Wear 320 (2014) 1–6

Contents lists available at ScienceDirect

Wear
journal homepage: www.elsevier.com/locate/wear

Correlation of rubber based conveyor belt properties and abrasive wear

rates under 2- and 3-body conditions
W. Molnar a,n, M. Varga a, P. Braun b, K. Adam c, E. Badisch a
a
AC²T research GmbH, Viktor-Kaplan-Straße 2C, 2700 Wr. Neustadt, Austria
b
Semperit Technische Produkte GmbH – Sempertrans, Triester Bundesstraße 26, 2632 Wimpassing im Schwarzatale, Austria
c
Voestalpine Stahl GmbH, voestalpine-Straße 3, 4020 Linz, Austria

art ic l e i nf o a b s t r a c t

Article history: Predicting the lifetime of a conveyor belt from lab-scale tests has become increasingly important, as the
Received 13 March 2014 cost for the belt represents up to 70% of the acquisition and maintenance costs of a transport system. In
Received in revised form practice, belt selection relies strongly on the well-established ISO 4649 abrasion test, where fixed
4 August 2014
corundum paper is utilised as the abrasive medium, resulting in 2-body abrasion. In the present article,
Accepted 11 August 2014
this is compared to the ASTM G65 test with rolling, round abrasive particles, leading to 3-body abrasion.
Available online 20 August 2014
To evaluate the lab-scale results, they were compared to a conveyor belt that had been used to transport
Keywords: sintered charge for eight years. The comparability and reproducibility of wear patterns encountered on
Tribology this particular belt was matched with the lab test and then correlated with mechanical properties of the
Conveyor belt
rubber materials.
2- and 3-body abrasion
It was found that the ISO 4649 tests, where abrasive wear is dominant, rarely reflect wear patterns
Fatigue
Rubber and wear mechanisms occurring in real applications. In contrast, the ASTM G65 3-body abrasion test
entails fatigue dominated wear, which is found in real applications. The ISO 4649 test results showed a
strong dependence on tensile strength and Shore A hardness, while tear strength was the most
influential factor for the ASTM G65 test.
& 2014 Elsevier B.V. All rights reserved.

1. Introduction the tribologically most stressed component of the belt and is

Conveyor belt transport systems play an important role in
mining, handling of bulk material, timber industry and many
other fields where large quantities of goods have to be transported
over distances ranging from few metres to several kilometres. The
main reason for their widespread use may lie in their constructive
and economical adaptability [1]. Unfortunately, the costs for
acquisition, maintenance, repair and renewal of a conveyor belt
often exceed even the costs of the rest of the transport system,
such as the steel supports and machines [2] attached to it,
considering their respective lifetimes. With a cost share of 30–
70% [3,4], the conveyor belt is of special economic importance. It is
therefore not surprising that enhancing the lifetime of conveyor
belts, which usually means reducing wear, is a priority.
Fabric conveyor belts usually consist of a wear resistant top
layer (‘top cover’), a fabric carcass providing tensile strength, skim
layers for adhesion between rubber and carcass, and a bottom
layer (‘bottom cover’) to cover the carcass and provide sufficient
friction to the drive pulley [5]. In practice, the top layer is always
n
Corresponding author.
E-mail address: molnar@ac2t.at (W. Molnar).
therefore of special interest. Usually a top layer's polymer basis
consists of Natural Rubber (NR) or Styrene-Butadien Rubber (SBR),
while Ethylene Propylene(Diene) Rubber (EP(D)M) or Acrylonitrile
Butadiene Rubber (NBR) is preferred in case of exposure to heat or
oil [5]. As NR/SBR conveyor belts are amongst the most commonly
used and cost-efficient conveyor belt materials, the present paper
will concentrate on these particular materials.
In principle, abrasive wear can be divided into two categories:
2- and 3-body abrasions [6]. The first body is usually denoted as
the material which exhibits the most wear during the test. The
second body, usually the harder material, causes wear on the first
body by transmission of direct or indirect forces. In case of 3-body
abrasion, a third body is also present, which consists of interfacial
elements either created autogenously or foreign matter introduced
into the system. Examples are wear debris, lubricants, entrained
particles or even reactive chemicals. Another classification scheme
for abrasion is the separation into high- and low-stress abrasions
[7]. High-stress abrasion occurs when the abrasive material is split
during the process, generating new sharp edges. Abrasion caused
by sand paper is still considered high-stress, even though the
abrasive does not split, because it remains sharp during the
process [8]. The ISO 4649 test is a good example for such a high-
stress test.

