Tribology International 38 (2005) 910–921

www.elsevier.com/locate/triboint

Tribology of polymers: Adhesion, friction, wear, and mass-transfer

N.K. Myshkin*, M.I. Petrokovets, A.V. Kovalev
Tribology Department, Metal–Polymer Research Institute of Belarus National Academy of Sciences, Kirov St. 32A, Gomel, Belarus
Available online 21 September 2005

Abstract

Tribological behavior of polymers is reviewed since the mid-20th century to the present day. Surface energy of different coatings is
determined with new contact adhesion meter. Adhesion and deformation components of friction are discussed. It is shown how load, sliding
velocity, and temperature affect friction. Different modes of wear of polymers and friction transfer are considered.
q 2005 Elsevier Ltd. All rights reserved.

Keywords: Polymers; Adhesion; Friction; Wear; Mass-transfer

1. Introduction molecules of two approaching surfaces. These forces

Friction is a very common phenomenon in daily life and
industry, which is governed by the processes occurring in
the thin surfaces layers of bodies in moving contact. The
simple and fruitful idea used in studies of friction is that
there are two main non-interacting components of friction,
namely, adhesion and deformation. This idea is basic in the
two-term model of friction, although the independence of
these components is a matter of convention. Such approach
is correct for any materials including polymers. Behavior of
polymers has distinguishing features, some of which were
described by Briscoe [1,2]. The present review is connected
with his works. The main concept should be mentioned. It
consists of three basic elements involved in friction:
(1) interfacial bonds, their type and strength; (2) shearing
and rupture of rubbing materials inside and around the
contact region; (3) real contact area [3,4].
2. Adhesion
2.1. Adhesion bonds, their formation and breakdown
When two surfaces are brought into contact, the surface
forces of attraction and repulsion act between the atoms and
* Corresponding author. Tel.: C375 232 77 46 46; fax: C375 232 77
52 11.
E-mail address: nkmyshkin@mail.ru (N.K. Myshkin).
0301-679X/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.triboint.2005.07.016
neutralize each other at some equilibrium separation z0.
When the distance between the surfaces is z!z0 they will be
repulsed from each other, when zOz0 they will be attracted.
Due to these forces the bonds formed between the
contacting surfaces are followed by junctions developed on
the real contact spots. Formation and rupture of the
junctions control the adhesion component of friction.
The simple model of the junction formation has been
proposed by Bowden and Tabor [3].
For the majority of polymers, the Van der Waals and
hydrogen bonds are typical [5,6].
The hydrogen bond develops at very short distance in
polymers containing the groups OH, COOH, NHCO and
others, in which the hydrogen atom is linked with an
electronegative atom. Under favorable conditions two
approaching atoms are linked together by a common proton
providing a strong and stable compound.
The junctions sheared under the applied tangential force
result in the frictional force. That is, work done by the
frictional force results from breakdown of the interfacial
bonds. In general case, the interfacial junctions (their
formation, growth and fracture) are influenced by nature of
the surfaces, surface chemistry and stresses in the surface
layers (loading conditions). The interfacial junctions
together with products of their fracture and the highly
deformed layers where shear deformation is localized, were
named by Kragelskii as a ‘third body’ [4], the concept,
which has been developed later in much broader sense by
Godet [7]. This term implies that the polymer involved in
the friction process may possess the properties, which differ
drastically from its bulk properties.
 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921 911

If the interfacial bonding is stronger than cohesive of the

weaker material, then this material is fractured and the
polymer transfer takes place. Otherwise fracture occurs at
the interface. As a rule, in polymers the surface forces and
forces acting between polymer chains are nearly equal and
fracture often occurs in the bulk of polymers. This is not
always the case. It was observed for metal–polymer contact
that metal is transferred to the polymer surface under certain
conditions [5,6].
Electrostatic attraction makes a contribution to the
adhesion of polymer contact when electric double layer is
formed owing to transition of electron from one surface to
another. The polymer may be acceptor or donor depending on
the origin of the counterbody. In contact with metal, for
example, the metal is the electron donor, and when the contact Fig. 1. The load on sphere as a function of approach: (1) Hertzian; (2) with
consideration for adhesion (JKL).
is broken, the polymer surface gets a negative charge [8].
where
2.2. Johnson–Kendal–Roberts (JKR) model

~ ~ 1 3RPc 2=3 3
dZd=d s ; PZP=P ;d
s s Z Pc;Z pRg:
3R K 2
There are some models designed for description of
adhesion in contact. The Johnson–Kendal–Roberts (JKR) Fig. 1 shows the graph of this dependence.
