Tribology International 47 (2012) 122–133

Contents lists available at SciVerse ScienceDirect

Tribology International
journal homepage: www.elsevier.com/locate/triboint

Adhesive wear and frictional performance of bamboo fibres reinforced

epoxy composite
Umar Nirmal n, Jamil Hashim, K.O. Low
Faculty of Engineering and Technology, Multimedia University, Jalan Ayer Keroh Lama, 75450, Melaka, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: The aim of the present work is to investigate the adhesive wear and frictional performance of BMBFRE
Received 14 February 2011 composite. Adhesive wear performance of BMBFRE composite was found to be superior for AP-O. The
Received in revised form frictional performance of BMBFRE composite was found to be greater at low sliding velocity for AP-O by
10 October 2011
44% relative to the high sliding velocity. The temperature differential of BMBFRE composite for AP-O to
Accepted 18 October 2011
Available online 25 October 2011
the sliding direction gave lower interface temperatures of about 31.4% and 13.2% as compared to R-O
and P-O. The predominant wear mechanisms for AP-O was back film transfer associated with the
Keywords: formation of fine grooves on the worn surfaces.
Fibres & 2011 Elsevier Ltd. All rights reserved.
Polymer
Sliding
Contact

1. Introduction The use of natural fibres in composite materials are increasingly

being considered in many applications such as sliding panels,
The projection of oil demand has decreased sharply over the linkages, bearing and bushing [13]. However, there are some
recent years due to depletion of oil prices coupled with strict considerations that need to be addressed before natural fibre-
environmental regulations pertaining CO2 emissions to the envir- reinforced polymeric composites gain widespread acceptance and
onment [1]. On the other hand, technological advancement in confidence for them to be commercially viable [14–17]. Among the
petrochemical engineering has produced many new petroleum considerations the wear and frictional performance of a particular
based products for the benefits of mankind. [2]. This had caused a composite is subjected to adhesive wear environment [14,15].
substantial damage to the ecosystem especially when plastic Studies on wear and frictional performance of natural fibres
materials constitute a large portion of domestic waste [3]. Being reinforced polymeric composites have not been carried out
non-biodegradable, huge consumption of polymers in food con- extensively on bamboo fibres. Up to the best of authors’ knowl-
tainers, electrical appliances, sanitary and toys is becoming a edge, only one study concerning bamboo fibres subjected to
major cause of concern [4,5]. adhesive wear mode was reported by Tong et al. [18]. The work
Many works have been carried out to alleviate the pollution revealed that the wear volume of bamboo fibres was a function of
created by non-biodegradable products. Such works include the applied normal load, sliding velocity and orientation of fibres with
use of biodegradable fillers as reinforcement in polymeric com- respect to the sliding direction. Wear volume was superior when
posites, which is now gaining much attention among researchers the fibres were orientated anti-parallel to the sliding counterface.
[6–8]. Natural fibres such as kenaf, sugarcane, betelnut, bamboo, It was claimed that bamboo fibres had high resistance of material
hemp, oil palm and coir fibres have become of great interest as removal process due to the support received by the outer surface
reinforcement alternatives in various kinds of thermoplastic and layer. The outer surfaces protected the inner soft surface from
thermoset matrices. This is because they are cheap and abundant further delamination throughout the test. The predominant wear
due to fast growth of plant, yield good adhesion strength with mechanisms were found to be micro-cracking and adhesion.
resin matrix, possess high aspect ratio and are light in weight, Besides bamboo fibres, many studies have concentrated on
have low abrasiveness to process equipments and more impor- other natural fibres as polymeric composite materials. Yousif et al.
tantly, biodegradable and recyclable; all this leads to less depen- [19] have conducted a similar study on adhesive performance of
dency on petroleum sources [9–12]. wear and frictional behaviour of multi-layered polyester compo-
site based on untreated betelnut fibre mats under wet contact
conditions. The work focused on normal (N-O) and parallel (P-O)
n
Corresponding author. Tel.: þ606 2523007; fax: þ606 2316552. orientations of fibre mats. It revealed that the wear performance
E-mail addresses: [email protected], [email protected] (U. Nirmal). was much superior in N-O compared to P-O while friction being

0301-679X/$ - see front matter & 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.triboint.2011.10.012
 U. Nirmal et al. / Tribology International 47 (2012) 122–133 123

