See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/270294221

Multi-response Optimization of Machining Parameters of Turning AA6063 T6

Aluminium Alloy using Grey Relational Analysis in Taguchi Method

Article  in  Procedia Engineering · December 2014

DOI: 10.1016/j.proeng.2014.12.242

CITATIONS READS

42 284

2 authors:

Jayaraman Ponnambalathandi L. Mahesh Kumar

Prathyusha Engineering College St.Peter's University
11 PUBLICATIONS   49 CITATIONS    12 PUBLICATIONS   64 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Optimization of cutting parameters in turning of AA6351 View project

All content following this page was uploaded by Jayaraman Ponnambalathandi on 17 February 2015.

The user has requested enhancement of the downloaded file.

 Available online at www.sciencedirect.com

ScienceDirect
Procedia Engineering 97 (2014) 197 – 204

12th GLOBAL CONGRESS ON MANUFACTURING AND MANAGEMENT, GCMM 2014

Multi-response Optimization of Machining Parameters of Turning

AA6063 T6 Aluminium Alloy using Grey Relational Analysis in
Taguchi Method
P. Jayaramana*, L. Mahesh kumarb
a
Research Scholar, St. Peter’s University, Associate Professor, Prathyusha Institute of Technology and Management,Chennai, India
b
Department of Mechanical Engineering, St. Peter’s University, Chennai, India

Abstract

This paper presents a novel approach for the optimization of machining parameters on turning of AA 6063 T6 aluminium alloy
with multiple responses based on orthogonal array with grey relational analysis. Experiments are conducted on AA 6063 T6
aluminium alloy. Turning tests are carried out using uncoated carbide insert under dry cutting condition. In the present work,
turning parameters such as cutting speed, feed rate and depth of cut are optimized considering the multiple responses such as
surface roughness (Ra), roundness (Ø) and material removal rate (MRR). A grey relational grade(GRG) is determined from the
grey analysis. Optimum levels of parameters have been identified based on the values of grey relational grade and then the
significant contribution of parameters is determined by ANOVA. To validate the test result, confirmation test is performed.
Experimental outcomes have proved that the responses in turning process can be enhanced efficiently through this fresh
approach.
© 2014
© 2014The TheAuthors.
Authors. Published
Published by Elsevier
by Elsevier Ltd.
Ltd. This is an open access article under the CC BY-NC-ND license
Selection and peer-review under responsibility of the Organizing Committee of GCMM 2014.
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the Organizing Committee of GCMM 2014
Keywords: Turning; Orthogonal array; Grey relational analysis; ANOVA; Optimization.

1. Introduction

The manufacturing industries are continuously challenged for achieving higher productivity and high quality

* Corresponding author. Tel.: +91-9840397361;

E-mail address: [email protected]

1877-7058 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the Organizing Committee of GCMM 2014
doi:10.1016/j.proeng.2014.12.242
198 P. Jayaraman and L. Mahesh kumar / Procedia Engineering 97 (2014) 197 – 204

