Journal of Minerals and Materials Characterization and Engineering, 2012, 11, 876-884
Published Online September 2012 (http://www.SciRP.org/journal/jmmce)
Multiple Response Optimization of Three-Body Abrasive
Wear Behaviour of Graphite Filled Carbon-Epoxy
Composites Using Grey-Based Taguchi Approach
K. M. Subbaya1*, N. Rajendra1, Y. S. Varadarajan1, B. Suresha2
1Department of Industrial & Production Engineering, The National Institute of Engineering, Mysore, India
2Department of Mechanical Engineering, The National Institute of Engineering, Mysore, India
Received April 11, 2012; revised May 20, 2012; accepted June 8, 2012
ABSTRACT
The three-body abrasive wear behaviour of carbon fabric reinforced epoxy (C-E) composites has been evaluated by the
addition of graphite (G) particles as a secondary reinforcement. Three-body abrasive wear test were conducted using
dry sand rubber wheel abrasion tester as per ASTM G-65 with three process parameters load, abrading distance and
filler content. To assess the abrasive wear behaviour of particulate filled C-E composites satisfying multiple perform-
ance measure, grey-based Taguchi approach has been adopted. The experiments were designed according to Taguchi’s
orthogonal array (L27). The grey relational analysis was applied to convert a multi response process optimization to a
single response. Using analysis of variance, significant contributions of process parameters have been determined. The
results indicate that the addition of graphite particles into C-E composite increased the wear resistance considerably. It
was observed that highest wear resistance of C-E composite was achieved with incorporation of 10 wt% of graphite
filler. Results indicate that the filler content and grit size of abrasive paper were found to be the most significant factor
which has influence on the abrasive wear of C-E composite. The worn surface features were examined through scan-
ning electron microscope to probe the wear mechanism.
Keywords: Composites; Three-Body Abrasion; Grey Relational Analysis; Scanning Electron Microscope
1. Introduction
Engineering polymers have attracted much interest in
structural applications for many years, because of their
superior properties such as light weight, strong, ease of
fabrication and low cost [1]. In recent years much atten-
tion has been devoted to explore the potential advantage
of thermoset matrices for new polymer composites [2].
One of such matrix is epoxy which has found a special
place in the family of thermoset engineering polymers
because of its excellent mechanical properties with
chemical and corrosion resistance [3]. Polymer compos-
ites are subjected to abrasive wear in many applications.
Abrasive wear as defined by ASTM, is due to hard parti-
cles or hard protuberances that are forced against and
move along a solid surface. Wear, in turn, is defined as
damage to a solid surface that generally involves pro-
gressive loss of material and is due to relative motion
between that surface and a contacting substance or sub-
stances. Abrasive wear occurs when hard asperities on
one surface move across a softer surface under load, pen-
etrates and remove material from the softer surface leav-
ing grooves [4]. The tribological performance of com-
posite material is usually related with the properties of
their reinforcement [5]. One of the traditional concepts to
improve the friction and wear behaviour of polymeric
materials is to enhance their hardness, stiffness and their
compressive strength and to reduce adhesion to the coun-
terpart material [6-9]. Many investigations have shown
that the incorporation of fiber reinforcement improved
the wear resistance and reduced the coefficient of friction
[10]. Mody et al. [11] in their investigation showed the
simultaneous existence of parallel and perpendicular
oriented carbon fibers in a woven configuration leads to a
synergic effect on the enhancement of the wear resis-
tance of composite. Carbon fiber is graphitized carbon
with the hexagonal planes of its crystals aligned perpen-
dicular to the fiber axis. The lubricating function of the
graphitized carbon is thought to be responsible for the
reduction of the friction coefficient and wear rate as the
composite slide against the matting surface. Besides the
lubricating function, carbon fiber also enhances the
thermal conductivity and mechanical properties of poly-
*Corresponding author.