model [9] (sometimes termed as the model of contact The main concepts and conclusions of the JKR theory
mechanics) and the Derjaguin–Muller–Toporov (DMT) were successfully tested with good agreement of practice
model [10] are best known. The comparative analysis of and theory. Below we will add some data obtained with the
the models [11,12] shows that the JKR theory is applied to contact adhesion meter.
bodies of micrometer and greater sizes having the properties
of polymers, mainly elastomers, whereas the DMT theory is
2.3. Measurement of adhesion
valid for bodies of nanometer sizes having the properties of
metals. Because of this, the JKR theory is briefly described
Measurement of the molecular forces acting between
below.
solids is one of the most difficult experimental tasks. Since
Elastic contact of sphere and half-space is analyzed with
the forces are small and distances at which they act are short
consideration of Van der Waals’ forces which together with
the measuring instruments should meet the specific
the applied load compress the mating bodies. The energy of
requirements. One of the main problems arising when
molecular interactions is taken to be equal to WmZKpa2g.
measuring the molecular forces is that the latter increase
The elasticity of bodies counteracts the action of surface
rapidly with decreasing the distance between the specimens
forces at compression. Using an energy balance the
under testing. Hence, the measurements should be carried
equilibrium of all the forces—load, surface ones, and elastic
out at a very small speed that cannot be done using the
reaction—is found and the equations for main contact
design of the common balance.
parameters are derived. Given this, a combination of
Deryagin et al. [8] proposed to solve the problem by
Hertzian pressure distribution (loading) and Boussinesq
applying the principle of a feedback balance. This design
distribution (unloading) is used. Such combination gives
with modification was used later in a number of experiments
compression in the middle of the contact and an infinite
intended to measure molecular attraction forces. In
tensile stress at the edges.
particular, Israelachvili’s surface force apparatus (SFA)
The relation between the load and the approach is
measures the surface separation by multiple beam inter-
frequently useful. It is written in the following dimension-
ferometry with accuracy G0.1 nm. The surface or inter-
less form:
facial energy can be measured with accuracy of about
10K3 mJ/m2 (see [13]). Nowadays, the molecular forces are
8 2 31=3 measured by an atomic force microscopy using a special
>
>  pffiffiffiffiffiffiffiffiffiffiffi  1 pffiffiffiffiffiffiffiffiffiffiffi  technique [14,15].
>
>
>
>
>
~
3 PC1 K1 4 ~
PC1 C1 5 ; dRK3 ~ K2=3
; When studying the surface forces we have developed the
< 9
~dZ 2 31=3 contact adhesion meter [16] (Fig. 2). When designing the
>
>  pffiffiffiffiffiffiffiffiffiffiffi  1 pffiffiffiffiffiffiffiffiffiffiffi apparatus we have chosen a vertical torsion balance with
>
>
>
>
>K 3 PC1 ~ C1 4 1K PC1 ~ 5 ; K1% d%K3
~ K2=3
; the negative feedback as a basic design scheme. This design
: 9
eliminates the problems with balancing and errors caused by
(1) friction in the balance support.
912 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
Fig. 2. Principal scheme of adhesion meter: (1) frame; (2) string; (3) holder;
(4) movable coil 4; (5) mirror; (6) laser; (7) expander of optical base; (8)
photodetector; (9) coil; (10) specimen; (11) table; (12) stepping motor; (13)
system of fine positioning driven by piezo-drive.
The device consists of a ball probe with diameter
0.5–10 mm, fixed on the arm of highly sensitive electro-
magnetic balance with negative feedback. The test sample is
brought up to the ball with a predefined contact force
(10–10,000 mN). The dependence of adhesion forces vs
distance between the ball probe and the sample is measured
during both approach and pull-off of the sample from the
ball. The interface forces, which tended to rotate the arm
with the ball, are compensated by an electromagnet located
on the opposite arm of the balance. The measurement of a
compensating current in the electromagnet allows us to
determine the acting interface forces. The sample move-
ment proceeds by a piezo-stack and could be varied from 0.1
to 10 nm/s in the range up to 10 mm.
Fig. 3. Stages of contact formation and main points on curve force–
distance: hAC, effective radius of surface forces; hBC, corresponds to the
tensile elastic deformation of the surface; hCD, corresponds to the elastic
mutual penetration with account for adhesive attraction between solids.