vice versa. For the case of N-O, there was a sign of fluctuation for dry contact condition. Three different orientations of bamboo fibres
friction coefficient at all applied loads. The fluctuation was mainly (i.e. random, parallel and anti-parallel) with respect to sliding
due to the different interfaces in contact with the counterface direction of the counterface have been considered. Tests were
during the sliding process, i.e. fibrous or resinous region was conducted at room temperature (3075 1C) under a constant applied
exposed to the counterface. The predominant wear mechanism at normal load of 30 N, different sliding distances (1–4 km) and sliding
low applied loads (30 N) was the debonding of fibre. There was no velocities (1.7–3.96 m/s) subjected to a smooth stainless steel
sign of fibre pullout or ploughing at the resinous region. At higher counterface (1250 HB).
range of applied loads (70–200 N), micro- and macro-cracking
were evidenced at the composite worn surfaces’ morphology.
Fracture of polyester was observed at the resinous region due to 2. Materials preparation
the high shear resistance of the composite at P-O.
In another study conducted by Nirmal et al. [15], treated 2.1. Preparation of fibres
betelnut fibres’ reinforced polyester composites were investigated
on the adhesive wear property. Different contact mechanisms (i.e. For the current work, bamboo fibres have been extracted
dry and wet) of the composite against the counterface were manually since it was reported in [24] that the manually extracted
studied using a Block on Disc tribology machine. The work fibres exhibit improved fibre density properties compared to
concluded that, under wet adhesive mode, the wear performance chemically extracted fibres. The node portions and thin layers of
is only controlled by the mechanical loading of the composite exodermis (bark) of the bamboo were removed by hand, leaving
against the counterface. In the case of dry adhesive mode, the only the hollow cylindrical portion of culm, which contains the
wear performance was heavily affected due to thermo-mechan- bamboo fibres, Fig. 1a. Following this, the cylindrical portion of
ical loading. Relevant to this, there was a sign of film transfer onto culm was peeled in the longitudinal direction to make strips of
the counterface and of high interface temperatures during the dry 0.5–1.5 mm thick and 1075 mm width. All the peeled strips were
test, which was not in the case for the wet tests indicating no film then soaked in water (hardness: 120–130 mg/l) for three days at
transfer and constant interface temperatures. Hence, under the room temperature (2875 1C) in order to soften them, Fig. 1b. Then,
wet test, the friction coefficient was reduced by about 95% as all strips were removed from the water and gently beaten by hand
compared to the dry tests. on a hard concrete slab in order to loosen and separate the bamboo
In another relevant work on experimental wear modelling of fibre. The resulting fibre bundle was scrapped with a sharp edged
lifeboat slipway, jute fibre reinforced epoxy composite [20,21] knife and combed following the fibres’ longitudinal direction to
was used in place of nickel/chromium coated steel lining. A pin on obtain the desired fibre strips for composite fabrication. All
plate tribo test machine was used to evaluate the newly devel- extracted bamboo fibres (cf. Fig. 1c) were carefully selected under
oped composite in terms of its wear and frictional performance. It a 10  NK Vision microscope for diameters between 30 and 60 mm.
was found that the friction coefficient of the developed composite This was done mainly to ensure the homogenous dispersion of
for unlubricated contact conditions was reduced by about 59% as fibres during the fabrication process. Prior to this, fibres with length
compared to the steel slipway lining. Thomas et al. [20] concluded of 10 mm (cf. Fig. 1d) and 150 mm (cf. Fig. 1e) have been prepared
that the existing steel slipway lining can be effectively replaced using a sharp stainless steel (model: Happy chef, HRJ-B) scissors for
by jute fibre reinforced epoxy composite since it reduces the different fibre orientations in the epoxy matrix, i.e. random (R),
friction generated along the slipway’s lining. parallel (P) and anti-parallel (AP) orientations. For the R-O, the fibre
Chin and Yousif [22] investigated the potential of kenaf fibres as length in the matrix is about 1070.5 mm. All extracted fibres had
reinforcement for tribological applications and found that the been post-cured in an oven for 5 h at 45 1C and at a humidity level
presence of kenaf fibres in the resin enhanced the wear perfor- of 80710% to ensure proper drying of fibres and to enhance surface
mance of neat epoxy by about 85% when the fibres were orientated wettability of the fibres with the epoxy matrix [15]. The density of
normal to the sliding direction. Moreover, the composite roughness the bamboo fibres was determined to be about 910 kg/m3.
after the test in normal orientation was highest compared to anti-
parallel and parallel orientations due to the fact that the composite 2.2. Preparation of composite
experienced high shear resistance during sliding when subjected to
higher applied loads and longer sliding distances. The resin used for the current work was liquid Dow Epoxy Resin
Dry sliding wear of coir fibre reinforced polyester composite has (DER) 331 with density 1120 kg/m3. The curing agent used was
been investigated by Yousif [23]. It was discovered that friction and JOINTMINE 905-3S. The epoxy dynamic viscosity was 13.5 cp. Both
specific wear rate were high for four layers compared to three layers epoxy resin and curing agent were supplied by Pan Asel Chemicals
of fibre mats in the polyester. At higher applied loads and longer (M) Sdn. Bhd. An aluminium mould (4 mm thickness) with dimen-
sliding distances, reinforcing the composite with three layers of fibre sions of 100  100  100 mm3 was used for the purpose of fabricating
mats exhibited low material removal process due to back film the composite. The inner surface of the aluminium mould was
transfer. Hence, the formation of back film transfer had assisted to sprayed with a thin layer of SPMRA (semi-permanent mould release
protect the composite sliding surfaces from severe wear, thus agent). Prior to this, the resin mixed with hardener (50 wt%) and fibre
increasing the wear resistance. Its predominant wear mechanism of length 10 mm (4 wt%) were uniformly mixed with an electric
was micro-ploughing at the resinous regions and debonding of fibres. stirrer (model: Braun AG, type 4–172) and poured slowly into the
From the above survey, it is clear that there is a need to mould. When the mould was completely filled with the mixture, a
achieve a better understanding on bamboo fibre composites for thin steel plate was placed on top of the mould’s opening. A pressure
tribological applications since there is less work reported. In of about 5 kPa was applied on the steel plate to ensure that the
addition, only a limited number of works governing the adhesive trapped air bubbles in the composite were completely forced out.
wear and frictional behaviour of natural fibres reinforced poly- With the pressure still being applied on the mould, the composite
meric composites had been reported in the literature. This serves block was left to cure for 24 h at room temperature (2875 1C). For
as a motivation for the current work. thoroughness in curing, the hardened composite was removed from
The goal of this study is to analyse the adhesive wear and the mould and post-cured in an oven at 60 1C for 1 h. The same
frictional performance of bamboo fibres reinforced epoxy compo- procedures were repeated to fabricate the composite block for parallel
sites using an indigenous Pin on Disc tribo testing machine under and anti-parallel orientations. However, hand lay-up technique was
124 U. Nirmal et al. / Tribology International 47 (2012) 122–133

Bamboo fibre culm Bamboo strips soaked in water

Extracted bamboo fibres Bamboo fibres with length of 10mm

Stacked bamboo fibres (in mat form) with length of 150mm

Fig. 1. Steps of bamboo fibre extraction.