K. M. SUBBAYA ET AL. 877
mer matrix which is believed to be beneficial to the wear
resistance as well [12].
A notable advance in the polymer industry has been
the use of fiber and particulate fillers as reinforcements
in polymer matrices. Many researchers found that glass
fiber and carbon fibers were effective reinforcements for
distinct effect on the friction and wear behaviour of
polymer composites [13]. The advantage of glass fiber
includes low cost, high strength as well as high chemical
resistance. On the other hand its low modulus, low fatigue
resistance. Its self-abrasiveness and its poor interaction
with some matrices limit this type of fiber for some tri-
bological applications. Carbon fibers not only show a
high strength and modulus but also have an excellent
heat stability and chemical inertness [14]. Furthermore,
carbon fiber is preferred as reinforcement/filler in tribo-
composites because it imparts reinforcement, conductiv-
ity and better friction. However, these fibers have some
disadvantages such as difficulty of uniform distribution,
poor interfacial adhesion etc. [15]. The modification of
tribological behaviour of fiber reinforced polymers by
the addition of functional fillers has been reported by
many researchers [16,17]. Particulate fillers are of con-
siderable interest, not only in the economic point of view
but also as modifiers especially in respect to improve
physical properties of polymer. It is well documented in
the literature that majority of fillers have a positive in-
fluence on mechanical properties [18]. Inorganic fillers
like SiC, SiO2, Al2O3, MoS2, graphite, addition into poly-
mer composite can promote hardness, wear resistance
and thin film formation on the counterpart during sliding
wear process. Most of the findings are based on ran-
domly oriented, unidirectional oriented or woven fabric
reinforced polymer composites. Surface treatment of
carbon fibers [19-21] is commonly used to improve the
fiber-matrix adhesion, interfacial shear strength, etc.
However, only few works are focussed on the effect of
surface modification of filler on abrasive wear properties
of fiber reinforced polymer composites. It is reported that
silane coupling agent promotes the interfacial adhesion
and interfacial toughness between glass fibers and
polytetrafluoroethylene (PTFE) and largely enhance the
tensile and tribological properties of glass PTFE com-
posites [22,23].
Majority of research studied detailed experimental work
with effect of one factor by keeping all other factors
fixed, this approach is not advisable because in actual
environment there will be combined effects of interacting
factors influencing the abrasive wear. Hence in this study
an attempt is being made to study the interacting effects
of factors along with the main effect. To achieve this, de-
sign of experiments based on Taguchi method is adopted.
This method is advocated by Taguchi and Konishi [24].
Taguchi’s technique uses special design of orthogonal ar-
rays to study the entire parameter space with only a small
number of experiments. Taguchi’s technique also helps
in optimizing the critical parameters [25].
Carbon fabric reinforcement and graphite filler are
good choice since in many applications both high
modulus and high strength are desired. In the present
work, the carbon fabric and graphite particles were
treated with silane coupling agent and their effect on
Tree-body abrasive wear of epoxy composites have been
evaluated using multi response grey-based Taguchi
method. The purpose of this work is to study the abrasion
resistance of C-E composite with silane treated graphite
particles under three-body abrasion using silica sand of
angular shape. It is expected that this research work can
be helpful to the use of epoxy composites in practice.
2. Experiment
2.1. Materials
The composites investigated in the present study, consists
of bi-directional carbon fabric of about 6 - 8 μm diameter
as reinforcement (T300). The carbon fiber surface is
treated with silane coupling agent to improve adhesion
between the matrix and fiber. Epoxy resin (LY556) with
room temperature curing hardener (HY951 grade) with
diluent DY021 (supplied by Hindustan Ciba Geigy) mix
was employed for the matrix material. The graphite par-
ticles of average size of about 20 - 25 µm were employed
as filler material. The details of composites selected, me-
asured density and hardness are listed in Table 1. Three-
body wear test samples of size 25 × 75 × 2.5 mm3 were
prepared from the laminate using a diamond tipped cutter.
2.2. Barcol Hardness
ASTM D2583 Barcol hardness test method [26] is used
to determine the hardness of reinforced thermosets. The
specimen is placed under the indentor of Barcol hardness
tester and a uniform pressure is applied to the specimen
until the dial indication reaches a maximum. The depth
of penetration is converted in to absolute Barcol numbers.
At least three specimens of each composition were tested
and the average values were listed in Table 1.
Table 1. Composites selected for study.
Material (designation)Epoxy
(wt%)
Filler
(wt%)
Density
(g/cm3)
Hardness
(barcol)
Carbon-epoxy (C-E) 40 - 1.412 73
Graphite filled
carbon-epoxy (5G-C-E)35 5 1.431 74
Graphite filled
carbon-epoxy (10G-C-E)30 10 1.465 76