The adhesion meter allows us to measure the force
interaction of surfaces in two regimes. When the surfaces
are separated the rupture of bonds (pull-off force) is fixed.
On the other hand, approaching and contact of surfaces
under the action of attractive forces can be also studied. This
process is shown in Fig. 3, which presents the force–
distance curve [17]. Here, point A shows the beginning of
interaction of approaching surfaces, the portion AB
corresponds to pure attraction without formation of the
real mechanical contact between the solids. After the point
B, the force interaction and elastic deformation occur
simultaneously. At the point C, the elastic force of
resistance to penetration becomes to dominate. The point
D corresponds to the moment when the elastic force of
resistance to penetration equals to the adhesive force of
mutual attraction.
Using the contact adhesion meter we measured the
surface energy of coating on silicone plate (crystal structure
(111), homeopolar semi-conductor). Materials of coatings
were octadecyltrichlorsilane (OTS) of 3 nm in thickness,
poly[styrene-b-(ethylene-co-butylene)-b-styrene (SEBS) of
8 nm in thickness, and epoxysilane of 1 nm in thickness.
Also, the organic self-assembled monolayers [(SAMs) of
ODPO4 and DDPO4] of 2 nm in thickness were studied with
titanium and its oxide as underlayers. Silicon ball (radius
1 mm) and titanium one (radius 1.5 mm) probes were used
as counterbodies.
Our measurements have demonstrated that the specific
surface energy of the test coatings depends on material of
coating and probe as well as on probe radius. The energy is
equal to 0.004 J/m2 for OTS and epoxysilane (probe radius
1 mm) and 0.003 J/m2 for epoxysilane and 0.002 J/m2 for
SEBS (probe radius 1.5 mm). The following results were
obtained for SAMs: 0.007 J/m2 for SiCTiOxCODPO4,
0.011 for SiCTiCODPO4, and 0.002 for SiCTiCDDPO4
(indenter 1 mm), as well as 0.004 J/m2 for ODPO4 (indenter
0.75 mm).
Obtained results can be divided into two groups [17].
First group includes the results obtained when using a
silicon probe. The distances corresponding to the surface
force interaction between silicon and a tested sample fall in
the range from 90 (silicon—SiCTiOxCODPO4) up to
115 nm (silicon—OTS).
The second group is based on measurements with
titanium balls as a probe. The distances lay in a range
from 14 (SiCTiOxCODPO4) up to 50 nm (SiCepoxysi-
lane).
This data show that the range of a force field for silicon
ball is larger than that for titanium.
Calculated values g for sample SiCepoxysilane are very
close to values which are obtained at interaction of these
samples with silicon and titanium probes, whereas the
values of g for coating SiCTiOxCODPO4 are much
outside.
The obtained data are in a good agreement with the
model presented earlier [17].
 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921 913

2.4. Deformation component of polymer friction occurs due to high hysteresis losses. This deformation
component is known as friction due to elastic hysteresis
Another source of the frictional force is attributed to [3,21].
deformation taking place when the asperities of two sliding The energy may be also carried away, for example, with
surfaces come into contact with each other. The surface elastic waves generated at the interface and outgoing to
asperities experience elastic, plastic or viscoelastic defor- infinity, owing to nucleation and development of micro-
mation depending on material behavior. At initial appli- cracks in material [1,22].
cation of load to polymer, the deformation will be mainly
plastic if polymer is in glassy state or mainly viscoelastic (or
even viscoplastic) if polymer is in highly elastic state. Here, 3. Friction of plastics
it should be noted that at the present time there is a strong
trend towards transition from macro- to micro- and
3.1. Real contact area
nanoscale that may give a new insight on the basic problems
of tribology.
When two surfaces approach each other, their opposing
Mechanical properties of contact materials should be
asperities with maximum height come into contact. As the
taken into account at any scale level but depending on this
load increases, the new pairs of asperities with lesser height
level such parameters as Young’s modulus and hardness can
make contact forming individual spots. The overall area of
differ not only in magnitude but also in their physical
these spots is known as the real contact area (RCA).
interpretation. This deformation is accompanied by dissipa-
When simulating the real contact with plastics, tempera-
tion of mechanical energy depending on deformation mode,
ture, and sliding velocity, to name but a few, should be taken
sliding conditions, rubbing materials, scale level of
into account [3,23–25]. Here, only the effect of temperature
mechanical properties, environment, and other factors.
is studied taking in mind that effect of velocity in general is
The hardness, plasticity index and elastic modulus for
considered similar.