http://dx.doi.org/10.1016/j.wear.2014.08.007
0043-1648/& 2014 Elsevier B.V. All rights reserved.
2 W. Molnar et al. / Wear 320 (2014) 1–6

In general, abrasive wear of polymers is a major topic in
literature. Here, 2-body abrasion often occurs during tribological
tests, while 3-body abrasion is more often present in practical
applications [9,10,11]. A good example for the dependence of
different intensities of wear is given by the work of Evans and
Lancaster [12], where low-density polyethylene showed the low-
est wear rate out of 18 different polymers against rough mild steel,
but the highest against coarse corundum paper.
Hence, abrasive wear of polymers is not an easy matter to discuss.
Also the work of Budinski [13], who investigated 21 different kinds of
polymers (but no rubber) with an ASTM G65 tester, revealed no
simple or conclusive correlations, although hardness, friction and
scratch resistance were taken into account. In general, the wear
processes in polymers are still widely recognised as incompletely
understood [14].
Another very important part that determines the abrasive
properties of rubber in conveyor belts is aging at elevated
temperatures. Oxygen reacts with the sulphur cross-links between
the rubber molecules and reduces the rubber's elastic properties.
At an uptake of just 1% oxygen, natural rubber loses practically all
of its elasticity and wear resistance is impaired [15].
The aim of this work is to compare the ISO 4649 and ASTM G65
tests with each other, as well as with real applications, to determine
suitable tests for the latter. Furthermore, the present paper corre-
lates wear rates with the mechanical properties of rubber and
enhances the understanding of the predominant wear mechanisms.
In practise, the abrasive properties of rubber conveyor belts are
mostly determined by conducting a 2-body abrasion test according
to ISO 4649 [16], where a sample is drawn over abrasive paper on
a rotating drum (Fig. 1a). The abrasive paper contains an edged
and fixed abrasive, which, unfortunately, rarely simulates the
abrasive stress applied in real conveyor belt applications. Different
tests are therefore needed to tailor a product's abrasive properties
to its actual application. One appropriate alternative is the 3-body
ASTM G65 abrasion test [17], which uses a loose abrasive and a
rubber wheel (Fig. 1b). Such a test rig also offers flexible parameter
variation, which can prove useful for reproducing wear phenom-
ena. Namely parameters such as velocity, temperature, applied
load, wheel material, as well as shape and size of the abrasive, can
be changed to fit relevant applications [18]. Hence, five commer-
cially available conveyor belts were acquired (Table 1) and their
mechanical properties were determined. The ASTM G65 standard
procedure was modified and the tests were run according to the
parameters listed in Table 2. The test is therefore referred to as
ASTM G65M to make it clear that it was modified.
For the ASTM G65M tester, 70  25  6 mm3 pieces were cut
out from the conveyor belt samples using a box cutter. Analo-
gously, cylindrical samples (Ø 16  6 mm2) were punched out of
the conveyor belts with a die cutter for the ISO 4649 tester. The
top layers of both kinds of samples were then tested according to
Table 1
List of SBR conveyor belt covers materials used in this study.