K. M. SUBBAYA ET AL.
878
2.3. Three-Body Wear Test
Three-body abrasive wear experiment was conducted us-
ing dry sand/rubber wheel abrasion test set up as per
ASTM G65 standard [27]. The schematic diagram of the
test set up is shown in Figure 1.
Initially the samples were cleaned with acetone, dried
and its initial weight was determined using a high preci-
sion digital electronic balance (0.0001 g accuracy). The
naturally deposited silica sand of grain size 212 µm in
angular particle shape was used as abrasive. The abrasive
was fed at the contacting face between the rotating rubber
wheel and test sample. The tests were conducted at a
rotational speed of 200 rpm. The rate of feeding abrasive
was maintained at 255 ± 5 g/min. The parameters and
levels are given in Table 2. The experiment was carried
out for three loads viz., 11 N, 23 N and 35 N. The abrad-
ing distances were maintained at 300 m, 600 m and 900 m.
After testing, the specimens were removed, cleaned, dried
and again weighed. The difference between the initial
weight and final weight of the specimen was computed,
this gives the material wear in terms of weight loss, was
then converted into volume loss using measured density
data. Further specific wear rate (Ks) was calculated from
equation [28]:
3
mNm
sV
KLD



(1)
where;
 V is the volume loss in m3, L is the load in
Newton and D is the abrading distance in meters.
Figure 1. Dry sand/rubber wheel abrasive test rig.
Table 2. Control factors and levels.
Factor Description
Level 1 Level 2 Level 3
A Filler (wt%)
0 5 10
B Load (N)
11 23 35
C Sliding distance (m) 300 600 900
In unlubricated sliding wear the value of wear coeffi-
cient (k) is the dimensionless quotient [29] of the specimens
was calculated according to the equation:
VH
kLD

 (2)
where ∆V is volumetric wear, H is hardness of the wear-
ing material, L is the normal load in Newton, and D is
abrading distance in metres.
2.4. Design of Experiments
Wear properties are highly dependent on its material
composition and variables under which it has to be per-
formed [30]. This study was selected as the target quali-
ties for optimization of the weight loss of composite ma-
terial using 212 grain size silica sand of angular shape as
abrasive. Three body abrasive wear test were conducted
to determine weight loss. Taguchi uses a special design
of orthogonal arrays to study the entire process parameter
space with only a small number of experiments [31-33].
Twenty seven experimental runs, based on Taguchi or-
thogonal arrays were used to find the best factor level
condition given in Table 3. The factor levels were as-
sessed according to selected wear loss. The influence of
the controllable process factors on the weight loss was
studied, based on the correlation between the process fac-
tors. By analyzing the grey relational grade matrix, the
most influential factors were identified. Finally the worn
surfaces were observed by scanning electron microscopy
(SEM) to understand wear mechanisms.
2.5. Grey System
Deng [34] proposed Grey relational analysis (GRA)
which is part of grey system. GRA is suitable for solving
problems with complicated inter relationships between
multiple factors and variables. According to Moran et al.
[35] grey relational analysis solves multi-attribute deci-
sion making problems by combing the entire range of
performance attribute values being considered for every
alternative into one single value. This reduces the origi-
nal problem into a single decision making problem. This
method quantifies the influences of various factors and
their relation which is called the whitening of factor rela-
tion. The information that is either incomplete or unde-
termined is called grey. The system having incomplete
information is called grey system. Grey based Taguchi
method follows the optimization method developed by
Dr. Genichi Taguchi. Taguchi method works for optimiza-
tion of a single performance characteristic. GRA is used
to combine the entire considered performance character-
istic in optimization problem [36]. Grey relational analy-
sis requires less data and can analyze many factors that
can overcome the disadvantages of statistical methods.