products in order to remain competitive. The desired shape, size and finished ferrous and non ferrous materials are
conventionally produced through turning the preformed blanks with the help of cutting tools that moved past the
work piece in a machine tool. Turning is an important and widely used machining processes in engineering
industries. Surface quality is an important performance characteristic to evaluate the productivity of machine tools
as well as machined components [1]. Surface roughness is the critical quality indicator for machined surfaces [2]. A
good quality turning surface can lead to improvement in strength properties such as fatigue strength, corrosion
resistance, assembly tolerance, wear rate, coefficient of friction, cleanability, thermal resistance and aesthetics etc.
At the same time, higher material removal rate (MRR) is desired by the industry to cope up with mass production
without sacrificing product quality in shorter time. Higher MRR is achieved through increasing the process
parameters like cutting speed, feed and depth of cut. However, very high cutting speed induces the larger power
which may exceed the power available in the machine tool. Also at the same time, the cutting temperature increases
with increase in the process parameters. This is detrimental to both the tool and the product as it causes dimensional
inaccuracies by thermal deformation, impairs the machined surfaces by inducing thermal effects and built-up-edge
formation and damages the tool sharpness and vibration. So, selection of appropriate process parameters plays a key
role in the effectiveness, efficiency and overall economy of manufacturing by machining to achieve these objectives
(higher MRR and product quality). In turning, parameters such as cutting speed, feed rate and depth of cut affects
surface finish, roundness and MRR. Generally, the suitable cutting parameters are determined based on experience
or by use of a handbook which does not guarantee optimal performance [3]. Many studies have been conducted to
determine the effect of machining parameters on surface roughness, but only few studies have been conducted to
determine the effect of machining parameters on roundness and MRR.
Ashok kumar sahoo et al. [4] have used grey relational analysis to perform multi-objective optimization of surface
roughness and MRR in turning of AA 1040 steel and determined that cutting speed is the most influencing
parameter affecting combined grey relational grade followed by depth of cut and feed rate. Chorng-Jyh Tzeng et al.
[5] have used Grey relational analysis to perform optimization of turning operations with multiple performance
characteristics such as roughness average, roughness maximum, and roundness. The depth of cut was identified to
be the most influencing parameter affecting the grey relational grade followed by cutting speed and feed rate. For
simultaneous optimization of multi-responses, Taguchi normalized quality loss function [6] and grey based Taguchi
method have been popularly used in drilling [7], arc welding process [8] and turning [9] and greatly improved
through this approach. The grey relational analysis based on grey system theory can be used for solving the
complex interrelationships among the multiple responses [10]. A gray relational grade is determined to evaluate the
multiple responses. In the present study, experimental details using the Taguchi method of parameter design have
been employed for optimizing multiple performance characteristics such as surface roughness, roundness and MRR
for turning of AA 6063 T6. Further, for optimization of multiple response characteristics, grey relational analysis
was used.

2. Material Testing and Specification

The composition of AA 6063 T6 is 0.6 wt.% Si, 0.34 wt.% Fe, 0.09 wt.% Cu, 0.09 wt.% Mn, 0.88 wt.% Mg, 0.092
wt.% Cr, 0.095 wt.% Zn, 0.092 wt.% Ti, 97.721 wt.% Al. This alloy is widely used in the manufacturing of doors,
extrusions, window frames and irrigation tubing. CNC lathe was used for machining. The tool used is uncoated
carbide insert tool. The specification of cutting tool is DCGT 11 T3 04. The surface roughness was measured using
the Surfcorder SE 1200, Surface profilometer. The machining was done under dry cutting condition.

3. Experimental Work

Taguchi’s L9 orthogonal array was used to design the experiments with three factors and three levels. Experiments
were conducted based on the Taguchi’s method which is a powerful tool used in design of experiments [11].
Taguchi advocates use of orthogonal array designs to assign the factors chosen for the experiment. The advantage of
Taguchi method is that it uses a special design of orthogonal arrays to study the entire parameter space with only a
small number of experiments. Compared to the conventional approach of experimentation, this method reduces
drastically the number of experiments that are required to model the response functions [12]. The assignment of the
levels to the factors and the various parameters used are given in Table 1. The experimental results for L9
orthogonal array are given in Table 2.
 P. Jayaraman and L. Mahesh kumar / Procedia Engineering 97 (2014) 197 – 204 199

Table 1. Assignment of the Levels to the Factors

Sl.No Factors Unit Symbols Level 1 Level 2 Level 3

1 Cutting speed m/min v 119.22 158.96 198.71
2 Feed rate mm/rev f 0.05 0.075 0.1
3 Depth of cut mm d 0.1 0.15 0.2

Table 2. Experimental Results for L9 Orthogonal Array

Trial Actual value Ra (μm) Roundness (Ø) MRR (cm3/min)

No.
v (m/min) f (mm/rev) d (mm)

1 119.22 0.05 0.1 0.380 3.63 0.2123

2 119.22 0.075 0.15 0.360 3.9 0.4302

3 119.22 0.1 0.2 0.535 6.4 0.6872

4 158.96 0.05 0.15 0.388 3.85 0.3865

5 158.96 0.075 0.2 0.509 5.36 0.6872

6 158.96 0.1 0.1 0.533 7.37 0.5208

7 198.71 0.05 0.2 0.459 4.12 0.5821

8 198.71 0.075 0.1 0.469 3.73 0.4899

9 198.71 0.1 0.15 0.556 7.1 0.8531

4. Multi-response optimization using grey relational analysis

Grey relational grade is employed to convert multi objective problem into a single objective. The scope of this
study was to identify the optimal combination of process parameters that concurrently minimize the surface
roughness (Ra), roundness (Ø) and maximize the material removal rate (MRR). In order to accomplish this, grey
relational analysis was employed. Grey relational analysis consists of the following steps:

Step 1: Grey relational generation

Normalizing the experimental data according to the type of performance response, is the first step involved in grey
relational analysis. If the target value of the actual sequence is infinite, such as MRR, it has “the-larger-the-better”
characteristic and the normalization of actual sequence can be performed as follows:

k 0
* xi (k )  min( xi ( k ))
xi ( k ) (1)
0 0
max( xi ( k ))  min( xi ( k ))
200 P. Jayaraman and L. Mahesh kumar / Procedia Engineering 97 (2014) 197 – 204

If a defined target value, OB, exists, the normalization of actual sequence can be performed as follows:

µxi0 (k ) OBµ
xi* (k ) 1  (2)
max[max( xi0 (k ))OB, OBmin( xi0 (k ))]

In the present work, as surface roughness and roundness were to be minimized (“the smaller-the-better” is a
characteristic of the actual sequence), the normalization of actual sequence should be performed as follows:

max( xi0 (k ))  xi0 (k )

xi* (k ) (3)
max( xi0 (k ))  min( xi0 (k ))

Where xi* (k ) is the value after grey relational generation (normalized value), max( xi0 (k )) and min( xi0 (k )) are the

largest and smallest values of xi0 (k ) for the kth response, respectively, k being 1 for surface roughness, 2 for

roundness and 3 for MRR.

The Table 3 shows the calculated data after grey relational generation. Greater normalized result implies better
performance and the finest normalized result should be equal to unity.

Step 2: Grey relational coefficient

The correlation between the best and actual experimental results, is denoted by grey relational coefficients. The
formulae to calculate the grey relational coefficient (ξi(k)) is given below:

' min  ]' max

[i (k ) (4)
'0i (k )  ]' max

0 [i (k ) d1 (5)

* (k ) and comparability sequence x* (k ) ,

where Δ0i(k) is the deviation sequence of reference sequence x0 i

'0i (k ) x0* (k )  xi* (k ) (6)

min min *
' min x (k )  x*j (k ) (7)
j  i k 0

max max *
' max x (k )  x*j (k ) (8)
j  i k 0
ζ is the distinctive coefficient (ζ Ԗ [0,1]) and is used to regulate the difference of the relational coefficient. In this
work ζ was taken as 0.5 and the grey relational coefficient calculated using Eq.(4) was given in table 3.
 P. Jayaraman and L. Mahesh kumar / Procedia Engineering 97 (2014) 197 – 204 201

Step 3: Grey relational grade

The formulae for calculating grey relational grade is as follows:

1 n
Di ¦ [ (k ) (9)
nk 1 i

where n is the number of performance characteristics. If the grey relational grade has greater value, it shows that the
concerned parameter combination is very nearer to the optimum value.

Step 4: Grey relational ordering

An order of 1 is allotted to greatest grey relational grade. Grey relational grades are calculated using Eq. (9) and
grey relational order was figured out in table 3. From table 3, we come to know that the control parameter’s setting
of 2(experiment 2) had the greatest grey relational grade and this indicates that experiment 2 was the optimal turning
factors setting for minimum surface roughness, roundness and MRR simultaneously among the chosen nine
experiments. The larger better S/N quality characteristics was considered for grey relational grade, since higher
multiple performance characteristics was our target. The level of a parameter with the highest S/N ratio gives the
optimal level. So the optimal process parameter setting for the multiple performance characteristic was v1f1d2. The
main effects plot for mean for GRG is shown in Fig. 1 and the main effects plot for S/N ratio for GRG is shown in
Fig. 2.