used for these orientations (parallel and anti-parallel), in which the jigsaw (Model: CD301-B1). A schematic view of the tribological test
first layer of the composite material was obtained by pouring the specimen showing its different orientation of bamboo fibre with
liquid epoxy (mixed with 50 wt% hardener) into the mould followed respect to the counterface sliding direction is displayed in Fig. 2.
by placement of bamboo fibres in mat form onto it (cf. Fig. 1e). A steel The density of the composite was determined to be 1169.2 kg/m3.
roller was used to even out the epoxy resin on the stacked bamboo
fibre mat while releasing air bubbles from the mixture and ensuring
homogenous dispersion of fibres throughout the fabrication process. 3. Experimental procedure
This procedure was repeated until a maximum thickness of 10 mm
was achieved (resulting with 10 layers of fibre mats with 4 wt% and Wear and frictional test of the BMBFRE (bamboo fibre rein-
11 layers of epoxy resin). Correspondingly, the average void content forced epoxy) composite was conducted on a Pin on Disc (POD)
in the fabricated composite was determined to be about 5.172%. It tribo test machine, Fig. 3. A load cell (model: Accutec B6N-50)
was computed based on ASTM D2734 standard. was adopted in proper position to measure frictional forces while
For the purpose of conducting the tribological tests, test speci- a digital friction indicator (model: Dibal VD-310) was integrated
mens with dimensions of 10 mm  10 mm  20 mm had been directly to the load cell to capture the frictional readings. An
prepared from the cured composite block using a Black and Decker infrared thermometer (model: Extech 42580) was fixed and
 U. Nirmal et al. / Tribology International 47 (2012) 122–133 125

FN
Weights
10mm Test specimen
10mm
Counterface

Randomly Pivot
Counter
distributed 20mm weight
bamboo fibre balancer Load
cell

Epoxy resin
Specimen holder Shaft
SD

R-O Fig. 3. Schematic illustration of the Pin on Disc (POD) tribo test machine.

FN
where Ws ¼Specific wear rate [mm3/N m]; DV¼Volume differ-
10mm ence [mm3]; FN ¼Normal applied load; D ¼Sliding distance [m].
10mm
The friction coefficient was determined using Eq. (2). Each test
was repeated for three times and the average of the measurements
was determined. In general, the standard deviation of the results for
Parallel orientation all the materials at all sliding distances is listed in Table 1.
of stacked bamboo
fibre Measured f rictional f orce
20mm m¼ ð2Þ
Normal Applied load
Epoxy resin
3.1. Examination of worn surfaces
SD
An NK Metallurgical microscope (model: MT 7100) was used
P-O to analyse the worn surfaces’ morphology. All observing condi-
FN tions were performed at room temperature of 2875 1C and at
10mm humidity level of 80710%.
10mm

4. Results and discussions

Anti-parallel
orientation of stacked 4.1. Wear performance of BMBFRE composite
bamboo fibre 20mm
The wear performance of R-O, P-O, AP-O and neat epoxy (NE)
as a function of specific wear rate (Ws) vs. sliding distance at
Epoxy resin different counterface sliding velocities subjected to an applied
normal load of 30 N is presented in Fig. 4.
SD From the figure, it can be concluded that Ws is quite sensitive to
the different sliding velocities for the three different orientations.
AP-O Generally, it can be seen that reinforcing neat epoxy with bamboo
Fig. 2. Schematic illustration for R, P and AP orientations test specimen subjected
fibres for the three different fibre orientations served to improve
to applied normal load (FN) and sliding direction (SD). the wear performance of the composite as compared to NE.
In the case of R-O, Ws was high after the test compared to P
and AP orientations for the different sliding velocities. This could
be due to the contact condition itself where the fibres received
pointed at the interface of the test specimen and the counterface less support from the epoxy matrix, which tended to lower the
to measure the interface temperature during the test. Before each wear performance. As a result, Ws was noticed to increase at
test, the stainless steel counterface (ASTM B611, 1250 HB) was about 3 km of sliding distance for R-O (cf. Fig. 4d) when it was
polished using a Silica carbide abrasive paper, grade No.1000, subjected to a much higher sliding velocity (3.96 m/s). It was
while the specimens were polished against abrasive paper of observed during the test that there was a high material removal
grade No.800 SiC to ensure proper intimate contact between the process from either the resinous or fibrous regions. This indicates
specimen and the counterface. Adhesive dry sliding tests were that higher sliding velocity contributed to high thermo-mechan-
conducted at room temperature (2875 1C) under applied normal ical loading between the interacting surfaces of the test specimen
load of 30 N, different sliding distances (1–4 km) and sliding and counterface, which caused the interfaces’ temperatures to
velocities (1.7–3.96 m/s). The weight loss for each specimen was build up. The high temperatures degraded the wear performance
determined before and after testing, using a 70.1 mg weight of the composite as it softened the epoxy matrix leading to poor
indicator (model: Shimadzu AW120). Specific wear rate (WS) was interfacial adhesion strength of the fibre/matrix and subsequent
calculated using Eq. (1). high material removal process.
From another viewpoint concerning the P-O, it can be said that
DV the wear performance was much more superior compared to R-O for
WS ¼ ð1Þ
FN D the different sliding velocities. However, from a visual examination
126 U. Nirmal et al. / Tribology International 47 (2012) 122–133

Table 1
Typical values of standard deviation for Ws and friction coefficient.

Fibre orientation R P AP

Speed, m/s 1.70 2.22 2.83 3.96 1.70 2.22 2.83 3.96 1.70 2.22 2.83 3.96

Ws, % 0.3071 0.3112 0.3123 0.4533 0.3535 0.3492 0.4563 0.3867 0.3055 0.3002 0.3201 0.3402
Friction, % 0.1188 0.1219 0.2132 0.2213 0.1271 0.2011 0.2333 0.2333 0.1210 0.1162 0.2193 0.2391

14 14

12 12

Ws, mm3/Nm [x 10-5]

10
Ws, mm3/Nm [x 10-5]

10

8 8

6 6

4 4

2 2
R-O AP-O P-O NE
P-O AP-O R-O NE
0 0
0 1 2 3 4 0 1 2 3 4
Sliding distance, km Sliding distance, km

16 18

14 16

12 14
Ws, mm3/Nm [x 10-5]

Ws, mm3/Nm [x 10-5]

12
10
10
8
8
6
6
4
4
2 2
P-O R-O AP-O NE
P-O R-O AP-O NE
0 0
0 1 2 3 4 0 1 2 3 4
Sliding distance, km Slding distance, km