2. Experimental Sample Shore A Tensile Tear Max. abrasion Rubber

number hardness strength strength according to data base
In order to compare wear patterns and wear mechanism to real [dimensionless] [MPa] (ISO 34-1 sheet (acc. ISO
method A) 4649) [mm³]
applications, a NR/SBR-based conveyor belt (1000 EP 800/4 10/3 D)
[MPa]
was investigated regarding wear patterns and reduction of the cover
thickness. This particular belt transported sintered materials in an A 55 18.3 11.8 120 SBR
open hall for eight years, at a output of
350 t/h. The remaining B 58 19.9 17.7 90 NR/
cover thickness was determined by comparing the loss of material in SBR/BR
C 59 22.9 14.0 120 NR/BR
the centre to that in the rim regions, while the wear patterns were
D 60 21.9 20.7 70 NR/BR
examined with a high-resolution stereo microscope. The results were E 56 16.7 10.2 100 SBR
then compared to wear patterns from lab-scale tests.

Fig. 1. Schematics of the ISO 4649 2-body abrasion test (a) and the ASTM G65 3-body abrasion test (b).
 W. Molnar et al. / Wear 320 (2014) 1–6
the routines outlined in the standards. Three samples of a kind
were tested to produce statistical representative values.
The applied parameters (Table 2) for the 3-body abrasion ASTM
G65M were 130 N normal load, a wheel velocity of 2.43 m/s, silica
sand (standard Ottawa-sand, 212–300 mm, round) as abrasive, a
Shore 85A rubber wheel and a total distance of 4309 m. For the
ISO 4649 test, the parameters were 10 N normal load, a drum
velocity of 40 rpm, corresponding to a total gliding velocity of
0.31 m/s, standardised sand paper as abrasive (edged) and a total
distance of 40 m.
The main differences between these two tests are the abrasives
used (edged and fixed vs. round and rolling) and thus abrasion
mechanism (2-body and 3-body abrasions), as well as a higher
sliding velocity and total distance for the ASTM G65M. Also, the
applied load is higher for the ASTM G65M and the loaded area of
the sample increases during the test, leading to a larger wear zone
than for the circular ISO 4649 samples.
After testing, the worn surfaces were investigated by stereo-
microscopy and the mass loss was determined gravimetrically.
Quantitative wear rates were calculated for the ASTM G65M and
ISO 4649 tests by using the density of the top layer rubber to
convert mass loss to volume loss. In order to improve comprehen-
sibility and comparability, the abrasion rates are given as volume
loss per distance covered.
3. Results and discussion
3.1. Damage analysis of a used conveyor belt
In practice, the wear a conveyor belt experiences is nearly
always a mixture of more than one wear mechanism. To investi-
gate these phenomena, a used conveyor belt, with a total thickness
of 18 mm in the original state, was examined (Fig. 2). The rim
regions were riddled with large cracks, while the centre exhibits a
much smoother surface due to the influence of transported goods
and scrapers (Fig. 2a). The cracks can be attributed to oxidative
hardening to 80 Shore A and constant troughing as depicted in
Fig. 2b. Practically no abrasion was observed in the rim region, as
the cover layer retained its original thickness of 11 mm (Fig. 2b).
This also explains the higher hardness in the crack-filled rim
Table 2
Parameters for the ISO 4549 and the ASTM G65M test.
ISO 4649
Load 10 N
Sliding velocity 0.31 m/s
Covered distance 40 m
Testing time
2 min
Abrasive Fixed corundum, edged; mesh 60; Moh's hardness: 9
Sample size Ø 16  6 mm2
Fig. 2. A picture of the conveyor belt used for eight years to transport sintered charge with deep cracks in the rim region is shown in (a). The diagram in (b) shows Shore A
hardness and remaining cover thickness respectively to the position in (a).
3
regions, where the material was not abraded and was therefore
subjected to hardening by oxygen uptake over its entire lifetime.
The cracks themselves were up to 2 mm wide and up to 200 mm
