Research Article

Multi-response optimization of turning

parameters for machining glass
fiberreinforced plastic composite rod

Advances in Mechanical Engineering

2015, Vol. 7(12) 110
The Author(s) 2015
DOI: 10.1177/1687814015620109
aime.sagepub.com

P Raveendran1 and P Marimuthu2

Abstract
Glass fiberreinforced plastics are extensively applied in engineering fields as a potential replacement to conventional
steels, owing to its corrosive resistance property and high specific strength. But machining is complicated due to its anisotropic properties and non-homogeneous structure. In machining processes, minimum surface roughness and less tool
wear are important factors influencing the quality of the surface, tool life, and productivity. Thus, the selection of tool
and optimizing machining parameters are essential for perfect finishing. The tool used in this study is TiCN/TiN coated.
The cutting parameters applied are cutting speed, feed, and depth of cut. As a dynamic approach, the multiple response
optimization is carried out using grey relational analysis and desirability function analysis for simultaneous evaluation.
These two methods are considered in optimization, as both are multiple criteria evaluation and not much complicated.
Analysis of variance is employed to classify the significant parameters affecting the response.
Keywords
Machining GFRP composites, surface roughness, tool wear, Taguchi, ANOVA, grey relational analysis, desirability function
analysis

Date received: 8 July 2015; accepted: 5 November 2015

Academic Editor: Liyuan Sheng

Introduction
In modern-day engineering, high demands are being
placed on components made of fiber-reinforced plastics
(FRP). Fiberglass composites are an economic alternative to stainless steel and other materials in highly corrosive industrial applications. Glass fiberreinforced
composite materials are used in various products
including sporting goods, marine bodies, plastic pipes,
storage containers, and construction works. Glass
fiberreinforced plastics (GFRP) are widely used in the
transport industries. GFRP rods are generally manufactured by pultrusion process and it needs further
machining for dimensional control. The machining of
GFRP composites is dissimilar from conventional
materials and it necessitates special considerations on
wear resistance of the tool. Surface evenness plays an

imperative role and is a factor of more importance in

the assessment of machining accuracy. Surface unevenness of a product could affect its functional attributes,
such as friction, wear, light reflection, transmission of
heat, and capability for lubrication. In order to get fine
surface finish and dimensional properties, it requires
optimization techniques to find best possible cutting
1

Department of Mechanical Engineering, Mahath Amma Institute of

Engineering & Technology, Ariyur, India
2
Department of Mechanical Engineering, Syed Ammal Engineering
College, Ramanathapuram, India
Corresponding author:
P Raveendran, Department of Mechanical Engineering, Mahath Amma
Institute of Engineering & Technology, Ariyur, Pudukkottai District
622101, Tamil Nadu, India.
Email: lnpraveendran@yahoo.com

Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 3.0 License
(http://www.creativecommons.org/licenses/by/3.0/) which permits any use, reproduction and distribution of the work without
further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/
open-access-at-sage).
Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

Advances in Mechanical Engineering

parameters and theoretical models for predictions.1 The

anisotropic property of fiber-reinforced composite
could be brought into more efficient usage by organizing the machining parameters. This provides a wide
range of choice for the manufacturers. For achieving
the highest performance for specified requirements, it is
necessary to find optimum parameters for machining
composites and to select the appropriate parameter values.2 Achieving an acceptable surface quality with conventional methods of machining has been found
extremely difficult due to the anisotropic and heterogeneous nature of these materials. Excessive tool wear
(TW) is prevalent and frequently induces fiber pullout
and surface ply delamination in the component part.
This limitation has provided both academic and industrial motivation for research on the application of traditional methods of machining to reinforced polymers.3
There are many studies on the machining of GFRPs.
Isxk4 obtained results on machining of the unidirectional glass fiberreinforced composite (UD-GFRP)
with cermet cutting tool and suggested optimum cutting
parameters to obtain improved surface quality. Also,
Palanikumar5 took efforts and found optimum cutting
parameters for reducing surface roughness (SR) using
Taguchis method. It presents a simple and systematic
approach toward optimizing design. Palanikumar and
P Davim6 formed a mathematical model to calculate
the wear in tool used for machining glass fiber
reinforced composite by means of regression analysis,
and to develop the model, analysis of variance
(ANOVA) is applied. Naveen Sait et al.7 performed
turning experiments with K20 grade cemented carbide
tool on filament wound GFRP pipes produced by hand
layup process. It is based on L18 orthogonal array to
present the influence of machining parameter.
Adam Khan and Senthil Kumar8 were made an
attempt to analyze the wear mechanism of alumina cutting tools while machining GFRP composite material
prepared using a filament winding process. The machinability of GFRP by means of tools made of various
materials and geometries was investigated experimentally by Lee9 and revealed that proper selection of cutting tool material. S Kumar et al.10 applied Taguchi
method to find optimum process parameters for turning
UD-GFRP rod using polycrystalline diamond (PCD)
tool. G Meenu and S Kumar11 developed a SR prediction model for the machining of UD-GFRP composite
using regression modeling and optimization by simulated annealing and revealed that feed is the primary
factor and has great influence on SR and then it is followed by cutting speed. They have also made an
attempt to model the SR and material removal rate
using principal component analysis to determine
the influence of cutting parameters on UD-GFRP composite with PCD tool.12 Chandrasekaran et al.13
have applied grey relational analysis (GRA) for