Fig. 4. Specific wear rate (Ws) vs. sliding distance for the three different fibre orientations (R, P and AP) and neat epoxy (NE) subjected to an applied normal load of 30 N at
different counterface sliding velocities. (a) Sliding velocity: 1.7 m/s; (b) sliding velocity: 2.22 m/s; (c) sliding velocity: 2.83 m/s; (d) sliding velocity: 3.96 m/s.

of Fig. 4a–d for P-O, it is observed that Ws is decreasing proportio- velocities. Interestingly, for AP-O, there was a sign of steady state
nately with the function of sliding distances and sliding velocities (i.e. achievement on Ws at about 3 km of sliding distance for all different
except for P-O subjected to 3.96 m/s of sliding velocity where Ws sliding velocities. This could have been due to the contact condition
increases with sliding distances). For the same reasons explained for itself where in AP-O, the fibres experienced low shear resistance in
R-O, it can be concluded that the different sliding velocities of the the rubbing zone, thus lowering the material removal process. This is
counterface had an influence on Ws. It is significant for P-O that the confirmed from the past published works on natural fibres reinforced
material removal rate was more predominant from the fibrous polymeric composites where it was revealed that high wear resis-
region rather than the resinous region. Due to the fact that the tance of natural fibres/composites can be achieved when they were
sliding force was parallel to the fibre’s orientation, there was an ease orientated AP to the sliding direction [15,25]. However, in a work
of removal from the fibrous regions during the sliding. This further done by Tong et al. [18], it was reported that the wear performance
increased the amount of epoxy resin to be in contact with the was superior when the bamboo fibres were orientated anti-parallel
counterface due to the absence of fibre, which contributed to a to the sliding direction. Arguably, it is to be highlighted here that in
higher thermo-mechanical loading, thereby ruining the wear perfor- [18], the tribo test specimens subjected to the counterface were
mance of the composite. purely based on bamboo fibres. For the current work, bamboo fibres
When the bamboo fibres were orientated anti-parallel to the were reinforced with epoxy resin for the purpose of preparing the
sliding direction, the wear performance was significantly superior test specimens. Therefore, the wear behaviour for the current work
compared to R and P orientations for the different counterface sliding was highly influenced by the contact mechanism formed by the
 U. Nirmal et al. / Tribology International 47 (2012) 122–133 127

fibrous and resinous regions against the counterface. In the case of The figure indicates that there is an improvement in friction
[18], the wear behaviour was only dominant from the fibrous coefficient for the three orientations of BMBFRE composite sub-
regions. Hence, the superiority of wear performance differs for the jected to different sliding velocities as compared to NE. Besides,
current work as compared to the one done in [18]. the trend of friction coefficient for NE is in agreement with the
Fig. 5 shows the effect on wear performance for the three one reported by Edwards [26]. In that work, it was revealed that
different fibre orientations (R, P and AP) and NE subjected to continuous film transfer (i.e. deformation of polymer test speci-
different ranges of sliding velocities (1.7–3.96 m/s). It can be seen men to the counterface) and back film transfer (i.e. deformed
from the figure that the wear performance of the NE and BMBFRE polymer on the counterface back into the polymer test specimen)
composite is poor at higher sliding velocity. On the contrary, the could result in a continuous decrease in friction coefficient with
composite experienced high wear performance at a lower sliding increasing sliding distance due to the deformation of asperities
velocity (i.e. 1.7 m/s) of about 28.8% for R-O, 39.7% for P-O and formed by the polymer test specimen subjected to the sliding
31.3% for AP-O compared to the ones at a higher sliding velocity, counterface.
i.e. 3.96 m/s. Interestingly, significant wear performance of the Further examination of Fig. 6 shows that, except for R-O, it is
BMBFRE composite was achieved for the three orientations when obvious for P and AP orientations to exhibit a gradual rise in
compared to NE. Their improvement is summarised in Table 2. frictional values for increasing sliding distances and counterface
In summary, the wear performance followed the order of sliding velocities. This is because material removal rate is high at
AP-O4P-O4R-O, where AP-O had excellent wear resistance (i.e. higher sliding velocities and thus the tendency for a third body
improvement of Ws was about 60% at 2.22 m/s of counterface sliding interaction between the interfaces is possible. On the contrary,
velocity as compared to NE for the same speed), P-O experienced when the BMBFRE composite test specimen (soft) was in contact
moderate wear resistance while poor wear resistance was incurred with the stainless steel counterface (hard), three different contact
for R-O for all different counterface sliding velocities. Further evi- mechanisms might have taken place. They are known as ‘cold
dence of the composite worn surfaces morphology will be examined welding’ and ‘rupture’ due to the uneven surface of the test
in detail in Section 4.4 with the assistance of the photo micrographs. specimen. Besides that, fine wear particles, which could not be
detected by the naked eyes, could possibly exist between the
interfaces. When the sliding started, rupture of the uneven sur-
face from the test specimen took place, causing a third body
4.2. Friction performance of BMBFRE composite between the interfaces. This third body (from the resinous or
fibrous region) with the trapped wear particles might have been
The frictional performance of NE and BMBFRE composite for R, in circular or linear motion between the interfaces during the
P and AP orientations as a function of sliding distances and sliding adhesive dry sliding resulting in the generation of film transfer
velocities is presented in Fig. 6. onto the counterface. Therefore, possible wear mechanism on the
top surface of the film transfer could be ‘galling’ (due to circular
12 motion of wear debris between the interfaces) or ‘scoring’ (due to
1.7m/s 2.22m/s 2.83m/s 3.96m/s
linear motion of wear debris between the interfaces). This can
cause high shear resistance of relative motion during the dry
10 adhesive test and thus contributing to a much higher value of
Average Ws, mm3/Nm [x10-5]

