Wear 247 (2001) 208–213

On the SEM features of glass–epoxy composite system

subjected to dry sliding wear
Kishore a,∗ , P. Sampathkumaran b , S. Seetharamu b , A. Murali b , R.K. Kumar b
aDepartment of Metallurgy, Indian Institute of Science, Bangalore-560012, India
b Materials Technology Division, Central Power Research Institute, Bangalore-560094, India
Received 2 October 1999; accepted 29 September 2000

Abstract
The scanning electron microscope (SEM) features of glass fiber–epoxy composites subjected to sliding wear for distances ranging from
500 m to 6 km are recorded in this work. The results have shown that there is an existence of noticeable features on the worn surfaces.
Thus for the longer run case interface separation is noticed, while for shorter runs matrix debris formation and occasional glass fiber
fragmentation are seen. The work also highlights the effect of load and sliding velocity on the wear loss pattern. © 2001 Published by
Elsevier Science B.V.
Keywords: Glass–epoxy; Pin-on-disc; Dry sliding wear; SEM observations

1. Introduction sliding wear of glass–epoxy (G–E) has been looked into.

The emphasis is to study the sliding wear behavior of the
The study of wear of polymers in general and polymer cross-section of glass fiber reinforced epoxy resin compos-
based systems in particular is finding increasing citations in ites and record the resulting surface features using a scan-
literature [1] due to the availability of wider choice of mate- ning electron microscope (SEM) [8] for structure-property
rials, ease of manufacturing, good strength and light weight. correlation, since the literature on this is scanty.
An area where their use has been found to be very effec-
tive is the situation involving sliding contact wear [2]. The
polymer based materials are preferred in recent years over 2. Materials and methods
metal based counterparts in view of their low coefficient 2.1. Fabrication of glass fibrous composite laminates and
of friction [3] and ability to sustain loads. This has given test specimens
an impetus to industrial production of the materials, as for
instance in the production of bearing components used in The G–E material used in this investigation was made
automobile industries [4] such as gears, cams, wheels, etc. from the hand lay-up technique using a compression mould-
The introduction of fibers as reinforcing agents in epoxy ing process [5,7]. The procedure consisted of placing the
[5] based polymer materials widen the scope for structural glass fibers with epoxy compatible finish on a substrate ma-
application in view of their prossessing useful properties terial which had a release coat applied on it. Weighed quan-
[6,7] like tensile strength in the range 240–280 MPa, flex- tities of room temperature curing epoxy resin (LY 556, sup-
ural modulus in the range 16–20 GPa, impact toughness plied by Hindustan Ciba-Geigy Ltd., Mumbai) plus hard-
(Izod) in the range 290–370 J/m, interlaminar shear strength ener mix was taken and smeared over the glass fabric. On
in the range 25–50 MPa in addition to light weight (den- this, another layer of the glass fabric was laid and the pro-
sity 2000–2400 kg/m3 ). However, the deployment of these cess continued. The whole lay up was covered with a mat
as components for use in actual service requires good under- finished fabric over which was placed a steel plate with the
standing of the processing related structure and its influence necessary release coat applied on it. The lay up assembly
on wear. Keeping this aspect in mind, the response to dry was pressed in a press. The excess resin was allowed to
squeeze out. The laminate was cured at ambient conditions
∗ Corresponding author. Tel.: 91-80-3344411/ext. 2683; for a period of about 24 h. The thickness of the laminate was
fax: 91-80-3341683. about 3 mm, obtained by using spacers. The cured material
E-mail address: [email protected] (Kishore). was cut using a diamond tipped cutter to yield wear test

0043-1648/01/$ – see front matter © 2001 Published by Elsevier Science B.V.

PII: S 0 0 4 3 - 1 6 4 8 ( 0 0 ) 0 0 5 3 7 - 8
 Kishore et al. / Wear 247 (2001) 208–213 209

coupons of size 6 mm × 6 mm. In the present investigation,

as the laminates were made by hand lay-up technique, the
mechanical properties are expected to be lower compared to
the values mentioned earlier.
2.2. Test set up and wear runs