multi-response optimization of machining parameters

for turning stainless steel. Kumar et al.14 used DistanceBased Pareto Genetic Algorithm (DBPGA) approach
to optimize tangential and feed force while turning UDGFRP composite with various process parameters
including feed rate, cutting speed, and depth of cut and
investigated the influencing factor. NS Vyas et al.15
investigated the influence of drilling parameters, such
as cutting speed, feed, and point angle on delamination,
thrust force, and circularity produced when drilling
GFRP composite as its application is getting increased.
From the earlier work done and literature, it shows
that most of the researchers have done the experiments
on filament winding GFRP pipes produced by hand
layup process and machined with costly tools such as
cubic boron nitride (CBN), PCD, cermet, and alumina
cutting tools and gave attention on individual objective. Studying the level of performance in using the
coated tools for turning GFRP rods and collective
objectives to minimize SR and TW has got less attention. So, in this study of research, low-priced TiCN/
TiN-coated tool has been used for turning pultruded
GFRP rods in dry conditions. The experiments were
intended based on Taguchis L27 orthogonal array, by
predetermined cutting conditions. The considering
parameters of cutting speed, feed, and depth of cut
have been optimized with multiple responses using
GRA and also using desirability function approach
with an aim to reduce TW and SR. To identify the significant parameters affecting the response, ANOVA
was applied.

Materials and methods

Materials
GFRP rods produced by pultrusion method, shown in
Figure 1, are used in this study. The diameter and
length of specimen are 40 and 280 mm, respectively,
having an L/D ratio 7, which were used for the

Figure 1. GFRP rod specimen.

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

Raveendran and Marimuthu

3
Table 2. Composition and coating specification of the tool
material.
Composition
Grain size
Hardness
Specification of coating

Co 9.5%, composite carbides

6.5%, Tungsten Carbide (WC) rest
12 mm
HV 1400
CVD 2 Ti (C, N) + Ti (C, N)
+ TiN + Ti (N, B) + Ti (C, N)
+ TiN; 6 mm

HV: Vickers hardness.

Figure 2. Scanning electron microscope image of the

composite.

Table 1. Properties of GFRP material.

Glass fiber
Glass content (%)
Matrix material
Epoxy content (%)
Density (g/cm3)
Tensile strength (N/mm2)
Shearing strength (N/mm2)

E-glass
7580
Epoxy resin
2025
1.9
1200
50

multiple response characteristics in machining metal

matrix composite, and this method for the optimization
of the multi-response problems is a very useful tool for
converting multi-responses into single response problem.16 The step-by-step procedure followed by Noorul
Haq et al.16 for converting multi-responses into single
response problem is used for the analysis:
Step 1. Signal-to-noise (S/N) ratio for the corresponding responses is calculated from equation (1).
This is applied in the parts where maximizing the
quality characteristic of attention is sought. Here,
the smaller-the-better-type problem is intended to
minimize the distinctive factors

S=Nratioh =  10 log10

GFRP: glass fiberreinforced plastics.

experiments. The work piece was turned for 90 mm in

all the trials. The reinforced fiber in the rod is E-glass
and resin is epoxy. The scanning electron microscope
image of the cross-section of the composite is shown in
Figure 2. The properties of composite rod samples used
are shown in Table 1.