K. M. SUBBAYA ET AL.
879
Table 3. Experimental layout using L27(313) orthogonal Array and performance results.
Expt. No. Load, A (N) Distance, B (m) Filler, C (wt%) Specific wear rate
(Ks)  10–5 (m3/N·m)
Wear coef.
(k)  10–3 Hardness (H)
1 11 300 0 6.6545 6.1887 73
2 11 300 5 6.312 5.9333 74
3 11 300 10 5.2757 5.0647 76
4 11 600 0 3.6818 3.424 73
5 11 600 5 3.3984 3.1945 74
6 11 600 10 3.1696 3.0429 76
7 11 900 0 2.6615 2.4753 73
8 11 900 5 2.4282 2.2825 74
9 11 900 10 2.2737 2.1827 76
10 23 300 0 3.6898 3.4315 73
11 23 300 5 3.2492 3.0543 74
12 23 300 10 3.1115 2.9871 76
13 23 600 0 1.9854 1.8465 73
14 23 600 5 1.7782 1.6715 74
15 23 600 10 1.7057 1.6375 76
16 23 900 0 1.4801 1.3765 73
17 23 900 5 1.3226 1.2433 74
18 23 900 10 1.183 1.1357 76
19 35 300 0 2.7257 2.5349 73
20 35 300 5 2.3876 2.2443 74
21 35 300 10 2.2076 2.1193 76
22 35 600 0 2.0547 1.9109 73
23 35 600 5 1.6923 1.5908 74
24 35 600 10 1.5114 1.4509 76
25 35 900 0 1.7698 1.6459 73
26 35 900 5 1.3346 1.2545 74
27 35 900 10 1.1892 1.1416 76
3. Results and Discussion Additions of graphite filler considerably decreased the
wear volume.
3.1. Abrasive Wear Volume
3.2. Data preprocessing The wear volume of unfilled C-E, 5G-C-E and 10G-C-E
composites tested with varying load 11 N, 23 N and 35 N
are shown in Figures 2(a)-(c). The wear volume tends to
increase with increasing abrading distance from 300 m to
900 m as well load from 11 N to 35 N. The higher filler
loaded, 10 wt% graphite filled C-E exhibited lowest wear
volume at all abrading distances and loads. Wear volume
data revealed that wear volume of unfilled C-E at higher
load is higher than that obtained at lower load. Severe
damage to matrix and fiber is the main reason for higher
wear at higher load. Higher contact pressure and elonga-
tion of softened matrix caused more debris and micro
cracks in the matrix. Softened polymer matrix could not
effectively protect the carbon fiber from peeling off
which aggravated the fiber removal. This resulted in the
decline of the wear resistance of unfilled C-E composite.
Data pre-processing is a process of transferring original
sequence to a comparable sequence. For this experimen-
tal research, data is normalized between zero and one.
Depending on characteristics of data sequence various
methodologies of data pre-processing available. In this
study the response to be optimized is specific wear rate,
wear coefficient and Hardness.
To obtain the optimal wear performance lower-the-
better quality characteristic has been used to minimize the
specific wear rate and wear coefficient, and can be ex-
pressed as:
  
 
*max
max min
ii
iii
xk xk
xk
x
kx

k
(3)
where


*
i
x
k and


i
x
kare sequences after data pre-

K. M. SUBBAYA ET AL.
880
processing and comparability sequence respectively, k =
1, 2 for wear rate and wear coefficient. i = 1, 2, 3, ···, 27
for experiment number 1 to 27.
Hardness of composite should follow higher—the bet-
ter criterion which can be expressed as:
  
 
*min
max min
ii
iii
xk xk
xk
x
kxk

 (4)
where

*
i
x
k and

i
x
k are sequences after data pre-pro-
cessing and comparability sequence respectively, k = 3 for
hardness. i = 1, 2, 3, ···, 27 for experiment number 1 to 27.
∆oi(k) is deviation sequence of reference sequence


*
0
x
kand comparability sequence

*
i
x
k. Deviation
sequence ∆oi can be calculated using equation:


**
oio i
kxkxk (5)
Table 4 represents the normalized data and deviation
sequence for 27 experimental runs.
(a) (b) (c)
Figure 2. Effect of abrading distance on wear volume of composites at (a) 11 N; (b) 23 N; (c) 35 N.
Table 4. Normalized data and deviation sequences.
Normalized data Deviation sequences (oi)
Expt
No. Ks × 10−4 (k) × 10−3 H Ks × 10−4 (k) × 10−3 H
1 0.0000 0.0000 0.0000 1.0000 1.0000 1.0000
2 0.0625 0.0505 0.3333 0.9375 0.9495 0.6667
3 0.2519 0.2224 1.0000 0.7481 0.7776 0.0000
4 0.5433 0.5471 0.0000 0.4567 0.4529 1.0000
5 0.5951 0.5925 0.3333 0.4049 0.4075 0.6667
6 0.6369 0.6225 1.0000 0.3631 0.3775 0.0000
7 0.7297 0.7348 0.0000 0.2703 2652 1.0000
8 0.7724 0.773 0.3333 0.2276 0.227 0.6667
9 0.8006 0.7927 1.0000 0.1994 0.2073 0.0000
10 0.5418 0.5456 0.0000 0.4582 0.4544 1.0000
11 0.6223 0.6203 0.3333 0.3777 0.3797 0.6667
12 0.6475 0.6336 1.0000 0.3525 0.3664 0.0000
13 0.8533 0.8593 0.0000 0.1467 0.1407 1.0000
14 0.8912 0.8939 0.3333 0.1088 0.1061 0.6667
15 0.9044 0.9006 1.0000 0.0956 0.0994 0.0000
16 0.9457 0.9523 0.0000 0.0543 0.0477 1.0000
17 0.9744 0.9787 0.3333 0.0256 0.0213 0.6667
18 1.0000 1.0000 1.0000 0.0000 0.0000 0.0000
19 0.7180 0.7230 0.0000 0.2820 0.2770 1.0000
20 0.7798 0.7806 0.3333 0.2202 0.2194 0.6667
21 0.8127 0.8053 1.0000 0.1873 0.1947 0.0000
22 0.8406 0.8465 0.0000 0.1594 0.1535 1.0000
23 0.9069 0.9099 0.3333 0.0931 0.0901 0.6667
24 0.9399 0.9376 1.0000 0.0601 0.0624 0.0000
25 0.8927 0.899 0.0000 0.1073 0.1010 1.0000
26 0.9722 0.9764 0.3333 0.0278 0.0236 0.6667
27 0.9988 0.9988 1.0000 0.0012 0.0012 0.0000