organic polymers [poly(methylmethacrylate), PMMA;
A contact of a rigid rough body and smooth elastic half-
poly(styrene), PS; poly(carbonate), PC; and ultra-high
space was examined [25,26]. The latter is taken to be
molecular weight poly(ethylene), UHMWPE] were
homogeneous and isotropic with Young’s modulus E,
obtained using the contact compliance method [18]. The
Poisson’s ratio n, and thermal expansion coefficient a. The
dependence of the imposed penetration depth, the maximum
problem for a heated sphere of radius R was solved by
load and the strain rate upon the hardness and elastic
Barber[27] with the following assumptions: temperature of
modulus for these polymers was described; typical
the heated sphere is T, while outside of the contact the
penetration depths were in the range from about 10 nm
surface of the half-space is free of the mechanical and
to 10 mm while the applied loads were less than 300 mN
thermal loads.
[19,20].
A desired solution (i.e. relations between the total load
Let us consider some deformation mechanisms.
P, circular contact radius a and approach of the sphere d) is
In the Bowden–Tabor model [3] for sliding friction, the
sought as a superposition of the solutions derived for the
asperities of the harder surface are assumed to plough the
problems formulated at the mechanical loading and
softer one. The ploughing resistance causes a force
heating.
contributing to the frictional force. This contribution is
Following Greenwood–Williamson’s approach [28], we
referred to as the ploughing component of friction, the
find the total area of all the contact spots (RCA) and the total
deformation term. A simple estimate for conical asperity of
load. The equations in non-dimensional form for normal
semi-angle q gives that the coefficient of friction due to the
distribution of asperity heights are written as
ploughing can be written
ð
N
2 1
fd Z cot q Ar =Aa Z pRDs pffiffiffiffiffiffi ðxKhÞexp½Kx2 =2dx; (2)
p 2p
h
The slope of surface asperities is less than 108, that is, the
semi-angle qO808, and the coefficient fd should be about ð
N
P ð1Kv2 Þ 4 1=2 3=2 1
0.05 and less. In elastic contact, fd is often assumed to be Z R Ds pffiffiffiffiffiffi ðxKhÞ3=2 exp½Kx2 =2dx
negligibly small. Aa E 3 2p
h
It should be noted that almost without exception
ploughing is accompanied by adhesion and under certain 2aT 1
C ð1 C vÞRDs pffiffiffiffiffiffi
conditions it may result in microcutting, that is, additional p 2p
work should be done increasing the friction.
ð
N
There are other mechanisms of energy dissipation at
deformation. So when a polymer with viscoelastic behavior ! ðxKhÞexp½Kx2 =2dx:
slides against a hard rough surface, the energy dissipation h
914 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
Fig. 4. The real contact area Ar/Aa vs temperature T at different non-
~ 10K4 : 1KPZ
dimensional pressure P; ~ 10; 2K8; 3K6; 4K4; 5K
~ 2 ðPZ
PZ ~ Pð1Kv2 Þ=ðEAa Þ.
Here, D is the surface density of asperities, Aa is the
apparent contact area, h is the separation, s is the root-mean
square of asperities heights.
As an example we calculated the behavior of nylon (elastic
modulus EZ1.5 GPa, Poisson’s ratio nZ0.38, hardness HZ
0.1 GPa, thermal expansion coefficient aZ1.0!10K4 KK1)
and modify it by changing the temperature-dependent moduli
of elasticity and Poisson’s ratios.
The effect of the excess temperature on the relative RCA
is shown in Fig. 4. The higher the temperature, the smaller is
the RCA. In fact, we have the situation which takes place in
a sliding contact where the so-called ‘thermoelastic
instability’ appears.
Thus, it was shown that the RCA of two bodies with
different temperatures becomes smaller when the tempera-
ture difference increases, and the RCA is always smaller
than that in the isothermal case. However, if the mechanical
behavior of material is sensitive to change of temperature
(for example, the elastic modulus for majority of polymers
drops with increasing the temperature), the decrease in RCA
obtained above may be ‘hidden’ by the RCA rise due to
decrease in mechanical characteristics.
For simplicity of the further analysis, we shall restrict our
consideration by the polymers with simple thermal–
rheological behavior. Then the simple exponential describes
Fig. 5. Effect of the rheologica parameter b on temperature-dependent real
contact area Ar/Aa: (1) bZ0/0020; (2) 0.0018; (3) 0.0016; (4) bZ0.0014.
its temperature-dependent modulus
E Z E0 expðKbTÞ; (3)
where b is a constant having dimensionality of the
reciprocal of temperature and conventionally termed the
rheological parameter.