Tool material
Machining studies were carried out in turning operation
of GFRP composite rod using chemical vapor deposition (CVD)-coated carbide tool of CERATIZIT make
of grade CTC 2135 HC-M30 and the type SNMG
120408 EN-TMR. This tool was preferred due to its
excellent toughness and wear resistance. The composition and coating specification of the tool material are
shown in Table 2. The cutting tool geometry is shown
in Figure 3.

n
1X
y2
n i = 1 ij

!
1

where n is the number of tests and yij is the response

value observed in ith experiment at jth test.
Step 2. To reduce the variability and to avoid the
results of adopting different units, yij is normalized
as Zij (0  Zij  1) by the following formula. A
proper value is subtracted from the values at same
array to compose the value of the array approximate to 1. The sensitivity of the normalization process in the sequencing results has been analyzed, as
the rank is affected by the process of normalization.
Therefore, the S/N ratio value is adopted when normalizing data in GRA. Equation (2). shows the
smaller-the-better characteristic


max yij , i = 1, 2, . . . n  yij




zij =
max yij , i = 1, 2, . . . n  min yij , i = 1, 2, . . . n
2
Step 3. For the normalized values of S/N ratio, grey
relational coefficient is considered from equation (3)

GRA
GRA is a measurement technique in grey system theory
that analyzes the degree of relation in a discrete
sequence. GRA had been used for optimization of

g yo k , yi k =

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

D min + jD max
Doj k + jD max

Advances in Mechanical Engineering

Figure 3. Cutting tool geometry.

where yo(k) is the reference sequence (yo(k) = 1,

k = 1, 2, ., m); yj(k) is the specific comparison
sequence; Doj = kyo (k)  yi (k)k, the absolute value of
and
yj(k);
the
variation
yo(k)
between
value
Dmin = min8j2i min8k yo (k)  yj(k), the smallest

of yj(k); Dmax = max8j2i max8k yo (k)  yj (k), the largest value of yj(k); and z is the distinguishing coefficient
which is defined in the domain 0  z  1 (the value
adjusted is based on the realistic needs of the system).
Step 4. Grey relational grade is made by equation
(4)
j =
g

m
1X
g
k i = 1 ij

j is the grey relational grade of jth experiment

where g
and k is the number of performance characteristics.

Step 2. The index of individual desirability for entire

responses can be united to form a single value
named composite desirability (dG) by equation (6)
p
dG = d1w1  d2w2      diwi

where di represents the individual desirability property

Yi, wi refers to the weight of the property, Yi in the
composite desirability, and w is the summation of the
individual weights.
Step 3. At last, the optimal parameter and its level
of combinations are to be determined: The higher
the composite desirability value, the better the product quality. Thus, based on the composite desirability (dG), the effects of parameter and the optimum
level for each parameter are estimated.

Experimental details
Desirability function analysis in the Taguchi method
Desirability function analysis (DFA) is one of the most
extensively used methods for the optimization of multiresponses problems. DFA is used to change the multiresponses problems into single response problems. As a
result, optimization of the complicated multi-response
problems could be converted into optimization of a single response problem termed composite desirability.17
Step 1. Individual desirability function (di) for the
corresponding responses has to be determined. For
the smaller-the-better, the desirability function can be
expressed as in equation (5). The value of ^y is
expected to be the smaller-the-better while ^y is less
than a particular criterion value, the desirability value
will be equal to 1; if the ^y exceeds a particular criterion value, the desirability value will be equal to 0
8
1,
>
r
<  ^y  y
max
di =
,
>
: ymin  ymax
0

if ^y  ymin
if ymin  ^y  ymax , r  0 5
if ^y  ymax

where the ymin denotes the lower tolerance limit of ^y,

the ymax representing the upper tolerance limit of ^y, and
r refers to the weight.

The turning experiments on GFRP rods were conducted in dry cutting conditions on KIRLOSKARmake TURNMASTER-35 center lathe with variable
speed and feed drive and 2.2-kW spindle power DC
motor. The turning operation was carried out using
SNMG 120408 EN-TMR, CVD-coated carbide tool
insert of grade CTC 2135. This tool was preferred due
to its excellent toughness and wear resistance.
KISTLER TYPE 9257B multi-component dynamometer, measuring cutting forces, was setup on the
lathe. Figure 4 shows the experimental setup for turning GFRP rod using a lathe.
Measurements of SR were taken at least five times,
and the mean value is recorded as the Ra value. The
Mitutoy- make SR tester of model SJ 210 had been
used to measure the SR as shown in Figure 5. The TW
was measured using optical tool makers microscope as
shown in Figure 6. The image of the TW after turning
GFRP is shown in Figure 7.
The most important cutting parameters affecting
surface finish are cutting speed, feed, depth of cut, and
wet and dry cutting conditions. The main objective of
this study is to establish a relation among cutting speed,
feed, and depth of cut responds on SR and TW. The
range of cutting parameter values is chosen based on
tool manufactures recommendation. In this study,

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

Raveendran and Marimuthu

Figure 4. Experimental setup.