K. M. SUBBAYA ET AL. 881
3.3 Grey Relational Coefficient and Grey
Relational Grade
Grey relational coefficient (GRC) expresses relationship
between ideal and actual normalized experimental results
and is given by:
 
min max
max
ioi
kk




(6)
where ∆oi(k) is deviation sequence of reference sequence

*
0
x
k and comparability sequence is

*
i
x
k,  distin-
guishing or identification coefficient. The purpose of grey
relational coefficient is to expand or compress the range
of grey relational coefficient [37]. If all parameters are
given equal preference,
 is taken as 0.5. Grey relational
coefficient (GRC) Table 5 for each experiment can be
calculated using Equation (6).
After obtaining GRC, the Grey relational grade (GRG)
is computed by averaging GRC corresponding to each
performance characteristic. Overall evaluation of multiple
performances characteristic is based on GRG is given by:

1
1n
ii
k
k
n


 (7)
Table 5. Grey relational coefficient and grey relational
grade.
Grey Relational Coefficient
Expt. No
Ks k H
GRG
1 0.3333 0.3333 0.3333 0.3333
2 0.3478 0.3449 0.4285 0.3737
3 0.4006 0.3913 1.0000 0.5973
4 0.5226 0.5247 0.3333 0.4602
5 0.5525 0.5509 0.4285 0.5106
6 0.5793 0.5698 1.0000 0.7163
7 0.6490 0.6534 0.3333 0.5452
8 0.6871 0.6877 0.4285 0.6011
9 0.7148 0.7069 1.0000 0.8072
10 0.5218 0.5238 0.3333 0.4596
11 0.5696 0.5683 0.4285 0.5221
12 0.5865 0.5771 1.0000 0.7217
13 0.7731 0.7803 0.3333 0.6289
14 0.8212 0.8249 0.4285 0.6915
15 0.8394 0.8341 1.0000 0.8911
16 0.9020 0.9129 0.3333 0.7478
17 0.9512 0.9591 0.4285 0.7796
18 1.0000 1.0000 1.0000 1.0000
19 0.6393 0.6435 0.3333 0.5387
20 0.6942 0.695 0.4285 0.6059
21 0.7274 0.7197 1.0000 0.8157
22 0.7582 0.7651 0.3333 0.6188
23 0.8430 0.8473 0.4285 0.7062
24 0.8926 0.889 1.0000 0.9272
25 0.8233 0.8319 0.3333 0.6628
26 0.9473 0.9549 0.4285 0.7769
27 0.9976 0.9976 1.0000 0.9984
It is evident from the Table 5 that experiment 18 has
the best multiple performance characteristic among the
twenty seven experiments since having the highest grey
relational grade. The process parameters are converted in
to signal to noise ratio. Figure 3, graphically represents-
the effect of three control factor in GRG. The combination
factor that optimizes wear process was identified as
A3B3C3. Parameter combination of 10 wt% graphite fi-
lled carbon-epoxy composite abraded for 900 m under a
load of 35 N results in least wear. Total mean of grey
relational grade for twenty seven experiments are given
in Table 6. The rank in the last column of the table
represents the significance of each parameter.
3.4 Analysis of Variance
The purpose of analysis of variance (ANOVA) is to in-
vestigate parameters which significantly affect the per-
formance characteristic. This is established by separating
the total variability of the grey relational grades, which is
measured by sum of the squared deviations from the total
mean of the grey relational grade, into contributions by
each wear parameters and listed in Table 7.
Figure 3. Effect of wear process parameter levels on grey
relational grade.
Table 6. Response table for grey relational grade.
GRG
Symbol ParameterLevel 1Level 2 Level 3
Main effect
(max − min)Rank
A Load 0.54940.7158 0.7390 0.1895 3
B Distance0.55200.6834 0.7688 0.7688 2
C Filler 0.55500.6186 0.8305 0.2755 1
Table 7. ANOVA results.
SymbolParameter DOF Sum of Square Contribution %
A Load 2 0.192408 24.170
B Distance 2 0.214650 26.970
C Filler 2 0.374583 47.070
A*B Load*Distance 4 0.010234 1.290
A*C Load*Filler 4 0.002567 0.322
B*C Distance*Filler 4 0.000220 0.028
Error 8 0.001202 0.150