long. At the edges, the cracks were already reaching through the
entire thickness of the belt (Fig. 3a).
The centre region featured a much smoother surface and no
cracks, but clear signs of abrasion can be seen. Only 1–2 mm of the
original 11 mm of the cover remained. Practically all abrasion
losses on this conveyor belt can be attributed to tribological
interactions with the transported goods and the scrapers used to
clean the belt. A closer investigation of the centre (Fig. 3b)
revealed a small number of grooves perpendicular to the running
direction, but also large areas covered by small pits. These
roughened areas are characteristic for fatigue wear, which origi-
nates from continuous rolling abrasive and stretching of the
material, but only affects the material surface and regions very
close to it [19]. This usually initiates and propagates cracks, but
there were no large cracks found in the central region. It can be
assumed that the abrasion rates in the centre exceed the rate of
crack propagation and thus prevent larger cracks. So the main
wear mechanism present in the centre of the belt can be seen as
3-body abrasion, where the rolling transported goods are the
rolling abrasives, and to a lesser extent as 2-body abrasion, where
particles of the transported goods stick to the scrapers and form
the fixed abrasive. The interaction between the bare scrapers and
conveyor belt can of course be seen as 2-body abrasion as well.
3.2. Comparison between 2- and 3-body abrasion behaviour in lab-
scale tests
To evaluate the different test methods, the abrasion rates
measured during the commonly used ISO 4649 tests were com-
pared to those from the ASTM G65M tests, as the latter offers more
adaptability. The results of both types of tests are presented in
Fig. 4. Samples investigated with the ISO 4649 tester experience
abrasion rates between 1.4 and 7.2 mm³/m, while ASTM G65M
tested samples experience much lower abrasion rates between
0.05 and 0.27 mm³/m. At first glance, both testing methods show
similar trends, but a closer look reveals important differences.
While having a very similar abrasion rate as B in the ASTM G65M
ASTM G65M
130 N
2.43 m/s
4309 m
30 min
Standard, round Ottawa sand; 212–300 mm; Moh's hardness: 7
70  25  6 mm3
4 W. Molnar et al. / Wear 320 (2014) 1–6
Fig. 3. Macro images taken from a conveyor belt after eight years of service. (a) Side view of a crack in the rim region, which is nearly reaching through the whole belt.
(b) Top view of the center region, where transported goods and scrapers cause abrasion. Here the running direction is from left to right.
Fig. 4. Comparison between wear rates received with the ASTM G65M and the ISO
4649 test.
test (0.032 mm³/m (B), 0.036 mm³/m (C)), the values from the ISO
4649 test clearly differ (2.6 mm³/m (B), 1.4 mm³/m (C)). Sample E
behaves similarly; while showing identical values as A in the ISO
4649 test (2.9 mm³/m), the values in the ASTM G65M test also
differ (0.042 mm³/m (A), 0.051 mm³/m (E)). Samples A and E
feature the highest wear rates when investigated by the ASTM
G65M test. Both samples are solely based on SBR as their rubber
material. Therefore it can be assumed, that SBR is more severely
affected by wear than NR/BR mixtures. Nonetheless sample A
experiences a lower wear rate in the 3-body abrasion than sample
E, what can be attributed to the difference in processing para-
meters and the slightly higher tear strength of sample A.
The least wear rates were exhibited by sample D, which is a NR/
BR mixture and slightly outperforms sample C in 3-body abrasion
due to its higher tear strength and gives atleast values similar to
those of sample C as the deviation overlaps. It can therefore be
concluded that NR and BR improve wear resistance considerably. It
is a generally accepted fact in the rubber industry that the addition