Figure 6. Tool wear measured using optical tool makers

microscope.

Figure 7. Image of the tools after turning GFRP.

Figure 5. Surface roughness testing.

three parameters set at three levels have been selected

for experimentation. The turning parameters and their
levels taken for both cases are presented in Table 3.

Result and discussion

Multiple response optimizations for turning GFRP
using GRA
The S/N ratio for SR and TW is computed using equation (1). Normalizing the S/N ratio values for SR and
TW is computed using equation (2). Grey relational
coefficient for the normalized S/N ratio values was calculated using equation (3), and the grey relational grade
can be computed by equation (4). Finally, the grades
were considered for optimizing the multi-response parameter design problem. The results are given in Table 4.
The higher grey relational grade gives the better product quality and on its basis, the factor effect can be
calculated and the optimal level for each controllable
factor can be determined. The main effects are

tabulated in Table 5, and considering maximization of

grade values in Table 4, the optimal parameter conditions obtained are v1, f1, and a1. The cutting speed set
as minimum level (50 m/min), the feed set as minimum
level (0.08 mm/rev), and depth of cut was set to the
maximum level (0.5 mm). The ANOVA of SR for
machining GFRP is given in Table 6, and it clearly
illustrates that the most significant factor that affects
the SR with a p value of 0.002 is depth of cut and second by the feed with a p value of 0.126. The ANOVA
for TW for turning GFRP is tabulated in Table 7, and
it clearly illustrates that the depth of cut most significantly affects the TW with a p value of 0.000 followed
by cutting speed with a p value of 0.022.

Multi-response optimization for turning GFRP using

DFA
For every response, the individual desirability (di) is
calculated depending upon the required quality characteristics. As all the responses are having minimization
objective, the smaller-the-better-type equation is chosen. The calculated individual desirability for each

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

Advances in Mechanical Engineering

Table 3. Machining parameters and levels.

Parameter

Designation

Level 1

Level 2

Level 3

Cutting speed (m/min)

Feed (mm/rev)
Depth of cut (mm)

v
f
a

50
0.08
0.5

75
0.12
1

100
0.16
1.5

Table 4. Grey relational analyses for GFRP.

Trial no.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

Experimental

S/N ratios

SR (mm)

TW (mm)

SR

2.942
3.047
4.079
3.062
2.779
3.433
3.304
2.686
4.129
3.052
3.242
3.058
2.864
3.126
4.385
2.806
3.306
3.411
2.342
2.565
3.58
3.297
3.434
3.967
3.694
4.07
4.185

0.01
0.02
0.025
0.015
0.02
0.04
0.02
0.025
0.05
0.02
0.03
0.035
0.025
0.03
0.035
0.025
0.03
0.035
0.025
0.03
0.035
0.025
0.03
0.035
0.025
0.0375
0.0425

29.3729
29.6774
212.211
29.7201
28.8778
210.714
210.381
28.5821
212.317
29.6917
210.216
29.7087
29.1395
29.8998
212.839
28.9618
210.386
210.658
27.3917
28.1817
211.078
210.362
210.716
211.969
211.35
212.192
212.434

Normalized values of S/N ratios

Grey relational coefficient

TW

SR

TW

SR

TW

40
33.9794
32.0412
36.4782
33.9794
27.9588
33.9794
32.0412
26.0206
33.9794
30.4576
29.1186
32.0412
30.4576
29.1186
32.0412
30.4576
29.1186
32.0412
30.4576
29.1186
32.0412
30.4576
29.1186
32.0412
28.5194
27.4322

0.393
0.453
0.956
0.462
0.295
0.659
0.593
0.236
0.977
0.456
0.560
0.460
0.347
0.497
1.080
0.311
0.594
0.648
0.000
0.157
0.731
0.589
0.659
0.908
0.785
0.952
1.000