K. M. SUBBAYA ET AL.
882
According to ANOVA the percentage of contributions
indicate the relative power of a factor to reduce variation.
The factor with high percent contribution has greatest
influence on the performance. The percentage contribu-
tion of filler content (47.07%) was found to be the major
factor affecting the wear performance. Whereas the abrad-
ing distance (26.97%) found to be second influential
factor followed by Load (24.17%) as shown in Figure 4.
Suresha et al. [38] investigated the effect of SiC filler
addition on two-body abrasive wear and concluded that
the maximum wear resistance was obtained for 7.5 wt%
SiC in G-E composite. The present work data by Grey
relational analysis are in good agreement with the experi-
mental investigations.
3.5. Surface Morphology
The worn surface of unfilled C-E composite Figures 5(a)
and (b), 10% graphite filled C-E composite Figures 6(a)
and (b) are examined foe two loads 11 N and 35 N at 900
m abrading distance. Unfilled C-E sample Figure 5(a)
showed higher degree of worn surface features compared
to 10% graphite filled Figure 5(b). The same observations
hold good for samples subjected to higher load of35N
shown in Figures 6(a) and (b). These are supportive to
wear volume shown in Figure 2. Filled C-E composites
Figure 4. Percentage contribution of factors on grey relatio-
nal grade.
(a)
(b)
Figure 5. SEM micrograph of worn surface of unfilled C-E
subjected to load (a) 11 N; (b) 35 N.
(a)
(b)
Figure 6. SEM micrograph of worn surface of 10 wt%
graphite filled C-E subjected to (a) 11 N; (b) 35 N.
at a load of 11 N, exhibited number of broken fiber with
minimum debris as compared to unfilled C-E composites.
Filled C-E composites subjected to higher load of 35 N,
markings of fibers were noticed where as unfilled C-E
Figure 5(b) show large number of broken fibers with lot
of distortion in the matrix with higher degree of debris
formation. The application of higher load has resulted in
the deformation in the form of disorientation of fibers
perpendicular to the abrading direction.