friction coefficient. However, further evidence on the worn

surfaces of the test specimens will be examined in Section 4.4.
8
Alternatively, a visual examination for R-O revealed a different
trend of friction coefficient with increasing sliding distances and
6 counterface velocities. From Fig. 6a and b, friction coefficient rised
gradually from 2 km of sliding distances till the end of the test
(i.e. 4 km). Here, it can be interpreted that the contact mechanism
4 was predominant from the resinous regions at low/moderate
sliding velocities. Hence, it is possible for a gradual rise in friction
coefficient when contact was made between materials with the
2 same atomic composition such as from the polymer/polymer
pairs and molecular structure [26]. When counterface sliding
velocity was increased (2.83–3.96 m/s), friction coefficient for
0 R-O tended to drop with increasing sliding distances (cf. Fig. 6c
R-O P-O AP-O NE and d). This could be due to the more gradual wear incurred for
Fig. 5. Average specific wear rate for the three different fibre orientations (R, P
R-O, which indirectly contributed to the significant improvement
and AP) and neat epoxy (NE) subjected to an applied normal load of 30 N at of intimating contact between the test specimen and counterface.
different counterface sliding velocities. Besides that, the achievement of low resistance to relative motion
could have been achieved due to the absence of mechanical inter-
locking at higher range of sliding velocities (i.e. 2.83–3.96 m/s)
Table 2
Improvement of Ws in percentage for R, P and AP orientations at different sliding and increasing sliding distance. In lieu of this, the behaviour on
velocities as compared to neat epoxy (NE). the friction coefficient for the higher range of sliding velocities is
in agreement with previous published works [27,28].
Improvement in Ws (%) For comparison purposes, Fig. 7 is presented on average
Sliding velocities (m/s)
friction coefficient as a function of different counterface sliding
R-O P-O AP-O
velocities. From the figure, AP-O exhibited superiority in frictional
1.70 24.3 50.0 52.9 performance followed by P-O and poor frictional performance
2.22 32.2 51.1 60.0 was incurred for R-O as compared to NE. Interestingly, it can be
2.83 28.7 40.4 43.6
seen that there exists a trend of increasing friction coefficients for
3.96 33.6 47.3 56.4
P and AP orientations with increasing sliding velocities. In the
128 U. Nirmal et al. / Tribology International 47 (2012) 122–133

1.2 1.2

1.0 1.0
Friction coefficient

Friction coefficient
0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2
R-O AP-O P-O NE
R-O P-O AP-O NE
0.0 0.0
0 1 2 3 4 0 1 2 3 4
Sliding distance, km Sliding distance, km

1.2 1.2

1.0 1.0
Friction coefficient

Friction coefficient
0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2
AP-O R-O P-O NE
AP-O R-O P-O NE

0.0 0.0
0 1 2 3 4 0 1 2 3 4
Sliding distance, km Sliding distance, km

Fig. 6. Friction coefficient vs. sliding distance for the three different fibre orientations (R, P and AP) and neat epoxy (NE) subjected to an applied normal load of 30 N at
different counterface sliding velocities. (a) Sliding velocity: 1.7 m/s; (b) sliding velocity: 2.22 m/s; (c) sliding velocity: 2.83 m/s; (d) sliding velocity: 3.96 m/s.

1.2 Table 3
1.7m/s 2.22m/s 2.83m/s 3.96m/s
Improvement of friction coefficient (m) in percentage for R, P and AP orientations
at different sliding velocities as compared to neat epoxy (NE).
1.0
Improvement in l (%)
Sliding velocities (m/s)
Average friction coefficient

R-O P-O AP-O

0.8
1.70 3.3 37.8 46.1
2.22 11.2 27.0 46.4
0.6 2.83 10.4 27.1 30.9
3.96 17.2 24.2 27.3

0.4 above findings, it can be said that friction performance of the

BMBFRE composite was superior at low sliding velocity for AP-O
by about 44% and P-O by about 35.7% as compared to the higher
0.2 sliding velocity for the same orientations. An opposite behaviour
was observed for R-O where frictional performance was poor by
about 9.89% at high sliding velocity compared to the low sliding
0.0
velocity. By comparing the friction performance of the BMBFRE
R-O P-O AP-O NE
composite with respect to NE, it can be seen in Fig. 7 that there
Fig. 7. Average friction coefficient for the three different fibre orientations (R, P was a reduction in friction coefficient when epoxy resin was
and AP) and neat epoxy (NE) subjected to an applied normal load of 30 N at reinforced with bamboo fibres. Their improvement (i.e. R-O, P-O
different counterface sliding velocities. and AP-O) is shown in Table 3 for the different counterface sliding
velocities. From the table, it is learnt that highest friction
case of R-O, no obvious trend was observed but at a lower sliding performance was attainable when the composite was subjected
velocity (i.e. 1.7 m/s), the composite experienced high frictional to 2.22 m/s of sliding velocity for AP-O; i.e. the improvement was
values while being vice versa for higher sliding velocity. From the about 46.4% as compared to NE.
 U. Nirmal et al. / Tribology International 47 (2012) 122–133 129

Table 4
100 Improvement of average interface temperature in
percentage for R, P and AP orientations at 2.22 m/s
90 sliding velocity as compared to neat epoxy (NE).

80 Improvement (%)

70 R-O 5.0
P-O 25.5
Temperature, °C

60 AP-O 34.7

50
temperatures to about 31.4% and 13.2% as compared to R and P
40 orientations. The improvement in temperature performance for
the BMBFRE composite as compared to NE is shown in Table 4.
30
From the table, AP-O exhibits high temperature performance; i.e.
20 the average interface temperature was reduced by about 34.7% as
AP-O P-O R-O NE compared to NE. This could be the reason why AP-O exhibited
10 superiority in wear and friction performance as compared to NE,
R-O and P-O.
0
0 1 2 3 4
4.4. Examination of the worn surfaces’ morphology
Sliding distance, km
The photo micrographs for the worn surfaces’ morphology at
100 different sliding distances and sliding velocities subjected to an
applied normal load of 30 N for the three orientations (R, P and
90 AP) are presented in Figs. 9–11, respectively.
Average interface temperature, °C