of BR improves elasticity and therefore decreases wear.
Neither during the ASTM G65M nor during the ISO 4649 test
fraction of the abrasive was encountered, what is not surprising as
the forces applied were too low during both tests and samples as
well as counterparts were too elastic.
The wear patterns from both types of tests are shown in Fig. 5.
The ISO 4649 test causes a classical wear pattern found in
literature (Fig. 5a, [23]) with river-like grooves perpendicular to
the gliding direction. Grooves in gliding direction are, however,
also visible. The inset in Fig. 5a* shows a SEM image of the sample,
including a characteristic rim which forms between the transver-
sal grooves and its damaged surroundings. Samples tested with
the ASTM G65M (Fig. 5b) do not feature such grooves, but instead
have a smoother surface with small pits. Such a wear pattern is
generated by fatigue wear due to rolling of round particles as
described by Myshkin et al. [19]. Signs of fatigue can also be seen
in the SEM image inset (Fig. 5b*), where the small pits are an
indication for broken-out particles.
The wear patterns produced by the ISO 4649 tests were not
encountered on the conveyor belt in Fig. 3b, which served as a
practical application reference. Instead, the surface much more
resembles the wear pattern produced with the ASTM G65M test.
However, the large cracks in the rim regions of the used
conveyor belt (Fig. 3a) could not be reproduced by either the ISO
4649 or the ASTM G65M tests. Therefore, and because the top
layer in the rim regions conserved its original thickness, it can be
assumed that this occurrence is not related to abrasion. These
cracks can then be attributed to the constant troughing the
conveyor belt experiences when in use. In this case not only
fatigue, but also tear propagation resistance, becomes important.
Nonetheless, these cracks did not deteriorate the performance of
the belt and are thus no reason for major concern.
The process parameters of both tests cannot be compared to
each other directly. The higher load (130 N compared to 10 N) and
also higher sliding velocity (2.43 m/s compared to 0.31 m/s) of the
ASTM G65M should produce higher abrasion rates than the ISO
4649 test, which is clearly not the case. Therefore the much higher
abrasion rates in the ISO 4649 test can be attributed to the
different abrasive and abrasion mechanisms encountered in these
tests. The ISO 4649 test employs a 2-body abrasion mechanism
with a very sharp and fixed abrasive. Wear in this case is mainly
caused by abrasion and tongue rupture, as is described by Fukahori
et al. [20] and illustrated in Fig. 6. This mechanism also creates
high-stress zones at the bottom of the tongue, where cracks
propagate into the material. The ASTM G65M, on the other hand,
employs 3-body abrasion by round, rolling particles, hence fatigue
wear [21].
Both tests can further be compared to exemplary sliding tests
described by Uetz et al. [23]. The fixed corundum in the ISO 4649
test can be seen as a larger number of more or less sharp needles
gliding over the samples. In this case, the tensile strength is the
determining factor influencing wear. The ASTM G65 test, on the
other hand, can be compared to a group of free-rolling balls gliding
over the sample, in which case the ratio between the coefficient of
friction and Young's modulus becomes the determining factor
regarding wear. In other words, it is determined by the fatigue
behaviour of the materials. Overall, the ASTM G65 test features
only 1.2–2.6% of the abrasion rate the ISO 4649 test produces,
which is not surprising, as fatigue is generally a much slower
process than abrasion.
3.3. Correlation of 2-body and 3-body abrasion to mechanical
properties
The correlation between the abrasion rates and the mechanical
properties of the five rubber samples is shown in Fig. 7, for both
 W. Molnar et al. / Wear 320 (2014) 1–6 5