0.000
0.431
0.569
0.252
0.431
0.861
0.431
0.569
1.000
0.431
0.683
0.778
0.569
0.683
0.778
0.569
0.683
0.778
0.569
0.683
0.778
0.569
0.683
0.778
0.569
0.821
0.899

0.560
0.524
0.343
0.520
0.629
0.431
0.458
0.679
0.339
0.523
0.472
0.521
0.591
0.501
0.316
0.616
0.457
0.436
1.000
0.761
0.406
0.459
0.431
0.355
0.389
0.344
0.333

1.000
0.537
0.468
0.665
0.537
0.367
0.537
0.468
0.333
0.537
0.423
0.391
0.468
0.423
0.391
0.468
0.423
0.391
0.468
0.423
0.391
0.468
0.423
0.391
0.468
0.378
0.357

Grey grade

0.780
0.531
0.406
0.592
0.583
0.399
0.497
0.573
0.336
0.530
0.447
0.456
0.529
0.462
0.354
0.542
0.440
0.413
0.734
0.592
0.399
0.463
0.427
0.373
0.428
0.361
0.345

GFRP: glass fiberreinforced plastics; S/N: signal to noise; SR: surface roughness; TW: tool wear.

Level

the minimization of composite desirability value, the

optimal parameter condition is obtained as v3f1a1 for
turning GFRP.

v
f
a

0.522
0.542
0.566

0.464
0.465
0.491

0.458
0.437
0.387

Confirmation test

Table 5. Main effects on grey grade for GFRP.

GFRP: glass fiberreinforced plastics; Bold figures represent the higher

value in the table.

quality characteristics with equation (5), is shown in

Table 8. With equation (6), composite desirability values (dG) are calculated. Finally, these values are taken
for optimization of multi-response parameter design
problem. The results are shown in Table 9. Considering

The objective of performing the confirmation test is to

identify the selected machining parameters, which will
make better results than the first part of the experiments.
The confirmation experiments are used to confirm that
the parameters and levels are chosen from an experiment.
If the average of the results arrived in the confirmation
test is within the range of confidence limits, then the
appropriate levels and significant parameters for obtaining the desired results are properly chosen.

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

Raveendran and Marimuthu

Table 6. Analysis of variance for SR.

Source

DF

SS

MS

v
f
a
Error
Total

2
2
2
20
26

0.2368
0.7805
3.0954
3.3839
7.4965

0.1184
0.3902
1.5477
0.1692

0.70
2.31
9.15

0.508
0.126
0.002

SR: surface roughness; DF: degrees of freedom; SS: sum square; MS: mean square.

Table 7. Analysis of variance for TW.

Source

DF

SS

MS

v
f
a
Error
Total

2
2
2
20
26

0.00020741
0.00020185
0.00113380
0.00044907
0.00199213

0.00010370
0.00010093
0.00056690
0.00002245

4.62
4.49
25.25

0.022
0.024
0.000

TW: tool wear; DF: degrees of freedom; SS: sum square; MS: mean square.

Table 8. Individual desirability (di) and composite desirability (dG).

Trial no.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

Individual desirability (di)

Composite desirability (dG)

TW

SR

1.000
0.750
0.625
0.875
0.750
0.250
0.750
0.625
0.000
0.750
0.500
0.375
0.625
0.500
0.375
0.625
0.500
0.375
0.625
0.500
0.375
0.625
0.500
0.375
0.625
0.313
0.188

0.706
0.655
0.150
0.648
0.786
0.466
0.529
0.832
0.125
0.652
0.559
0.650
0.744
0.616
0.000
0.773
0.528
0.477
1.000
0.891
0.394
0.533
0.465
0.205
0.338
0.154
0.098

TW: tool wear; SR: surface roughness.

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

0.957
0.915
0.744
0.931
0.936
0.764
0.891
0.921
0.000
0.915
0.853
0.838
0.909
0.863
0.000
0.913
0.847
0.806
0.943
0.904
0.787
0.872
0.833
0.725
0.823
0.684
0.607

Advances in Mechanical Engineering

Table 9. Parameters effects for composite desirability (dG).

Parameters

v
f
a

Levels
1

0.784
0.873
0.906

0.772
0.759
0.862

0.798
0.721
0.586

corresponding grey relational grade is 0.78 dB. The

result is obtained within the CI of the expected optimum performance characteristics.
The optimal TW is estimated at the selected optimal
setting of parameters A1B1C1. The predictable mean at
the optimum settings could be obtained from equation
(12)
mRa = A1 + B1 + C1  2  TRa

Bold figures represent the higher value in the table.