A close visual examination of Fig. 9a indicates that at short

80
sliding distance (2 km) and low sliding velocity (1.7 m/s), there
70 was an evidence of back film transfer on the test specimen
associated with wear debris scattered on the fibrous region. At
60 3 km of sliding distance (cf. Fig. 9b), there was an evidence of torn
and loose fibres. When the sliding distance escalated to 4 km, the
50
wear was more severe as the rapid back film transfer had left a
40 high amount of wear debris on the worn surfaces, which could
have been as a third body in the interfaces during the test
30 (cf. Fig. 9c). In regard to this, parts of the bamboo fibres (i.e. at
the fibrous regions) were adhered due to the parallel motion of
20 wear debris with respect to the sliding direction, Fig. 9d. This could
have caused rapid material removal and hence lowering the wear
10
performance for R-O. When the test was performed at 2.22 m/s for
0 the same orientation, R-O, Fig. 9e reveals that there was an
R-O P-O AP-O NE evidence of plastic deformation at the resinous region indicating
the high intimate contact that had been achieved during the sliding.
Fig. 8. Temperature performance for the three different fibre orientations (R, P However, due to the absence of fibre, which enhanced the brittle
and AP) and neat epoxy (NE) subjected to an applied normal load of 30 N at
nature of epoxy resin, fracture at the resinous regions was noticed
counterface sliding velocity of 2.22 m/s. (a) Temperature profiles vs. sliding
distances; (b) average interface temperature. on the worn sample. At 4 km of sliding distance, Fig. 9f suggests
that the fibres were polluted with wear debris (i.e. could have been
from the worn epoxy resin), which caused a rapid rise in interface
4.3. Temperature differential in BMBFRE composite temperature. This eased the softening process of the epoxy resin,
which failed to hold the fibres firmly thereby causing rapid
The temperature differential for NE and BMBFRE composite at delamination of fibres associated with debonding of fibres at the
the three different fibre orientations subjected to an applied resinous regions. When counterface velocity was set at 2.83 m/s,
normal load of 30 N and 2.22 m/s of counterface sliding velocity lower sliding distance (cf. Fig. 9g) had caused plastic deformation
is shown in Fig. 8. In Fig. 8a, the error of temperature measure- on the epoxy resin associated with the signs of fracture at the
ment was determined to be about 73 1C for the three orienta- resinous regions. At 3 km of sliding distance (cf. Fig. 9h), part of the
tions and neat epoxy. fibre was broken from the core of the bamboo fibre. This could have
Generally, average interfaces’ temperature builds up with been due to the brittle behaviour of the third body generated
increasing sliding distances, Fig. 8a. This is purely due to the during the sliding, which caused part of the fibre to chip off. At
thermo-mechanical loading, which is the main governing factor 4 km of sliding distance (cf. Fig. 9i), however, the fibre was
for adhesive dry sliding [15,18,19,23,27,28]. However, when the completely broken apart from the main fibre bundle, which con-
side force was anti-parallel to the fibre’s orientation, temperature tributed to a rise in material removal process. At a higher sliding
differential was superior (i.e. low interfaces temperature) com- velocity (i.e. 3.96 m/s), Fig. 9j and k suggests that the wear was
pared to R and P orientations, Fig. 8b, due to the contact much gradual. The predominant wear mechanisms included broken
mechanism of the BMBFRE composite against the counterface. fibre, back film transfer and plastic deformation associated with
This implies that when the bamboo fibres were orientated AP to fractures on the generated back film transfer. Thus, from a macro-
the sliding direction, there was a large reduction of material scopic point of view, it can be concluded that higher sliding velocity
removal from the resinous region. This lowered the interfaces’ had caused the generated film transfer on the worn surfaces to
130 U. Nirmal et al. / Tribology International 47 (2012) 122–133

`
Bf

Pd
F

F
R
Wd
BFt
Fr
SD
SD 0.01mm 10x SD 0.01mm 10x 0.01mm 10x

Pd
F+ Db
Tf
Bf
Dl

De
Lf
BFt

SD 0.01mm 10x SD 0.01mm 1 0x SD 0.01mm 10x

R Pd
Cr
BFt
Ld
Pd

Cr
Db
Sf
Ld
Fr
BFt
Fr

SD 0.01mm 10x SD 0.01mm 10x SD 0.01mm 10x

De

Db

Wd

Bf

Sf

SD 0.01mm 10x SD 0.01mm 10x

Fig. 9. Photo micrographs of BMBFRE composite tested in R-O under 30 N of applied normal load at different sliding distances and different counterface sliding velocities.
(a) 2 km, 1.7 m/s; (b) 3 km, 1.7 m/s; (c) 4 km, 1.7 m/s; (d) 4 km, 1.7 m/s; (e) 2 km, 2.22 m/s; (f) 4 km, 2.22 m/s; (g) 2 km, 2.83 m/s; (h) 3 km, 2.83 m/s; (i) 4 km, 2.83 m/s;
(j) 3 km, 3.96 m/s; (k) 4 km, 3.96 m/s. Remark: SD—Sliding Direction (BFt: back film transfer, F: fibrous region, Wd: wear debris, Pd: plastic deformation, R: resinous region, Fr:
fracture, Lf: loose fibre, Tf: torn fibre, De: debonding, Fþ Db: fibre and debris, Dl: delamination, Cr: crack, Sf: spoil fibre, Ld: large debris, Bf: broken fibre, Db: debris).

detach easily forming large debris, which was removed from the was sign of back film transfer on the worn surfaces of the composite
interfaces throughout the test (i.e. 4 km of sliding distance is when it was subjected to 2 km of sliding distance and 1.7 m/s of
achieved). This caused ease in relative motion and could possibly sliding velocity. Moreover, there was a massive debonding of fibre,
be the reason why frictional performance for R-O was low at higher which could have been from the generated wear debris (i.e. worn
sliding velocity as compared to the low and moderate counterface epoxy debris) travelling in the parallel motion of the sliding force.
sliding velocities (cf. Fig. 7). This is clearly confirmed by Fig. 10b where many debris can be seen
Fig. 10 depicts the different wear modes when the BMBFRE scattered on the worn surfaces at 4 km of sliding distance. When the
composite was tested in P-O at various sliding distances and counter- counterface velocity was increased to 2.22 m/s at 2 km of sliding
face velocities. A close examination of Fig. 10a indicates that there distance, Fig. 10c, a patch of back film transfer (black in colour) is
 U. Nirmal et al. / Tribology International 47 (2012) 122–133 131

