (Zinc coated copper wire) with
vertical configuration has been used and discarded once used. High MRR in WEDM without
wire breakage can be attained by the use of zinc coated copper wire because evaporation of
zinc causes cooling at the interface of work piece and wire and a coating of zinc oxide on the
surface of wire helps to prevent short-circuits (Sho et al., 1989).
Table 1: the chemical composition of EN31 Alloy steel.

Material C Cr Mn Si Fe
EN31 0.95 1.45 0.60 0.22 Balance %wt

143


2.2. Schematic of Machining

All the experiments were conducted on SPRINTCUT (AU) WITH PULSE
GENERATOR ELPULS 40A DLX CNC Wire-cut EDM machine made by
ELECTRONICA company. All the three axes of the machine are servo controlled and can
be programmed to follow a CNC code which is fed through the control panel. All three axes
have an accuracy of 1µm. Through an NC code, machining can be programmed. During the
experimentation the work piece is considered as positive terminal, where as the tool (wire) is
connected to negative terminal of the source. The size of the work piece considered for
experimentation on the wire-cut EDM is 125 mm x 25 mm x 5 mm. A small gap of 0.025
mm to 0.05 mm is maintained in between the wire and work-piece. The high energy density
erodes material from both the wire and work piece by local melting and vaporizing. The di-
electric fluid (de-ionized water) is continuously flashed through the gap along the wire, to
the sparking area to remove the debris produced during the erosion. A collection tank is
located at the bottom to collect the used wire erosions and then is discarded. The wires once
used cannot be reused again, due to the variation in dimensional accuracy.

2.3. Experimental procedure

The work piece is a rectangle plate having dimensions of 5mmx25mmx125mm.
Work piece is machined with zinc coated copper wire, used as a cutting tool, having
diameter of 0.20mm and de-ionized water as a dielectric fluid, into 5mmx5mmx25mm
pieces. Each piece is cut with different input process parameter combination considering
each combination as a separate job. The parameters are varied basing on the Taguchi
Orthogonal Array design. During this operation the respective output parameters or
responses, cutting speed, surface roughness, and dimensional deviations, machining time,
gap current and kerf are gathered. Out of these some like cutting speed, Gap current, gap
voltage were gathered from system display screen and others like MRR, Surface Roughness,
dimensional deviations were calculated separately. Surface roughness Ra values were
measured using Mitutoyo Surface Tester. Dimensional Deviations were measured using
Digital Micro Meter.

2.4. Process parameters and design

Input process parameters such as Pulse On time (TON), Pulse Off time (TOFF), Peak
Current (IP), Spark gap Voltage Setting (SV), Wire tension setting (WT), Wire Feed rate
setting (WF), Servo Feed Setting (SF), Flushing pressure of dielectric fluid (WP) used in this
study are shown in Table 2. Each factor is investigated at three levels to determine the
optimum settings for the WEDM process. These parameters and their levels were chosen
based on the review of literature and as per the few preliminary pilot experiments that were
carried out by varying the process parameters to find their significance and relevance to the
response parameters. In the present study most important output performances in WEDM
Cutting speed was considered for optimizing machining parameters. The gathered
experimental values are recorded as shown in table 3 in line with the L27 Orthogonal Array
design.

144


III. EXPERIMENTAL DATA ANALYSIS

3.1 Analysis of Variance (ANOVA)

In any Experimentation with several variables it is really hard to know which
variable is exactly influencing and which is not. The reason being in WEDM type of
experimentation each parameter will have its own control on the other since they are
acting together as one unit. In spite of that it is necessary to find, if at all, any variation
exists in the experimentation, then by which variable and to take necessary decisions can
be made concerning that parameters. For that purpose ANOVA is use. ANOVA is a
statistical method that can interpret experimental data. It categorizes machining
parameters into significant and insignificant ones.