The calculation was performed at several loads with
different values of b. Some results are presented in Fig. 5. It
is seen that the temperature-dependent RCA may pass the
minimum with increasing the temperature. Existence of this
minimum and its value depend on a combination of the
thermal (a) and mechanical properties of contacting
materials. An increase in the parameter b, i.e. a sharpening
of dependence of the modulus on temperature, results in
degeneration of the minimum (ascending RCA wE curve),
all other things being the same (Fig. 5).
3.2. The two-level model of contact
There is no question that the surface asperities are not
smooth, i.e. they have smaller asperities of nanoscale size
(order of 10 nm), which result from molecular and super-
molecular structure of polymers. In this case, the RCA
should be estimated based on two-level model of Archard
type. The combination of two levels, roughness and
microroughness, was examined. To take into account
microroughness on the large asperities (roughness), the
solution of Greenwood and Tripp [29] for contact of rough
spheres was used as the governing equation for a single
asperity. Analysis of the two-level model has shown that the
highest asperities of the first level (roughness) are coming
into contact and form the individual contact spots. Yet,
contrary to traditional view the spots are not continuous but
multiply connected, that is, each of the spots consists of a set
of smaller spots the total area of that was conditionally
named ‘physical contact area’. This area results from the
contact of microasperities (micro/nanoscale roughness) and
it is less that the real contact area by an order of magnitude.
The strong interaction between the mating surfaces may
occur within the microspots of the physical contact. The
contribution of this physical and may be chemical
interaction to total resistance to relative displacement of
rubbing surfaces may be very significant.
The two-level model with adhesion is of interest for
precision engineering. It can be developed in the same
manner as for the above model. A rough sphere models an
asperity of the first level. The load–contact radius relation
for a single microasperity is assumed to result from JKR or
DMT theory. Without going into detail of calculation
procedure which is described elsewhere [12,30] we simply
present some results dealing with the physical contact area.
Numerical experiment shows that for polymers the
physical contact area increases several times due to the
surface forces (Fig. 6). This rise is 1.5 times for high-density
polyethylene, 3 times for low-density polyethylene, and 2
 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
Fig. 6. Effect of surface forces on physical contact area: (1) low-density
polyethylene; (2) high-density polyethylene; (3) nylon; solid lines—
calculation with consideration for surface forces; dotted lines—without
consideration for surface forces.
times for nylon. Under the same conditions, a rise in the real
contact area is substantially less. For example, for high-
density polyethylene the RCA increases by 10% when the
root-mean-square roughness s is 0.5 mm and by 30% when
sZ0.08 mm.
3.3. Effect of load on friction
It is a common knowledge that the friction force is
proportional to the normal applied load (the first law of
friction). Experiments of a number of researchers have
shown that this law is valid for some polymers tested under
certain conditions (Table 1). Thus, the friction coefficient
remains practically constant at load in the range 10–100 N
Table 1
The effect of load on friction coefficient
Author(s)
Bowers, Clinton, and Zisman [32]
Shooter and Thomas [33]
Shooter and Tabor [31]
Rees [34]
Bartenev [22], Schallamach [35]
Kragelskii [4]
915
when a steel ball of radius 6.35 mm slides over polytetra-
fluoroethylene (PTFE), polymethylmethacrylate (PMMA),
polyvinylchloride (PVC), polyethylene (PE), and nylon
[31]. Other authors have obtained similar results with
the same materials as well as with some others, e.g. with
PTFE, polytrifluorochlorethylene (PTFCE), PVC, polyvi-
nylidene chloride (PVDC), PE at the load 2–15 N [32], with
PTFE, PMMA, polystyrene (PS) and PE at the load 10–40 N
[33], and so on.
Outside this range, on the left and right, the proportion-
ality between friction force and applied load breaks down.
Thus, it was shown that in the range of moderate loads
0.02–1 N the friction coefficient decreases with increasing
the load [34]. Such a behavior may be explained by elastic
deformation of the surface asperities [4]. Of interest is the
fact that the similar behavior is characteristic of rubbers for
which the elastic deformation is typical [35].
On the other side of the proportionality range, the friction
coefficient increases with increasing the load. This is often
explained by plastic deformation of asperities in contact.
Thus, friction of polymers as a function of load varies in
the manner, which was described by Kragelskii [4]. That is,
the friction coefficient passes a minimum, which corre-
sponds to transition from elastic contact (left descending
branch of the curve) to plastic one (right ascending branch
of the curve).