The test coupon was then inserted and glued in the

thickness direction using an adhesive to pins of size 6 mm
diameter × 25 mm length and having groove in one end of
the pin to rigidly hold the sample (Fig. 1a). This mounting
arrangement was different from the earlier experimental Fig. 2. A graph showing weight loss vs. load for a sample subjected to
work [8,9], wherein the flat portion of the alternate lay-up 3 m/s sliding velocity and 4 km sliding distance.
sequence of glass and epoxy, comes in contact with the
counter surface. In the present case, the cross section of
glass–epoxy sample makes contact with the counter sur- respectively. The weight changes in a sample was measured
face as detailed in Fig. 1b. The counter surface used in the by recording the initial and final weights. This procedure
present work is an alloy steel disc of hardness 62 HRC and of recording of weights was followed for at least four test
having surface roughness of Ra = 10 ␮m. The initial weight samples for a fixed sliding velocity, load and distance run.
of such test coupon/pin assembly was recorded accurately The scatter band has been provided for each of the velocity
using a digital electronic balance after cleaning. After fix- and load employed in this work. The repeatability of test
ing both the disc and the sample pin in their respective wear runs were established by determining the coefficient of
positions, the normal load to the pin was applied through variation [10] which was well with in the acceptable limit of
a pivoted loading lever with a string and pan assembly [9]. 10%. For SEM studies, the wear test coupons subjected to
The required loads were applied by placing known dead select combinations of load and sliding speeds were sputter
weights on the pan. The test was conducted by selecting coated with a thin conducting layer of gold.
the test duration, load and velocity. At the end of the test,
the sample was again weighed in the same digital balance
after cleaning using an air blower.
3. Results and discussion
2.3. Test parameters
The weight loss values recorded are presented in Figs. 2
The values of load, velocity and sliding distance employed and 3. Fig. 2 shows the effect of load in the range 20–60 N
in this work, varied from 20–60 N, 2–4 m/s and 0.5–6 km, on the weight loss of test coupon for a total sliding distance
of 4 km at 3 m/s sliding velocity. The weight changes with
respect to sliding velocities in the range 2–4 m/s at 40 N are
depicted in Fig. 3. In both the cases (Figs. 2 and 3) the trend
for the weight loss is to show an increase with load and ve-
locity, respectively. These findings are further supplemented

Fig. 1. (a) Schematic showing the mounting of the test piece in the pin
holder assembly and (b) inset of the G–E test sample showing glass fabric Fig. 3. A plot describing weight loss vs. sliding velocity for a sample at
arrangement. 40 N load.
210 Kishore et al. / Wear 247 (2001) 208–213

Fig. 4. SEM photograph showing the features of worn surface at 500 m sliding distance, 40 N and 3 m/s.

by the surface features as seen in SEM photomicrographs for the exposed longitudinal fibers to fragment thus bringing
taken at the central region of the worn surfaces. out the effect of distance on the wear surface features. The
Fig. 4 records the SEM features of the worn surfaces for matrix (Fig. 5) retains less of the wide spread, continuous
the 0.5 km at constant values of load (40 N) and velocity mass-like features seen in Fig. 4. Further the matrix debris
(3 m/s). The spread of the matrix and faint cracks, with some appear (Fig. 5) in increasing amount and the discontinuous
debris (marked ‘P’) are noticed. However, a detailed exam- glass fibers that are fragmented are distributed at random
ination of the 1 km run sample showed (Fig. 5) a tendency orientations (for example, at regions marked ‘F’).

Fig. 5. Surface features showing the SEM observation for 1 km run at 40 N and 3 m/s.
 Kishore et al. / Wear 247 (2001) 208–213 211

Fig. 6. SEM surface features for 2 km run at 40 N, 3 m/s showing debris on the sample surface.

Increasing amounts of debris are noticed in longer run longer sliding distance of 6 km (40 N, 3 m/s) is depicted
sample (2 km, Fig. 6). The spread of the debris partly in Fig. 8, wherein in addition to debris formation and bro-
masks the fibers at regions marked ‘K’ on top center of ken fibers, crater formation due to combined factors like
the photograph. Fig. 7 features situation prevailing for still the long wearing-in process coupled with widening of an
longer run case (4 km). Both transverse and longitudinal existing defect (for instance a void) and its emergence at
fibers can be noticed. Also visible are a few broken fibers the surface leading to hollow appearance. Consequently,
(marked by arrows). The test sample subjected to still in this photomicrograph, the subsurface (underneath) layer

Fig. 7. SEM surface features depicting fragmentation of fibers with debris on the surface of sample at 4 km sliding distance, 40 N and 3 m/s.
212 Kishore et al. / Wear 247 (2001) 208–213

Fig. 8. SEM picture for 6 km run sample at 40 N, 3 m/s showing a crater-like feature.