BFt
BeF
F

De

SD 0.1mm 5x SD 0.01mm 10x

Hr

We
F
Db

SD 0.01mm 10x SD 0.1mm 5x

R BFt
Pd

Bf

SD 0.01mm 10x SD 0.1mm 5x

Bf We

SD 0.1mm 5x SD 0.01mm 10x

Fig. 10. Photo micrographs of BMBFRE composite tested in P-O under 30 N of applied normal load at different sliding distances and different counterface sliding velocities.
(a) 2 km, 1.7 m/s; (b) 4 km, 1.7 m/s; (c)2 km, 2.22 m/s; (d)4 km, 2.22 m/s; (e) 2 km, 2.83 m/s; (f) 4 km, 2.83 m/s; (g) 2 km, 3.96 m/s; (h) 4 km, 3.96 m/s. Remark:
SD—Sliding Direction (BFt: back film transfer, BeF: bent fibre, F: fibrous region, Pd: plastic deformation, R: resinous region, De: debonding, Bf: broken fibre, Db: debris, Hr:
hollow region, We: wear epoxy).

noticed at the fibrous region. This caused part of the fibres to be At 2.83 m/s of sliding speed, the fibres were bent towards the sliding
ruined due to the micro-ploughing by the hard asperities namely direction (cf. Fig. 10e). This could have caused high shear resistance in
from the worn epoxy into the fibrous region. Thus, the bamboo fibres the rubbing zone, which contributed to the high values of Ws and
were broken into smaller segments, which are evidenced in Fig. 10d. friction coefficient. For the same sliding speed at higher sliding
132 U. Nirmal et al. / Tribology International 47 (2012) 122–133

We

Fg

R BFt

F F

SD 0.1mm 5x SD 0.1mm 5x

R
Cf

Wd
Af F
Fg

SD 0.1mm 5x SD 0.01mm 10x

Fig. 11. Photo micrographs of BMBFRE composite tested in AP-O under 30 N of applied normal load at different sliding distances and different counterface sliding
velocities. (a) 3 km, 1.7 m/s; (b) 4 km, 2.22 m/s; (c) 3 km, 2.83 m/s; (d) 3 km, 3.96 m/s. Remark: SD—Sliding Direction (BFt: back film transfer, Fg: fine grooves, F: fibrous
region, We: wear epoxy, Af: adhered fibre, R: resinous region, Cf: core fibre, Wd: wear debris).

distance (i.e. 4 km), Fig. 10f reveals different modes of wear. It can be poor wear and frictional performance). When the counterface
seen as a large part of hollow region marked ‘Hr’ on the worn sliding velocity was set at 3.96 m/s, Fig. 11d reveals that the fibres
surfaces. This is because, the fibres had a high tendency of fibrous were still in good shape. There were fine grooves, which could have
removal when they were subjected in parallel to the sliding force. been from the micro-ploughing effects by the hard epoxy wear
Hence, when the fibres were detached, the neighbouring resinous asperities generated during the sliding. In summary, the wear
region did not receive any support from the fibre. This could lead to a performance of BMBFRE composite in AP-O was much superior as
large amount of material removal during the test. Remarkably, the compared to R and P orientations. The main reason is due to the
wear was more severe as confirmed by Fig. 10g and h when the fibre orientation itself where it had a crucial role in influencing the
sliding velocity was increased to 3.96 m/s. There was sign of broken wear and friction for the different sliding velocities.
fibre, large amount of plastic deformation at the resinous regions
followed by sign of worn epoxy debris on the worn samples. This can
explain why the Ws and friction coefficient were high for P-O at 5. Conclusion
higher sliding velocity as compared to low sliding velocity for the
same orientation. The wear performance of BMBFRE composite followed the
Interestingly, for AP-O, the wear was ‘so-called’ superior as order of AP-O 4P-O4R-O where AP-O had excellent wear resis-
evidenced by the photo micrographs in Fig. 11 when compared to tance, P-O experienced moderate wear resistance while poor wear
the ones in Figs. 9 and 10. At low sliding speed (cf. Fig. 11a), there resistance was incurred for R-O for all different counterface
was evidence of fine grooves (marked ‘Fg’) on the worn surface. This sliding velocities as compared to NE. The friction performance
could be due to the scoring or galling motion of the generated wear of BMBFRE composite was improved by about 44% at low sliding
debris at the interface. However, the fibre was still in good shape. velocity for AP as compared to the higher sliding velocity. For R-O,
On examining Fig. 11b, several dark patches of back film transfer the frictional performance was poor by about 9.89% at high
had been developed in the fibrous regions. This phenomenon could sliding velocity as compared to the low sliding velocity.
have prevented the large amount of material removal from the The interface temperature differential of BMBFRE composite
fibrous regions since the back film transfer served as a thin layer of for AP-O was low by about 31.4% and 13.2% as compared to R and
shield preventing the fibres to be ruined or detached. Similar P orientations. The main reason was due to the fact that the fibres
finding was reported in [23] where smooth patches of polyester have minimised the amount of asperities interacting between the
film were evidenced on the worn samples, which shielded the interfaces, thereby, lowering the interface temperatures (i.e. low
composite from severe delamination at the fibrous regions. How- effect to thermo-mechanical loading). This is the reason why
ever, fractions of bamboo fibres from the core were adhered in the AP-O exhibited superiority in wear and friction performance as
same direction of the sliding force at higher sliding speed, Fig. 11c. compared to R and P orientations.
This could contribute to a high resistance of motion during the AP-O exhibited superiority in wear (i.e. Ws improved by about
sliding and can indirectly influence the Ws and friction values (i.e. 60%), frictional (i.e. friction coefficient improved by about 46.4%)
 U. Nirmal et al. / Tribology International 47 (2012) 122–133 133