We should bear in mind that the load can vary the
temperature of viscoelastic transitions in polymers and
thereby the mechanism of friction.
Materials and load Graphical representation
2–15 N, steel–polymer (PTFE, PFCE, PVC, PVDC, PE) f
N
10–40 N, steel–polymer (PTFE, PE, PMMA, PC) f
N
10–100 N, steel–polymer (PTFE, PE, PMMA, PVC, f
nylon)
N
Steel–polymer (PTFE, PE, nylon) f
N
Theory, steel–rubber f
N
Theory, steel–rubber f
N
916 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
Table 2
The effect of sliding velocity on friction coefficient
Author(s)
Shooter and Thomas [33]
Milz and Sargent [41]
Fort [43]
White [39]
Flom and Porile [37,38]
Oloffson and Gralben [36]
Bartenev and Lavrentev [22], challamach [40]
3.4. Effect of sliding velocity on friction
It is agreed that the friction force is independent of the
sliding velocity. This proposal is valid with a good
approximation only in the case where the contact
temperature varies insignificantly and, as a result, the
interface does not change its behavior. But separation of
the effect of velocity and friction temperature presents
significant difficulties. So, the results obtained by different
researchers should be analyzed with caution. Examples of
the great diversity of available results are shown in
Table 2.
Speed-independent friction was revealed only within a
limited range of velocities (0.01–1.0 cm/s) for PTFE, PE,
PMMA, and PS [33] as well as for fiber–fiber contact [36].
But more complex relationships between friction and
sliding velocity are most often observed. Such relationships
can be connected with viscoelastic behavior of polymers.
In the range of low velocities, the viscous resistance in
the contact zone increases with increasing velocity. When
the contact pressure is high, the abnormally viscous flow is
observed which leads to a sharp rise of viscosity due to
velocity increase (see, e.g. [37–39]). Molecular-kinetic
considerations also lead to the same dependence [22,40].
Materials and sliding velocity Graphical representation
0.01–1.0 cm/s, steel–polymer (PTFE, PE, PMMA, PC)
4–183 cm/s, polymer–polymer, 1—nylon, 2—PC f 2
1
v
10K5–10 cm/s, steel–polymer (PETF) f
v
0.1–10 cm/s, steel–polymer (1—PTFE, 2—nylon) f
2
1
v
1.1–180 cm/s, steel–polymer (PTFE) f
v
1.5 cm/s, polymer–polymer (fibers) f
v
Theory, steel–rubber f
In the range of high velocities, elastic behavior is
prevalent in the contact zone and, as a result, the friction
force depends only slightly on the velocity or it decreases
with velocity (see, e.g. [41,42]). In addition, it should be
borne in mind that the duration of contact is short at high
velocity leading to a further decrease in the friction force.
In the intermediate range of velocities, all the above
factors are in competition, and a maximum appears in the
friction force–sliding velocity curve, position of which
depends on the relaxation properties of polymer (see, e.g.
[43]).
It should be recognized that the friction force–sliding
velocity relationship depends essentially on the test
temperature [44]. When the tests are conducted near the
glass-transition temperature (high mobility of polymer
segments), the sliding velocity has a pronounced effect on
friction, whereas at lower temperature (segments of the
main chain are frozen) friction hardly depends on the sliding
velocity.
3.5. Effect of temperature on friction
Polymers as viscoelastic materials are very sensitive to
frictional heating. It is well known that friction is a typical
 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
dissipative process in which mechanical energy is converted
into heat (up to 90–95% according to the available
experimental data). The thermal state of friction contact is
frequently a decisive factor when evaluating the perform-
ance of a friction unit.
It is commonly believed that heat generation at friction
results from the deformation of material in the actual contact
spots. Some processes with their molecular mechanism
relating to the transformation of mechanical energy into
heat can proceed via plastic deformation, hysteresis,
dispersion, and viscous flow.
Another source of heat can be attributed to origination
and breakdown of adhesion bonds. These processes are most
probably non-equivalent energetically, and the energy
difference may cause the generation or absorption of heat.
Often it is believed that the temperature effect on friction
can be taken into account using the mechanical character-
istics of polymers measured at certain temperatures. In
support of this assumption, a correlation of friction
coefficient with hardness and shear strength was found for
some polymers [43–45]. Such correlation is valid only when
temperature produces no effect on adhesion [46].
Some friction patterns as functions of temperature are
presented in Table 3.
The basic mechanism of friction of polymers in the
highly elastic state over smooth surfaces is adhesion.