Table 3. Normalized values, grey relational coefficients and grey relational grades of responses

Tri Normalized values of responses Grey relational coefficients GRG Orde

al r
No Ra Ø MRR Ra Ø MRR

1 0.89796 1 0 0.83051 1 0.33333 0.72128 2

1
2 1 0.92781 0.34004 1 0.87384 0.43105 0.76830

8
3 0.10714 0.25936 0.7411 0.35897 0.40302 0.65885 0.47361

3
4 0.85714 0.94118 0.27185 0.77777 0.89474 0.40712 0.69321

7
5 0.2398 0.53743 0.7411 0.39676 0.51944 0.65885 0.52502

9
6 0.11735 0 0.48143 0.36162 0.33333 0.49088 0.39528

5
7 0.4949 0.86898 0.57709 0.49746 0.79237 0.54176 0.61053

4
8 0.44388 0.97326 0.43321 0.47343 0.94923 0.46870 0.63045

6
9 0 0.07219 1 0.33333 0.35019 1 0.56117
202 P. Jayaraman and L. Mahesh kumar / Procedia Engineering 97 (2014) 197 – 204

Main Effects Plot for Means

Data Means

cutting speed(m/min) feed rate(mm/rev)

0.70
0.65
0.60

0.55
Mean of Means

0.50

119.223 158.964 198.705 0.050 0.075 0.100

depth of cut(mm)
0.70

0.65
0.60
0.55
0.50

0.10 0.15 0.20

Fig. 1. Main effects plots for Mean for GRG

Main Effects Plot for SN ratios

Data Means

cutting speed(m/min) feed rate(mm/rev)

-3

-4

-5
Mean of SN ratios

-6

119.223 158.964 198.705 0.050 0.075 0.100

depth of cut(mm)
-3

-4

-5

-6

0.10 0.15 0.20

Signal-to-noise: Larger is better

Fig. 2. Main effects plot for S/N ratio for GRG.

Hence, the finest combination values for maximizing the multiple performance characteristics or grey relational
grade (GRG) were cutting speed of 119.22 m/min, feed rate of 0.05 mm/rev and depth of cut of 0.15 mm. The
response table for the means of grey relational grade is shown in Table. 4. The response table for the S/N ratios of
grey relational grade is shown in Table. 5. ANOVA output of the multiple performance characteristics was given in
Table. 6. From the analysis of this table, it could be concluded that feed rate followed by depth of cut and cutting
speed, are significantly affecting the grey relational grade.
 P. Jayaraman and L. Mahesh kumar / Procedia Engineering 97 (2014) 197 – 204 203

Table 4. Response Table for Means for GRG

Level v (m/min) f (mm/rev) d (mm)

1 0.6544 0.6750 0.5823

2 0.5378 0.6413 0.6742
3 0.6007 0.4767 0.5364
Delta 0.1166 0.1983 0.1378
Rank 3 1 2

Average mean= 0.5976

Table. 5. Response Table for S/N ratios for GRG

Level v (m/min) f (mm/rev) d (mm)

1 -3.873 -3.435 -4.969

2 -5.614 -3.964 -3.497
3 -4.437 -6.524 -5.458
Delta 1.741 3.088 1.961
Rank 3 1 2

Table. 6. ANOVA for grey relational grade (Multiple response characteristics)

Source DF Seq SS Adj SS Adj MS F P % contri-

bution
v 2
0.020422 0.020422 0.010211 104.83 0.009 17.35%

f 2
0.067553 0.067553 0.033777 346.76 0.003 57.365%

d 2
0.029555 0.029555 0.014778 151.71 0.007 25.105%

Error 2
0.000195 0.000195 0.000097

Total 8 0.117725

S = 0.00986947 R-Sq = 99.83% R-Sq(adj) = 99.34%

5. Confirmation Experiments

After the optimal level has been selected, one could predict the optimum response using the following equation:
n
J predicted J m  ¦ (Ji  J m ) (10)
i 1
where γm is the total mean S/N ratio, γ୧ is the mean S/N ratio at optimal level, n is the number of main design
parameters that affect the quality characteristics. The purpose of this confirmation experiment is to verify the
improvement in the quality characteristics. Based on the Eq.(10) the grey relational grade (GRG) is predicted for the
optimal combination of parameters (v1-f1-d2) and its value is 0.8084. Lastly confirmation test was conducted using
the optimum combination of parameters (v1-f1-d2). Table. 7. shows the comparison of predicted multiple
performance characteristics (GRG) with the actual one. The grey relational grade for the confirmation experiment is
found to be 0.7717.
204 P. Jayaraman and L. Mahesh kumar / Procedia Engineering 97 (2014) 197 – 204