3.2 S/N Ratio

In any experimentation the expectation will be that the desired values are to
obtained as outputs, if not at least the mean output values correlate to the desired ones.
This desired value is termed as signal and the deviation from the desired or signal is
termed as noise. The signal is the mean for the response and the square deviation for the
output is noise. So, the ratio of mean to square deviation is S/N ratio. This is calculated
as a logarithmic transformation of the loss function as equation 1. As per the objective of
the study, maximization of Cutting Speed, it is clear that ‘the higher value’ of the
experimental data and its corresponding input parameters are of the optimum machining
performance characteristics. It is denoted by ‘η’ with a unit of dB.
ଵ ଵ
η = -10 log ∑୬
୧ୀଵ (1)
୬ ௬ଶ

MINITAB 15 software was used to analyze the experimental data. ANOVA, S/N ratio
calculation are calculated using Minitab 15. Table 3 shows the L27 OA along with the
experimental Response Cutting Speed value and the S/N ratio for the cutting speed
obtained from Minitab. Table 4, 5 shows the S/N ratio, mean S/N ratio of the
parameters according to the level of each parameter.

a. Result Analysis

As discussed earlier higher η value corresponds to better performance. The level
with greatest η value is the optimal level of machining parameters. Table 6 is the Anova
result. Figures 2, 3 show graphically the effect of the eight control factors on Cutting
Speed. From the results it is evident that parameters at level A3B1C3D1E1F1G1H1,
optimal level, gives maximum Cutting Speed. It is also evident from Table 4, 5 that WT
WF WP are having less significant impact on the Cutting Speed. Here WT is totally
insignificant, WF, WP show less influence.

145


IV. CONFIRMATION OF EXPERIMENTS

The final step in the DOE process is confirmation experiment. This to validate the
conclusions drawn during the analysis phase. An experiment conducted with a specific
combination of the factors and levels previously evaluated. Its response value is compared
with the predicted value of the same with the Minitab. To have more confirmation a
mathematical general non linear regression model of the form given below is considered
which gives the relationship between response and the input variables.

Response= C * Aâ1 * Bâ2 * Câ3…. Where C is constant A, B, C.. are the input process
variables, and a1,a2,a3… are the coefficients. This is solved by using a custom made
Regression code which runs on Fortron. From the available experimental data the suggested
model will be as follows:

Y=7.593 *A^0.4534*B^-0.2993*C^9.13E-02*D^-0.114*E^-2.64E-02*F^-2.61E-02*G^-
0.3082*H^ 4.48E-02

With this equation the response Y is obtained. With optimal parameters. The optimal model
obtained above, is also subjected and the value is obtained and compared with the
experimental values. The comparative statement of the Predicted Optimal Value, its
experimental value, its Math model value are tabled in table 7. Table 8 gives the respective
S/N ratio values. From Table7, 8 it is clearly evident that the results obtained from the
optimal that optimal suggested value has an edge over.