Another friction mechanism appears when a polymer
transforms from the highly elastic into glassy state.
Mechanical losses contribute more to the bulk re-
deformation of the surface layers on the polymer; the
volume mechanical component contributes more when the
polymer is heated almost to the glass-transition tempera-
ture, Tg, until it becomes comparable with the contribution
of adhesion.
Table 3
The effect of temperature on friction coefficient
Author(s)
Shooter and Thomas [33]
Ludema and Tabor [45]
King and Tabor [46]
Schallamach [35]
917
The above relations were observed for amorphous
polymers and rubbers. The similar dependence was also
obtained for crystalline polymers [46].
4. Wear of polymers
4.1. Wear modes
The changes in surface layer arise from mechanical
stresses, temperature and chemical reactions. Polymers due
to their specific structure and mechanical behavior are more
sensitive to these factors.
The local temperature at the interface may be substan-
tially higher than that of the environment, and may also be
enhanced at the asperity contacts by transient ‘flashes’ or
‘hot-spots’. The temperature exerts an influence on wear of
polymers. Thus, it was shown that a number of polymers
sliding against steel pass a minimum at characteristic
temperature [47].
The above-listed mechanisms are the basis for wear
process. Yet, the great diversity of the mechanisms and their
interrelation make impossible the rigorous classification f
wear processes [4,48,49]. It is generally recognized that the
most common types of wear of polymers are abrasion,
adhesion, and fatigue.
4.2. Abrasive wear
The key aspect of abrasive wear relates to cutting or
plowing of the surface by harder particles or asperities.
These cutting points may either be embedded in the
counterface, or loose within the contact zone. The former
Material and test temperature Graphical representation
20–80 8C, steel–polymer (1—PS, 2—PTFE) f 1
T, °C
K50 to C150 8C, steel–polymer (1, 2—PCTFE, 3—PP) f 3 1
2
1—vZ3.5!10K5 cm/s, 2, 3—vZ3.5!10K2 cm/s
T, °C
K40 to C20 8C, steel–polymer (1—PE, 2—PTFE) f 2
8C
1
T, °C
20–200 8C, steel–rubber f
T, °C
918 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
case is commonly called two-body abrasion, and the latter,
three-body abrasion.
Abrasion displays scratches, gouges, and scoring marks
on the worn surface, and the debris produced by abrasion
frequently take on the appearance of fine cutting chips
similar to those produced during machining, although at a
much finer scale. Most of the models associated with
abrasive wear incorporate geometric asperity descriptions,
so that wear rates turn out to be quite dependent on the shape
and apex angles of the abrasive points moving along the
surface. The sources of the abrasive solid are numerous, and
the nature of the abrasive wear in a given tribosystem will
depend to some extent on the manner in which the abrasives
enter the tribosystem: whether they are present in the
original microstructure as hard phases, enter the system as
contaminants from outside, or generated as debris from the
contact surfaces as they wear.
There are two distinct modes of deformation when an
abrasive particle acts on the plastic material. The first mode
is plastic grooving, often referring to as ploughing, in which
a prow is pushed ahead of the particle, and material is
continually displaced sideways to form ridges adjacent to
the developing groove. No material is removed from the
surface. The second mode is named cutting, because it is
similar to micromachining and all the material displaced by
the particle is removed as a chip.
There is another approach to description of abrasive
wear. Experiments have shown that the abrasive wear rate is
in proportion to 1/su3u where su is the ultimate tensile stress
and 3u is the corresponding strain (Fig. 7). The correlation
was found by Lancaster and Ratner [47,50] and is often
referred to by their names.
In two-body abrasion, some asperities produce plough-
ing, the rest shows cutting, depending on two controlling
factors: the attack angle of the particle and the interfacial
shear strength expressed as the ratio between the shear stress
at the interface and the shear yield stress of plastically
deformed material. In the case of three-body abrasion the
free abrasive readily penetrates the polymeric surface,
Fig. 7. The Ratner–Lancaster correlation for abrasion: su, 3u are the
ultimate tensile stress and elongation (from Briscoe BJ, Tribol Int; August
1981: 231–43).
which begins to operate as an emery cloth resulting in
increasing the wear of countersurface.
4.3. Adhesion wear and friction transfer
Adhesion wear results from the shear of the friction
junctions. The fundamental mechanism of this wear is
adhesion, important component of friction outlined above.