The expected mean of weighted grey relational grade

on a confirmation experiment is as equation (13)

Confirmation test of SR and TW based on GRA

The optimal SR is predicted in the selected optimal setting of parameters, and A1B1C1 is the optimum parameter. The predictable mean at the optimum settings
could be obtained from equation (7)
mRa = A1 + B1 + C1  2  TRa

Equation (8) is referred for the prediction of mean in

weighted grey relational grade of confirmation test
hRa = hA1 + hB1 + hC1  2  hTRa

13

where hTRa is grey relational grade overall mean of

TW, and hV1, hF1, and hD1 are the mean weighted grey
relational grade corresponding to cutting speed v1, feed
rate f1, and depth cut a1.
The CI for the expected results from the confirmation test is calculated using equation (14)
s


1
1
+
CI = Fa 1, fe Ve
neff
R

14

where Fa (1, fe) is the F ratio, at a significant level of

a%, a is the risk, fe is the error degrees of freedom, Ve
is the error mean square, neff is the effective total number of experiments, and R is the number of confirmation tests
neff =

where Fa (1, fe) is the F ratio at a significant level of a%,

a is the risk, fe be the error degrees of freedom, Ve is the
error mean square, neff is the effective total number of
tests, and R is the number of confirmation experiments
neff =
Total number of observation
1 + total degree of freedom associated with items used

10
Therefore, the 95% CI must be given by
hRa  CI  hRa  hRa + CI

hRa = hA1 + hB1 + hC1  2  hTRa

where hTRa is grey relational grade overall mean of SR,

and hV1, hF1, and hD1 are the mean weighted grey relational grade of corresponding cutting speed v1, feed rate
f1, and depth cut a1.
Confidence interval (CI) for the predictable results
from the confirmation experiment is formed using
equation (9)
s


1
1
+
CI = Fa 1, fe Ve
neff
R

12

11

hRa = hA1 + hB1 + hC1  2  hTRa

= 0:522 + 0:542 + 0:566  2  0:481 = 0:668B
0:668  1:19  hRa  0:668 + 1:19
 0:522dB  hRa  1:858 dB
The confirmation test is conducted at the optimum
settings to confirm the SR for turning. SR value found
at the optimum setting of A1B1C1 is 2.942 and the

Total number of observation

1 + total degrees of freedom associated with items used

Therefore, the 95% CI must be given by

hRa  CI  hRa  hRa + CI
hRa = hA1 + hB1 + hC1  2  hTRa
= 0:522 + 0:542 + 0:566  2  0:481 = 0:668
0:668  1  hRa  0:668 + 1
 0:332  hRa  1:668
The confirmation experiment is performed at the
optimum settings to check the TW for turning. TW
value found at the optimum setting of A1B1C1 is 0.01,
and the corresponding grey relational grade is 0.78 dB.
The result is obtained within the CI of the expected
optimum performance characteristics.

Confirmation experiment for SR and TW based on

DFA
The confirmation experiment is performed at the optimum settings to check the SR for turning. SR value

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

Raveendran and Marimuthu

found at the optimum setting of A3B1C1 is 2.342, and

corresponding composite desirability is 0.943. The
obtained result is within the CI of the expected optimum performance characteristics. TW value at the
optimum setting of A3B1C1 was found as 0.025, and
the corresponding composite desirability is 0.943. The
obtained result is within the CI of the expected optimum performance characteristics.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.

Funding
The author(s) received no financial support for the research,
authorship, and/or publication of this article.

Conclusion

References

This present investigation is focused on effective turning of GFRP using TiCN/TiN-coated tool with multiresponse optimization of cutting parameters. From this
study, using GRA, DFA, and ANOVA, the following
results can be concluded:

1. Palanikumar K. Application of Taguchi and response

surface methodologies for surface roughness in machining glass fiber reinforced plastics by PCD tooling. Int J
Adv Manuf Tech 2008; 36: 1927.
2. Isxk B and Kentli A. Multicriteria optimization of cutting
parameters in turning of UD-GFRP materials considering sensitivity. Int J Adv Manuf Tech 2009; 44:
11441153.
3. Bagci E and Isxk B. Investigation of surface roughness in
turning unidirectional GFRP composites by using RS
methodology and ANN. Int J Adv Manuf Tech 2006; 31:
1017.
4. Isxk B. Experimental investigations of surface roughness
in orthogonal turning of unidirectional glass-fiber reinforced plastic composites. Int J Adv Manuf Tech 2008;
37: 4248.
5. Palanikumar K. Cutting parameters optimization for
surface roughness in machining of GFRP composites
using Taguchis method. J Reinf Plast Comp 2006; 25:
17391751.
6. Palanikumar K and Paulo Davim J. Mathematical model
to predict tool wear on the machining of glass fibre reinforced plastic composites. Mater Design 2007; 28:
20082014.
7. Naveen Sait A, Aravindan S and Noorul Haq A. Optimization of machining parameters of glass-fibre-reinforced
plastic (GFRP) pipes by desirability function analysis
using Taguchi technique. Int J Adv Manuf Tech 2009; 43:
581589.
8. Adam Khan M and Senthil Kumar A. Machinability of
glass fibre reinforced plastic composite using aluminabased ceramic cutting tools. J Manuf Process 2011; 13:
6773.
9. Lee E-S. Precision machining of glass fibre reinforced
plastics with respect to tool characteristics. Int J Adv
Manuf Tech 2001; 17: 791798.
10. Kumar S, Meenu G, Satsangi PS, et al. Multiple regression model for cutting force in turning UD-GFRP using
polycrystalline diamond cutting tool. IJAET 2012; III:
108115.
11. Gupta M and Kumar S. Prediction of surface roughness
in turning of UD-GFRP using mathematical model and
simulated annealing. IJAET 2013; IV: 417426.
12. Gupta M and Kumar S. Investigation of surface roughness and MRR for turning of UD-GFRP using PCA and
Taguchi method. Eng Sci Tech 2015; 18: 7081.
13. Chandrasekaran K, Marimuthu P, Raja L, et al. Multi
response optimization of machining parameters for

1.

2.

3.

4.

From the results of using ANOVA, the significant cutting parameters affecting the SR is
found as depth of cut followed by feed, and the
significant cutting parameters affecting the TW
is found as depth of cut followed by cutting
speed.
Multi-response optimization using the GRA
was performed for turning GFRP and found
the optimum setting of cutting speed at 50 m/
min, feed at 0.08 mm/rev, and depth of cut as
0.5 mm for minimization of SR and TW.
For the optimization of multi-response problems in turning GFRP, DFA is a very useful
tool, and the optimal setting parameters for
minimization of SR and TW are found as cutting speed set at 100 m/min, feed set at 0.08 mm/
rev, and depth of cut set as 0.5 mm.
The most optimal among these two methods in
this study could be found by equation (15)

ma  mc + mb  md 
m c + m d

15

where ma are mb are the predictable means of S/N value

of SR and TW, respectively, at optimal condition of
the used method. mc and md are the predictable means
of S/N value of SR and TW, respectively, at one normal condition and
m=A+B+C  2  T
where A, B, and C are the average S/N value at optimal
speed, feed, and depth of cut, respectively, and T is the
total average of S/N value. From this equation, the percentage of improved optimal condition to the normal
one taken is 32.89% when using GRA and 17.42%
when using DFA.

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016

10

Advances in Mechanical Engineering

turning stainless steel using coated tools. Appl Mech

Mater 2014; 573: 644648.
14. Kumar S, Gupta M and Satsangi PS. Multiple-response
optimization of cutting forces in turning of UD-GFRP
composite using Distance-Based Pareto Genetic Algorithm approach. Eng Sci Tech 2015, http://dx.doi.org/
10.1016/j.jestch.2015.04.010
15. Vyas NS, Patel AR and Gajera HM. An investigation on
thrust force and circularity of GFRP sheet by applying
regression analysis. IJSRD 2015; 3: 278283.

16. Noorul Haq A, Marimuthu P and Jeyapaul R. Multi

response optimization of machining parameters of drilling Al/SiC metal matrix composite using grey relational
analysis in the Taguchi method. Int J Adv Manuf Tech
2008; 37: 250255.
17. Naveen Sait A, Aravindan S and Noorul Haq A. Optimization of machining parameters of glass-fibre-reinforced
plastic pipes by desirability function analysis using Taguchi technique. Int J Adv Manuf Tech 2009; 43: 581589.

Downloaded from ade.sagepub.com at PEC UNIVERSITY OF TECHNOLOGY on January 26, 2016