Table.2 Levels for various control factors

Sl.No. PARAMETERS SYMBOL LEVEL1 LEVEL2 LEVEL3 UNITS

1 Pulse On time (A) TON 122 125 128 µ sec

2 Pulse Off time (B) TOFF 53 58 63 µ sec

3 Peak Current (C) IP 130 180 230 Ampere

4 Spark gap Voltage(D) SV 20 30 40 Volts

5 Wire Tension (E) WT 2 3 4 Kg-f

6 Wire Feed rate (F) WF 4 5 6 m/min

7 Servo Feed (G) SF 500 1300 2100 mm/min

8 Dielectric Flushing pressure (H) WP 2 3 4 Kg/cm2

146


Table.3 Experimental results using L27 OA

CSPEED CS S/N
TON TOFF IP SV WF WT SF WP
(CS) RATIO
1 1 1 1 1 1 1 1 2.84 9.06637
1 1 1 1 2 2 2 2 2.34 7.38432
1 1 1 1 3 3 3 3 1.79 5.05706
1 2 2 2 1 1 1 2 2.74 8.75501
1 2 2 2 2 2 2 3 2.26 7.08217
1 2 2 2 3 3 3 1 1.68 4.50619
1 3 3 3 1 1 1 3 2.7 8.61727
1 3 3 3 2 2 2 1 2.23 6.9661
1 3 3 3 3 3 3 2 1.56 3.86249
2 1 2 3 1 2 3 1 1.72 4.71057
2 1 2 3 2 3 1 2 2.7 8.61976
2 1 2 3 3 1 2 3 2.26 7.08217
2 2 3 1 1 2 3 2 1.93 5.71115
2 2 3 1 2 3 1 3 2.84 9.06637
2 2 3 1 3 1 2 1 2.45 7.78332
2 3 1 2 1 2 3 3 1.61 4.13652
2 3 1 2 2 3 1 1 2.7 8.61756
2 3 1 2 3 1 2 2 2.19 6.80888
3 1 3 2 1 3 2 1 2.45 7.78332
3 1 3 2 2 1 3 2 1.91 5.6347
3 1 3 2 3 2 1 3 2.83 9.03573
3 2 1 3 1 3 2 2 2.19 6.80888
3 2 1 3 2 1 3 3 1.55 3.80663
3 2 1 3 3 2 1 1 2.7 8.61652
3 3 2 1 1 3 2 3 2.35 7.42136
3 3 2 1 2 1 3 1 1.79 5.05706
3 3 2 1 3 2 1 2 2.76 8.81818

Table.4 S/n ratios with the levels for each parameter

Level TON TOFF IP SV WF WT SF WP
1 6.811 7.153 6.7 7.263 7.001 6.957 8.801 7.012
2 6.948 6.904 6.895 6.929 6.915 6.94 7.236 6.934
3 6.998 6.701 7.162 6.566 6.841 6.86 4.72 6.812
Delta 0.187 0.452 0.462 0.697 0.16 0.096 4.081 0.2
Rank 6 4 3 2 7 8 1 5
147


Table.5 mean s/n values for each parameter at its levels

Level TON TOFF IP SV WF WT SF WP
1 2.237 2.316 2.212 2.343 2.281 2.27 2.755 2.284
2 2.266 2.26 2.251 2.263 2.258 2.264 2.302 2.258
3 2.281 2.209 2.322 2.178 2.246 2.251 1.727 2.243
Delta 0.044 0.106 0.111 0.165 0.034 0.019 1.028 0.041
Rank 5 4 3 2 7 8 1 6

Table.6 Anova

Source DF Seq SS Adj SS Adj MS F P
TON 2 0.00885 0.00885 0.00443 2.28 0.153
TOFF 2 0.05092 0.05092 0.02546 13.12 0.002
IP 2 0.05669 0.05669 0.02834 14.61 0.001
SV 2 0.12318 0.12318 0.06159 31.75 0.000
WF 2 0.00556 0.00556 0.00278 1.43 0.283
WT 2 0.00180 0.00180 0.00090 0.46 0.642
SF 2 4.78108 4.78108 2.39054 1232.16 0.000
WP 2 0.00765 0.00765 0.00383 1.97 0.190
Error 10 0.01940 0.01940 0.00194
Total 26 5.05513
S = 0.0440468 R-Sq = 99.62% R-Sq(adj) = 99.00%

Main Effects Plot for Means
Data Means

TO N TO F F IP
2.8

2.4

2.0

122 125 128 53 58 63 130 180 230
Mean of Means

SV WF WT
2.8

2.4

2.0

20 30 40 2 3 4 4 5 6
SF WP
2.8

2.4

2.0

500 1300 2100 3 4 5

Fig2.Mean effect values of Cutting speed

148


Main Effects Plot for SN ratios
Data Means

TO N TO F F IP
9

7

5

Mean of SN ratios
122 125 128 53 58 63 130 180 230
SV WF WT
9

7

5

20 30 40 2 3 4 4 5 6

SF WP
9

7

5

500 1300 2100 3 4 5

Signal-to-noise: Larger is better

Fig3. S/N ratios of Cutting speed

Table 7 Results of the confirmation experiment for Cutting Speed
Sl No Experiment Math Model
Cutting Speed Value obtained at Initial
1 conditions A2B2C2D2E2F2G2H2 2.29 2.11