clearly exposed due to the parting off of the top layer is On the other hand, increasing the load to 60 N (Fig. 10) has
noticed. The matrix debris seen in the photomicrograph the effect of showing the features resembling interface sepa-
could be an additional factor in furthering the wear process ration (across the top region of the photograph). Besides the
[11]. matrix has a coloration change probably effected by the fric-
The effect of changing the load to either 20 or 60 N at tional heat developed. Summarizing, the photomicrographs
sliding distance (4 km) and velocity (3 m/s) is brought out in (Figs. 7, 9 and 10) presented earlier corroborate the general
Figs. 9 and 10, respectively. Compared to Fig. 7, the surface trend seen in weight changes in that the matrix debris for-
features depicts (Fig. 9) less of fiber exposure thus highlight- mation and fragmentation of fibers are larger when the value
ing the effect of decreasing the applied load from 40 to 20 N. of the applied load is higher.

Fig. 9. SEM picture depicting worn surface features for 4 km run sample at a load of 20 N and velocity of 3 m/s.
 Kishore et al. / Wear 247 (2001) 208–213 213

Fig. 10. SEM photograph showing features at 60 N, 3 m/s, for 4 km distance run sample.

4. Conclusions References

The worn surface features reflect characteristically the [1] Scott D. Henry, (Ed.) in “Materials for friction and wear
applications”, ASM Handbook, ASM International, Materials Park,
effect of total distance run during slide wear experiments.
Pa., U.S.A. 18 (1992) 820.
Thus small distances run sample show a varying trend [2] R. Ramesh, Kishore, R.M.V.G.K. Rao, Dry sliding wear studies in
for the matrix from continuous mass-like feature to small glass fiber reinforced epoxy composites, Wear 89 (1983) 131.
amount of debris formation. The longest run sample, on [3] S.K. Biswas, Kalyani Vijayan, Friction and wear of PTFE — a
the other hand, display interface separation features besides review, Wear 158 (1992) 193.
[4] Klaus Friedrich, Wear of reinforced polymers by different abrasive
debris getting detached. This situation leads to an increased
counterparts, K. Friedrich (Ed.), in: Friction and Wear of
wear and a general coloration change on the surface. The Polymer Composites, Composite Materials Series, Vol. 1, Elsevier,
samples subjected to wear for the sliding distances in the Amsterdam, 1986, p. 233.
range 0.5 m–6 km show features like increase in matrix [5] H. Lee, K. Nevilee, Handbook of Epoxy Resins, Mc Graw-Hill, New
debris formation, fibers breakage and exposure of both york, 1967, pp. 1–5.
[6] Hans Domininghaus, Plastics for Engineers-Materials, Properties,
longitudinal and transverse fibers due to the wear of the
Applications, Hanser publishers, Munich, 1993, p. 659 (J. Haim, D.
matrix. Hyatt, Trans.).
[7] Kishore, S. Venkatraman, Impact behavior of glass–epoxy composites
containing foam material, J. Appl. Polym. Sci. 67 (1998) 1565.
Acknowledgements [8] Kishore, P. Sampathkumaran, S. Seetharamu, A. Murali, R.K. Kumar,
Dry sliding wear behavior of glass–epoxy Composite, J. Reinf. Plast.
Compos. 18 (1) (1999) 55.
The authors would like to acknowledge the assistance of [9] P. Sampathkumaran, S. Seetharamu, A. Murali, R.K. Kumar, Kishore,
Mr. D. Mollaiah for the help rendered and the Chairman Sliding wear studies in glass–epoxy system through scanning
of the Department of Metallurgy (Prof. D.H. Sastry) for microscopic observations, Bull. Mater. Sci. 21 (4) (1998) 335.
the keen interest shown in the work. The authors thank the [10] Robert A. Storer, (Ed.) in Annual book of ASTM Standards, Sec.
3, 03.02, “Wear testing with a pin-on-disc apparatus” ASTM G 99,
CPRI management for the permission extended to publish
Pa, U.S.A. (1997) p. 392.
this paper. The authors also thank Mr. S. Vynatheya and Mr. [11] B. Vishwanath, A.P. Verma, C.V.S. Kameswara Rao, Effect of matrix
M.R. Mohan for the help rendered in taking SEM pictures content on strength and wear of woven roving glass polimeric
and in conducting experiments, respectively. composites, Compos. Sci. Technol. 44 (1992) 77.