K. M. SUBBAYA ET AL. 883
4. Conclusions
The grey based Taguchi multi response method was ap-
plied in this study to optmize the three-body abrsive wear
of unfilled and graphite filled carbon-epoxy composites.
The results are summarized as follows:
 Taguchi method, a simple systematic and efficient
methodolgy used for optimizing three control factors
to set optimal levels for meeting the objective with
minimum number of experimental runs, to study the
abrasive wear characteristics of composites.
 Optimal wear parameters have been determined by
Grey relational grade for multiperformance charateri-
stics: specific wear rate, wear coefficient and hardness.
 ANOVA of GRG for multiperformance charateristics
revealed that factors like filler content (C), abrading
distance (B) and applied load (A) are significant in
order of priority to minimize the wear.
 Controlling factors with A3B3C3 combination caused
minmum wear and higer hardness.
 Graphite filler found to possess good filler characteri-
stics and increases the wear resistance of C-E com-
posite.
REFERENCES
[1] C.-P. Fung, “Manufacturing Process Optimization for
Wear Property of Fiber Reinforced Polybutylene Tereph-
thalate Composites with Grey Relational Analysis,” Wear,
Vol. 254, No. 3-4, 2003, pp. 298-306.
doi:10.1016/S0043-1648(03)00013-9
[2] B. Suresha, G. Chandramohan and P. V. Mohanram,
“Role of Fillers on Three Body Abrasive Wear Behavior
of Glass Fabric Reinforced Epoxy Composites,” Polymer
Composites, Vol. 10, No. 8, 2009, pp. 1106-1113.
doi:10.1002/pc.20662
[3] M. Cirino, K. Friedric and P. B. Pipes, “Effect of Fiber
Orientation on the Abrasive Wear Behavior of Polymer
Composite Materials,” Wear, Vol. 121, No. 2, 1998, pp.
127-135. doi:10.1016/0043-1648(88)90038-5
[4] Z. Lu and A. M. Hager, “Recent Advances in Polymer
Composites,” Tribology, Vol. 190, 1996, pp. 139-144.
[5] A. A. El-Sayed, M. G. El-Sherbiny, A. S. Abo-EL-Ezz, G.
A. Aggag, “Friction and Wear Properties of Polymeric
Composite Materials for Bearing Applications,” Wear,
Vol. 184, No. 1, 1995, pp. 45-53.
doi:10.1016/0043-1648(94)06546-2
[6] B. Suresha, Siddaramaiah, Kishor, S. Seetharamu and P.
Sampath Kumaran, “Investigations on the Graphite Filler
on Dry Sliding Wear and Abrasive Wear Behavior of
Carbon Fabric Reinforced Epoxy Composites,” Wear,
Vol. 267, No. 9-10, 2009, pp. 1405-1414.
[7] H. W. Chang, “Correlation of Wear with Oxidation of
Carbon-Carbon Composites,” Wear, Vol. 80, No. 9-10,
2009, pp. 7-14. doi:10.1016/0043-1648(82)90083-7
[8] C. Lhymn, “Tribological Behavior of Unidirectional
Polyphenylene Sulphide Carbon Fiber Reinforced Lami-
nate Composites,” Wear, Vol. 117, No. 2, 1987, pp. 343-
359.
[9] B. S. Tripaty and M. J. Furey, “Tribological Behavior of
Unidirectional Graphite Epoxy and Carbon-PEEK Com-
posites,” Wear, Vol. 162, 1993, pp. 343-359.
[10] J. K. Lancaster, “The Effect of Carbon Fiber Reinforce-
ment on Friction and Wear of Polymer,” Applied Physics,
Vol. 1, 1968, pp. 549-555.
[11] P. B. Mody, T. W. Chan and K. Friedrich, “Effect of
Testing Condition and Microstructures on the Sliding
Wear of Graphite Fiber/Peek Matrix Composites,” Jour-
nal of Materials Science, Vol. 23, No. 12, 1998, pp. 4319-
4330. doi:10.1007/BF00551926
[12] S. Chand, “Review Carbon Fibers for Composites,”
Journal of Materials Science, Vol. 35, No. 6, 2000, pp.
1303-1313. doi:10.1023/A:1004780301489
[13] Y. J. Shi, X. Feng, H. Y. Wang and X. H. Lu, “Tri-
bological Properties of PTFE Composites Filled with Sur-
face Treated Carbon Fiber,” Journal of Materials Science,
Vol. 42, No. 20, 2007, pp. 8465-8469.
doi:10.1007/s10853-007-1767-7
[14] J. Bijwe and R. Rattan, “Carbon Fabric Reinforced Poly-
etherimide Composite: Optimization of Fabric Content
for Best Combination of Strength and Adhesive Wear
Performance,” Wear, Vol. 262, 2007, pp. 749-758.
[15] X. Cheng, Y. Xue and C. Xie, “Tribological Investigation
of PTFE, Composite Filled with Lead and Rare Earth-
Modified Glass Fiber,” Materials Letters, Vol. 57, 2003,
pp. 2553-2557.
[16] B. Suresha, Siddaramaih, Kishoe, S. Seetharamu and P.
Sampathkumaran, “Influence of Graphite Filler on Dry
Siding Wear of Carbon Fabric Reinforced Epoxy Com-
posites,” Wear, Vol. 267, No. 9-10, pp. 1405-1414.
doi:10.1016/j.wear.2009.01.026
[17] B. Suresha, B. N. Ramesh, K. M. Subbaya, B. N. Ravi-
kumar and G. Chandramohan, “Influence of Graphite
Filler on Two Body Abrasive Wear Behavior of Carbon
Fabric Reinforced Epoxy Composite,” Materials and De-
sign, Vol. 31, No. 4, 2010, pp. 1833-1841.
doi:10.1016/j.matdes.2009.11.006
[18] B. Suresha, B. N. Ramesh, K. M. Subbaya and G.
Chandramohan, “Mechanical and Three Body Abrasive
Wear Behavior of Carbon Epoxy Composite with and
Without Graphite Filler,” Journal of Composite Materials,
Vol. 44, No. 21, 2010, pp. 2509-2519.
doi:10.1177/0021998310369589
[19] Q. Z. Zhang, Y. W. Liu, Y. D. Huang, L. Liu and J. W.
Bao, “The Effect of Carbon Fiber Surface Properties and
Electron Beam Curing of Epoxy Resin Composites,”
Composites Science and Technology, Vol. 62, No. 3,
2002, pp. 331-337. doi:10.1016/S0266-3538(01)00222-6
[20] W. Qin, Z. Q. Zhang and X. H. Wu, “Effect of Cold
Plasma Treatment on the Mechanical Properties of RTM
Component,” Journal of Materials Science and Technol-
ogy, Vol. 19, 2003, pp. 281-286.
[21] K. Anthony, “Composite Materials,” Journal of Ceramic
Processing Research, Vol. 2, 2001, pp. 147-154.
[22] M. Cirino, K. Friedrich and R. B. Pipes, “The Abrasive