and temperature (i.e. interface temperature improved by about [10] Joshi SV, Drzal LT, Mohanty AK, Arora S. Are natural fiber composites
34.7%) performances at 2.22 m/s of counterface sliding velocity as environmentally superior to glass fiber reinforced composites?, Composites
Part A: Applied Science and Manufacturing 2004;35:371–6.
compared to NE. [11] Paul W, Jan I, Ignaas V. Natural fibers: can they replace glass in fiber
In examination of the worn surfaces’ morphology, the wear reinforced plastics? Composites Science and Technology 2003;63:1259–64.
performance of BMBFRE composite in AP-O was much superior [12] Gowda, TM, Naidu, ACB. and Rajput, C, Some mechanical properties of
compared to R and P orientations. The main reason was likely due untreated jute fabric-reinforced polyester composite, vol. 30, pp. 277–284,
1999.
to the high shear resistance incurred by the bamboo fibre that [13] Chand N, Fahim M, Tribology of Natural Fiber Polymer Composites, Hard-
influenced the wear and friction for the different sliding velocities. cover ed., 2008.
The predominant wear mechanisms for R-O were back film [14] Nirmal SG, Yousif BF. Wear and frictional performance of betelnut fibre-
reinforced polyester composite. Proceedings of the Institution of Mechanical
transfer, plastic deformation, fracture at the resinous region, torn
Engineers, Part J: Journal of Engineering Tribology 2009;223:183–94.
fibre, debonding, delamination of bamboo fibres and micro- [15] Nirmal U, Yousif BF, Rilling D, Brevern PV. Effect of betelnut fibres treatment
cracks. For P-O, back film transfer on the worn surfaces associated and contact conditions on adhesive wear and frictional performance of
with plastic deformation, debonding and broken fibre were polyester composites. Wear 2010;268:1354–70.
[16] Yousif BF, Nirmal U. On the mechanical properties of a treated betelnut fibre-
evidenced through photo micrographs. AP-O on the other hand
reinforced polyester composite. Proceedings of the Institution of Mechanical
exhibited back film transfer with the formation of fine grooves on Engineers, Part C: Journal of Mechanical Engineering Science 2010;224:
the worn surfaces and adhered bamboo fibres parallel to the 1805–14.
sliding force were noticed. [17] Yousif BF, Nirmal U, Wong KJ. Three-body abrasion on wear and frictional
performance of treated betelnut fibre reinforced epoxy (T-BFRE) composite.
The study concluded that AP-O exhibited superior wear and
Materials & Design 2010;31:4514–21.
frictional performance as compared to R and P orientations and it [18] Tong J, Arnell RD, Ren LQ. Dry sliding wear behaviour of bamboo. Wear
is a good alternative for sliding component as it gives good 1998;221:37–46.
contact with lower resistance to wear and tear. [19] Yousif BF, Alvin D, Yusaf TF. Adhesive wear and frictional behaviour of
multilayered polyester composite based on betelnut fiber mats under wet
contact conditions. Surface Review and Letters 2009;16:407–14.
References [20] Thomas B, Hadfield M, Austen S. Experimental wear modelling of lifeboat
slipway launches. Tribology International 2009;42:1706–14.
[21] Thomas B, Hadfield M, Austen S. Wear observations applied to lifeboat
[1] OPEC. World Oil Outlook. 2010 pp. 1–281.
slipway launches. Wear 2009;267:2062–9.
[2] Bolton W. Engineering Materials Technology, vol. 3. UK: Butterworth Heine-
[22] Chin CW, Yousif BF. Potential of kenaf fibers as reinforcement for tribological
mann; 1998.
applications. Wear 2009;267.
[3] Almeida EVR, Frollini E, Castellan A, Coma V. Chitosan, sisal cellulose, and
[23] Yousif BF. Frictional and wear performance of polyester composites based on
biocomposite chitosan/sisal cellulose films prepared from thiourea/NaOH
coir fibres. Proceedings of the Institution of Mechanical Engineers, Part J:
aqueous solution. Carbohydrate Polymers 2010;80:655–64.
[4] Huber GW, Iborra S, Corma A. Synthesis of transportation fuels from biomass: Journal of Engineering Tribology 2009;223:51–9.
chemistry, catalysts and engineering. Chemical Reviews 2006;vol. 106: [24] Murali MRK, Mohana RK. Extraction and tensile properties of natural fibers:
4044–98. Vakka, date and bamboo. Composite Structures 2007;77:288–95.
[5] Shackelford JF. Introduction to Material Science for Engineering. New Jersey: [25] Yousif BF, Lau STW, McWilliam S. Polyester composite based on betelnut
Prentice Hall; 2000. fibre for tribological applications. Tribology International 2010;43:503–11.
[6] JWhite JR, De SK. Short Fiber-polymer Composites. England: Woodhead; 1996. [26] Edwards KL. An overview of the technology of fibre-reinforced plastics for
[7] Billmeyer FW. Textbook of Polymer Science.Third ed. NY, USA: Wiley design purposes. Materials & Design 1998;19:1–10.
Interscience; 1984. [27] Srivastava VK, Pathak JP. Friction and wear properties of bushing bearing of
[8] Baiardo M, Zini E, Scandola M. Flax fibre-polyester composites. Composites graphite filled short glass fibre composites in dry sliding. Wear 1996;197:
Part A: Applied Science and Manufacturing 2004;35:703–10. 145–50.
[9] Corbiere NT, Gfeller LB, Lundquist L, Leterrier Y, Manson JAE, Jolliet O. Life [28] Unal H, Mimaroglu A, Serdar V. Dry sliding performance of thermoplastics
cycle assessment of biofibres replacing glass fibres as reinforcement in against reinforced unsaturated polyester (BMC): In use in electrical contact
plastics. Resources, Conservation and Recycling 2001;33:267–87. breakers components. Wear 2006;261:841–7.