Fig. 5. Stereo images of typical worn surfaces of sample E tested with the ISO 4649 (a) and ASTM G65M (b). The insets show representative SEM images of a tongue caused
by (a) the ISO 4649 and the much smoother surface caused by (b) the ASTM G65M with small pits caused by fatigue wear.

Fig. 6. Experimental simulation of tongue rupture, where viscous oil works as a substitute for rubber (Fukahori et al. [22]). From left to right one can observe bending of the
tongue due to adhesion (a), stretching the tongue (b) and finally rupture of the tongue tip (c).

Fig. 7. Correlations between the mechanical properties of tensile strength, tear strength and Shore A hardness and the respective abrasion rates received with the ISO 4649
(a–c) and the ASTM G65M (d–f).

the 2-body ISO 4649 and the 3-body ASTM G65M tests. Linear
regressions were applied to the data, and the fitted lines with their
respective equations and coefficients of determination are shown
in Fig. 7
For the ISO 4649 test, good coefficients of determination were
achieved for the correlation between the abrasion rates and tensile
strength (R² ¼0.86, Fig. 7a) and Shore A hardness (R² ¼0.82,
Fig. 7c). This supports the importance of tensile strength for
gliding abrasion, as stated by Uetz et al. [23]. As Shore A hardness
determines the resistance against a penetrating body, it seems
logical that a higher hardness allows the fixed abrasive to affect a
lower volume of the sample. The slopes of both fitted lines are also
very similar (  0.31 for tensile strength and  0.37 for Shore A
hardness), which means that both mechanical properties are about
equally influential. The correlation with tear strength (Fig. 7b), on
the other hand, achieved a coefficient of determination of only
0.39, rendering it an unreliable variable to predict abrasion rates
in this test.
6 W. Molnar et al. / Wear 320 (2014) 1–6
Analysis of the correlation between the abrasion rates mea-
sured for the ASTM G65M tests and tensile strength, tear strength
and Shore A hardness (Figs. 7d–f, respectively) revealed a good
coefficient of determination for tear strength (
0.96). The good
correlation between the abrasion rate and the tear strength is not
surprising. The ASTM G65M mainly involves a fatigue-related wear
process and once the first cracks form during the process, tear
strength becomes the determining factor for the removal of
particles and formation of pits as seen in Fig. 5b. Tensile strength
(R²
0.59) and Shore A hardness (R²
0.69) reach considerably
lower coefficients of determination and can therefore only be seen
weak predictors of abrasion rates.
Although a proper law to predict abrasion rates of multiple
polymers as a function of their mechanical properties does not
exist at present, atleast the determining factors for each test were
identified. Regression analysis showed a good correlation of
abrasion rates with tensile strenght and Shore A hardness of the
materials for the ISO 4649 test and determined tear strength as the
most important factor determingin abarasion rates in the ASTM
G65M tests. However, it was not possible to find a mechanical
property which correlates with the wear rates of both tests as the
abrasion mechanism of both tests are too dissimilar.
4. Conclusion
Based on a comprehensive damage analysis of used conveyor
belts and experimental wear results from lab-scale testing under
2-body and 3-body-abrasive conditions, the following major con-
clusions can be drawn:
 The used belt showed rim regions riddled with large cracks,
which originated from constant troughing of the hardened belt,
while the central region featured a much smoother surface,
caused by rolling goods and scrapers.
 The ISO 4649 test involves 2-body abrasion, while the ASTM
G65M test features 3-body abrasion. The ASTM G65M method
replicates the damage mechanisms encountered under real
operating conditions much better, as the wear mechanism
encountered there is much more fatigue related than the ISO
4649 test, where mainly abrasion occurs.
 Abrasion rates from experiments conducted with the ASTM
G65M method are much lower than for those with the ISO
4649 method, at values of only 1.2–2.6% of the ISO 4649
abrasion rates. The large difference can be attributed to the
aggressiveness of the sharp-edged abrasive. The ASTM G65M
test, on the other hand, mainly causes fatigue wear due to the
round, rolling abrasive.
 Regression analysis showed that the abrasion rates correlate