K. M. SUBBAYA ET AL.
884
Wear Behavior of Continuous Fiber Polymer Compos-
ites,” Journal of Materials Science, Vol. 22, No. 7, 1987,
pp. 235-247. doi:10.1007/BF01082134
[23] X. H. Cheng, Y. J. Xue and C. Y. Xie, “Friction and
Wear of Rare-Earth Modified Glass Fiber Filled PTFE
Composites in Dry Reciprocating Sliding Motion with
Impact Loads,” Wear, Vol. 253, No. 7-8, 2002, pp. 869-
877. doi:10.1016/S0043-1648(02)00217-X
[24] G. Taguchi and S. Konishi, “Taguchi Methods: Orthogo-
nal Arrays and Linear Graphs,” American Supplier Insti-
tute Inc., Dearborn, 1987.
[25] G. Taguchi, “Introduction to Quality Engineering,” Asian
Productivity, Tokyo, 1990.
[26] ASTM D2583, “Standard Test Method for Indentation
Hardness of Rigid Plastic by Means of a Barcol Im-
presser,” 2007.
[27] ASTM G65, “Standard Test Method for Wear Testing
with a Dry sand Abrasive Test Apparatus,” 2010.
[28] A. Ravikiran, “Wear Quantification,” ASME, Vol. 122,
2000, pp. 650-656.
[29] K. H. Zum Gahr, “Wear by Hard Particles,” Tribology
international, Vol. 31, No. 10, 1998, pp. 587-596.
doi:10.1016/S0301-679X(98)00079-6
[30] S. S. Mahapatra and S. Datta, “A Grey Based Taguchi
Method for Wear Assessment of Red Mud Filled Polyes-
ter Composites,” International Journal of Modeling and
Optimization, Vol. 1, No. 1, 2011, pp. 80-88.
[31] C. Hung, C.-L. Chang and L. Chen, “Applying Grey Re-
lational Analysis to the Wender Evaluation Model,” In-
ternational Journal of the Computer and the Internet and
Management, Vol. 11, No. 3, 2003, pp. 45-53.
[32] C.-L. Yang, “Optimizing the Glass Fiber Cutting Process
Using the Taguchi Methods and Grey Relational Analy-
sis,” New Journal of Glass and Ceramics, Vol. 1, 1987,
pp. 13-19.
[33] K. H. Chang and C. S. Wu, “A Grey Time Series Model
on Forecasting the Chinese Near Year Effect in the Tai-
wan Stock Market,” Journal of Grey System, Vol. 8, No.
3, 1996, pp. 235-260.
[34] J. Deng, “Introduction to Grey System,” Journal of Grey
System, Vol. 1, 1989, pp. 1-24.
[35] J. Moran, E. Granada, J. L. Miguez and J. Poterio, “Use
of Grey Relational Analysis to Asses and Optimize Small
Biomass Boilers,” Fuel Processing Technology, Vol. 87,
No. 2, 2006, pp. 123-127.
[36] S. Biswas and A. Satpathy, “Tribo Performance Analysis
of Red Mud Filled Glass-Epoxy Composites Using Ta-
guchi Experimental Design,” Materials and Design, Vol.
30, 2009, pp. 2841-2853.
[37] Y. Kuo, T. Yang and G.-W. Huang, “The Use of Grey
Based Taguchi Method for Optimizing Multiresponse
Optimization Problems,” Engineering Optimization, Vol.
4, No. 6, 2008, pp. 517-528.
doi:10.1080/03052150701857645
[38] B. Suresha, G. Chandramohan, J. P. Prakash, V. Balu-
samy and K. Shankaranarayanasamy, “Role of Fillers on
Friction and Slide Wear Characteristics in Glass-Epoxy
Composite System,” Journal of Minerals Materials Char-
acterization and Engineering, Vol. 5, No. 1, 2006, pp. 87-
110.