Table 7: Results of Confirmation Experiment for GRG

Initial cutting Optimal cutting parameters

parameters Prediction Experiment

Level v2f2d2 v1f1d2 v1f1d2

GRG 0.4913 0.8084 0.7717

S/N Ratio -6.1731 -1.8475 -2.2510

Improvement of SN ratio = 3.9221 dB

6. Conclusions

The surface roughness (Ra), roundness (Ø) and material removal rate (MRR) were measured under different
cutting conditions for diverse combinations of machining parameters. The final conclusions arrived, at the end of
this work are as follows:

x
From this analysis, it is revealed that feed rate, depth of cut are prominent factors which affect the turning
of aluminium alloy. The feed rate (P = 57.365%) is the most influencing factor in determining the multiple
performance characteristics or grey relational grade (GRG) followed by depth of cut (P = 25.11%) and
cutting speed (P =17.35%).
x The best multiple performance characteristics was obtained with uncoated carbide insert when turning
aluminium alloy with the lower cutting speed of 119.22 m/min, lower feed rate of 0.05 mm/rev and
medium depth of cut of 0.15 mm with the estimated multiple performance characteristics (GRG) of 0.8084.
The experimental value of GRG for this combination of parameters is 0.7717.
x The percentage of error between the predicted and experimental values of the multiple performance
characteristics during the confirmation experiments is almost within 4.7 %.
x The improvement in the S/N ratio from the initial cutting parameters to the optimal cutting parameters is
3.9221 db for GRG.
x The value of multiple performance characteristics obtained from confirmation experiment is within the
95% confidence interval of the predicted optimum condition.
References

[1] P.G. Benardos, G.C. Vosniakos, Predicting surface roughness in machining: a review, International Journal of Machine Tool and
Manufacturing, Vol.43, 2007, pp. 833-844.
[2] I.P. Arbizu, C.J.L. Perez, Surface roughness prediction by factorial design of experiments in turning processes, Journal of Materials
Processing Technology, Vol.143, 2003, pp. 390-396.
[3] M.P. Groover, Fundamentals of Modern Manufacturing- Materials Processing and Systems, First Ed, Prentice-Hall , Englewood, Cliffs, 1966.
[4] Ashok Kumar Sahoo, Achyuta Nanda Baral, Arun Kumar Rout, B.C. Routra, Multi-objective optimization and predictive modelling of
surface roughness and material removal rate in turning using Grey Relational and Regression Analysis, Procedia Engineering, 38 (2012) 1606-
1627.
[5] Chorng-Jyh Tzeng, Yu-Hsin Lin, Yung-Kuang Yang, Ming-Chang Jeng, Optimization of turning operations with multiple performance
characteristics using the Taguchi method and Grey relational analysis, Journal of Materials Processing Technology, 209 (2009) 2753-2759.
[6] C.Y. Nian, W.H. Yang, Y.S. Tarng. Optimization of turning operations with multiple performance characteristics, Journal of Materials
Processing Technology, 95 (1999) 90-96.
[7] A. Noorul Haq, P. Marimuthu, R. Jeyapaul, Multi response optimization of Machining parameters of drilling Al/SiC Metal matrix composite
using grey relational analysis in the Taguchi method, International Journal of Advanced Manufacturing Technology, 37 (2008) 250-255.
[8] S. Datta, A. Bandyopadhyay, P.K. Pal, Modeling and optimization of features of bead geometry including percentage dilution in submerged
arc welding using mixture of fresh flux and fused slag. International Journal of Advanced Manufacturing Technology, 36 (2008) 1080-1090.
[9] C.L. Lin, Use of Taguchi method and grey relational analysis to optimize turning operations with multiple performance characteristics,
Materials & Manufacturing Processes, 19/2 (2004) 209-220.
[10] R. Jeyapaul, P. Shahabudeen, K. Krishnaiah, Quality management research by considering multi-response problems in the Taguchi method -
a review. International Journal of Advance Manufacturing Technology, 26 (2005) 1331–1337.
[11] P. J. Ross, Taguchi Techniques for Quality Engineering, First Ed, McGraw-Hill, New York, 1996.
[12] K. Palanikumar, Cutting parameters optimization for surface roughness in machining GFRP composites using Taguchi’s method, Journal of
Reinforced Plastic Composites, Vol.25, 2006, pp. 1739-1751.

View publication stats