with the tensile strength and Shore A hardness for the ISO 4649
tests, while tear strength was the most influential factor for the
ASTM G65M tests.
In conclusion, it can be said that the ISO 4649 procedure,
although well established and commonly used, entails a 2-body
abrasion mechanism, which rarely reflects the situation encoun-
tered in real conveyor belt applications. The ASTM G65M test, on
the other hand, leads to a 3-body abrasion mechanism, which is
mainly based on fatigue. This is much more similar to practical
applications. Nonetheless, the deviation of the results still lacks
refinement and will be subject of further studies.
Acknowledgements
This work was funded by the “Austrian Comet-Program”
(governmental funding program for pre-competitive research,
Project XTribology, no. 824187) via the Austrian Research Promo-
tion Agency (FFG) and was carried out at the “Austrian Center of
Competence for Tribology” (AC2T research GmbH). The authors
thank Frank Schütze for performing the abrasion tests and
voestalpine Stahl GmbH, Stricker GmBH & Co KG, Wanggo Gum-
mitechnik GmbH and Semperit Technische Produkte GmbH for the
active research cooperation.
References
[1] H. Westphal, Optimierung von Fördergurten mit Textil-Einlagen, Hebezeuge
Fördermittel 25 (3) (1985) 81–85.
[2] H.P. Lachmann, Fördegurte, Aufwand und Verfügbarkeit, Braunkohle 33 (6)
(1981) 178–184.
[3] R. Hartlieb, v. Wallthor, Beiträge zu einer betriebssicheren Gestaltung von
Gurtförderanlagen im Steinkohlebergbau unter Tage, Berg. Hüttenmänn.
Monatshefte 120 (8) (1975) 371–380.
[4] A. Mattig, K. Wagner, Die Werkstoffprüfung von Fördergurten in neuer Sicht,
Gummi–Asbest–Kunstst. 15 (1962) 448–461.
[5] F. Röthemeyer, F. Sommer, Kautschuk Technologie, 2nd ed., Carl Hanser Verlag,
München Wien, 2006.
[6] M. Petrica, E. Badisch, T. Peisitt, Abrasive wear mechanisms and their relation
to rock properties, Wear 308 (2013) 86–94.
[7] A.P Harsha, An investigation on low stress abrasive wear characteristics of
high performance engineering thermoplastic polymers, Wear 271 (2011)
942–951.
[8] J.D Gates, Two-body and three-body abrasion: a critical discussion, Wear 214
(1998) 139–146.
[9] R.I Trezona, I. Hutchins, Three-body abrasive wear testing of soft materials,
Wear 233–235 (1999) 209–221.
[10] K.G. Budinski, Adhesive transfer to abrasive particles in abrasion testing, Wear
271 (2011) 1258–1263.
[11] J.H. Tylczak, J.A. Hawk, R.D. Wilson, A comparison of laboratory abrasion and
field wear results, Wear 225–229 (1999) 1059–1069.
[12] D.C Evans, J.K. Lancaster, Treatise on Materials Science and Technology, in:
D. Scott (Ed.), The wear of polymers, 13, Academic Press, New York USA, 1979,
pp. 85–139.
[13] K.G. Budinski, Resistance to particle abrasion of selected plastics, Wear 203–
204 (1997) 302–309.
[14] G.M. Bartenev, V.V. Lacrentev, in: K.C. Ludema, L. H Lee (Eds.), Friction and
Wear of Polymers, Elsevier, Amsterdam, 1981, p. 239.
[15] M. Edgea, N.S. Allen, R. Gonzalez-Sanchez, C.M. Liauwa, S.J. Reada, R.
B. Whitehouse, The influence of cure and carbon black on the high tempera-
ture oxidation of natural rubber I. Correlation of physico-chemical changes,
Polym. Degrad. Stab. 64 (1999) 197–205.
[16] ISO 4649, Rubber, Vulcanized or Thermoplastic – Determination of Abrasion
Resistance using a Rotating Cylindrical Drum Device, 2002.
[17] ASTM G65 – 04(2010): Standard Test Method for Measuring Abrasion Using
the Dry Sand/Rubber Wheel Apparatus, ASTM International, West Consho-
hocken, PA, 2010, 10.1520/G0065-04R10.
[18] M. Varga, H. Rojacz, H. Winkelmann, H. Mayer, E. Badisch, Wear reducing
effects and temperature dependence of tribolayer formation in harsh envir-
onment, Tribol. Int. 65 (2013) 190–199.
[19] N.K. Myshkin, M.I. Petrokovets, A.V. Kovalev, Tribology of polymers: adhesion,
friction, wear, and mass-transfer, Tribol. Int. 38 (2005) 910–921.
[20] Y. Fukahori, H. Yamazaki, Mechanism of rubber abrasion Part 3: how is friction
linked to fracture in rubber abrasion? Wear 188 (1995) 19–26.
[21] M. Linz, H. Winkelmann, K. Hradil, E. Badisch, F. Mücklich, Directional
development of residual stress and surface fatigue during sliding contact,
Eng. Fail. Anal. 35 (2013) 678–685.
[22] Y. Fukahori, P. Gabriel, J.J.C. Busfield, How does rubber truly slide between
Schallamach waves and stick–slip motion? Wear 269 (2010) 854–866.
[23] H. Uetz, J. Wiedemeyer, Tribologie der Polymere, Hanser, München, 1984.