Cutting Speed Value obtained at Optimum
2 Conditions A3B1C3D1E1F1G1H1 3.36 3.23

3 Improvement of Cutting Speed obtained 1.46 times 1.53 times

Table 8 Results of the confirmation experiment for Cutting Speed
Sl Experiment Math Model
No
S/N RATIO Value obtained at Initial conditions
1 7.19 6.48
S/N RATIO Value obtained at Optimum Conditions
2 10.52 9.58

3 Improvement of S/N RATIO obtained 3.23 3.10

V. CONCLUSION

In this study the following are achieved.
1 The effects of TON, TOFF, IP, SV, SF, WT, SF, WP are investigated on the EN31 Alloys
Steel for Cutting speed with which to estimate the speed of material removal.
2. With the help of ANOVA, S/N ratio and Math model the optimal input parameter
combination for the cutting speed on the WEDM machined arrived, which will be useful for
the people who do not have much idea of WEDM can use for the selection of input
parameters.

149


REFERENCES

[1] 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 Materials Processing
Technology 152 (2004) 316-322.
[2] Can Cogun (2004) a study on kerf and material removal rate in wire electrical
discharge machining based on Taguchi method. Journal of Materials Processing Technology.
[3] S. S. Mahapatra, Amar Patnaik, (2006) Optimization of wire electrical discharge
machining (WEDM) process parameters using Taguchi method. International Journal of
Advanced Manufacturing Technology.
[4] H. Singh, R. Garg (2009) Effects of process parameters on material removal rate in
WEDM, Journal of achievements in material and manufacturing engineering, Volume 32,
Issue.
[5] Sarkar S, Mitra S, Bhattacharyya B (2005) Parametric analysis and optimization of
wire electrical discharge machining of γ-titanium aluminide alloy. Journal of Materials
Processing Technology 159:286–294.
[6] M.S. Hewidy, T.A. El-Taweel, M.F. El-Safty, Modelling the machining parameters of
wire electrical discharge machining of Inconel 601 using RSM, Journal of Materials Process-
ing Technology 169 (2005) 328-336.
[7] Mohd Amri Lajis, H.C.D. Mohd Radzi, A.K.M. Nurul Amin (2009) The Implemen-
tation of Taguchi Method on EDM Process of Tungsten Carbide. European Journal of
Scientific Research, ISSN 1450-216X Vol.26 No.4 (2009), pp.609-617.
[8] U. D. Gulhane, A. B. Dixit, P. V. Bane and G. S. Salvi, “Optimization Of Process
Parameters For 316l Stainless Steel Using Taguchi Method And Anova” International Journal
of Mechanical Engineering & Technology (IJMET), Volume 3, Issue 2, 2012, pp. 67 - 72,
ISSN Print: 0976 – 6340, ISSN Online: 0976 – 6359, Published by IAEME.
[9] Rodge M.K, Sarpate S.S and Sharma S.B, “Investigation on Process Response and
Parameters in Wire Electrical Discharge Machining of Inconel 625” International Journal of
Mechanical Engineering & Technology (IJMET), Volume 4, Issue 1, 2013, pp. 54 - 65, ISSN
Print: 0976 – 6340, ISSN Online: 0976 – 6359, Published by IAEME.

150

Parametric optimization for cutting speed – a statistical regression modeling for wedm

More Related Content

What's hot

Viewers also liked

Similar to Parametric optimization for cutting speed – a statistical regression modeling for wedm

More from IAEME Publication

Parametric optimization for cutting speed – a statistical regression modeling for wedm