This wear process evolves in formation of adhesion
junction, its growth and fracture. A distinguishing feature
of this wear is that transfer of material from one surface to
another occurs due to localized bonding between the
contacting solid surfaces. Bely et al. [6] noted that the
transfer of polymer is the most important characteristic of
adhesive wear in polymers. It is reasonable that the
processes associated with other wear types (fatigue,
abrasion and so on) accompany the adhesive wear.
The phenomenon of friction transfer is observed for
nearly all materials (metals, ceramics, and polymers) and
their combinations. The point is that whether the transfer
produces an influence on tribological behavior of the
friction pair. In this case, the consequences of material
transfer may be significantly distinct [51–53]. If small
particles of micrometer size are transferred from one surface
to the other, then wear rate varies to only a small extent.
Under certain conditions, the situations take place when thin
film of soft material is transferred onto the hard mating
surface, for example, polymer on metal. Results of the
transfer may be as follows. If the transferred polymer film is
carried away from steel surface and is newly formed, the
wear rate is increasing. In the case that the film is held in
place, the friction occurs between similar materials that may
result in seizure. Spreading of polymer on steel shift gives
rise to abrupt jump of friction force, but the wear changes
insignificantly.
It has been known that under certain conditions the hard
material is transferred on the soft surface. For example,
bronze is transferred on polymer. The transferred hard
particles are embedded in soft material and serve as
abrasive, which scratches the parent’s material.
Polymers are most susceptible to friction transfer when
rubbing both against metals and polymers. As an illustration
let us consider friction between polytetrafluoroethylene
(PTFE) and polyethylene (PE) [6]. Experiments were
carried out on wear tester with cylindrical block-shaft
(conforming contact) geometry. It has been discovered that
PTFE is transferred in the form of flakes of very small size in
the initial period of friction. The thickness of the transferred
layer increases monotonically and then oscillates about a
mean value whose magnitude and amplitude of oscillations
depend on the test conditions, especially on load and sliding
velocity (Fig. 8).
The transferred polymer fragment may exhibit a wide
variety of forms depending on polymer properties and
friction conditions. For example, even near ideally spherical
particle is observed during adhesive wear. Such particle is
 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
Fig. 8. The thickness of the transferred layer of PTFE as a function of
friction time (loadZ0.05 MPa, sliding velocityZ0.35 m/s).
likely to owe its origin to flexibility of chain molecules of
the polymer, thermal condition at the interface, and some
other factors, which remain unknown.
Once more consequence of polymer transfer is a change
in roughness of both surfaces in contact. The roughness of
polymer surface undergoes large variation during the
unsteady wear until the steady wear is reached, while
metal surface roughness is modified due to transfer of
polymer [54].
4.4. Fatigue wear
Fatigue is known to be a change in the material state due
to repeated (cyclic) stressing which results in progressive
fracture. Its characteristic feature is accumulation of
irreversible changes, which give rise to generation, and
development of cracks. The similar process takes place at
friction accompanying nearly all the wear modes. A friction
contact undergoes the cyclic stressing at rolling and
reciprocal sliding. In addition, each asperity of friction
surface experiences sequential loading from the asperities of
countersurface. As a consequence, two varying stress fields
are brought about in surface and sub-surface regions with
different scales from the diameter of apparent contact area in
the first case to that of local contact spot in the second.
These fields are responsible for material fatigue in these
regions that leads to the generation and propagation of
cracks and the formation of wear particles. This process is
named friction fatigue. Unlike the bulk fatigue, it spans only
surface and sub-surface regions. The loss of material from
solid surfaces owing to friction fatigue is referred to as
fatigue wear. It has been known that the fatigue cracks are
initiated at the points where the maximum tangential stress
or the tensile strain takes place. The theoretical and
experimental studies show that under contact loading the
maximum tangential stress position is dependent on friction
coefficient.
With low friction coefficient, the point where the shear
stress is maximum is located below the surface (f!0.3).
919
Fig. 9. Fatigue damage of solid surface of epoxy resin at friction coefficient
fZ0.17 (from Bogdanovich PN, Sov J Frict Wear 1982;3(2).
When the friction coefficient increases (fO0.3), the point
emerges on surface. On the other hand, if a solid is subjected
to combine normal and tangential loading, the surface and
sub-surface regions appear where the tensile strain and
thereby frictional heating occur. Therefore, cracks may be
nucleated in surface and/or below it (Fig. 9).
The initiation of the fatigue cracks is assisted by defects,
which are responsible for stress concentration. These are
scratches, dents, marks and pits on the surface, and
impurities, voids, cavities in sub-surface region.
Both surface and sub-surface cracks which open due to
repeated stressing will gradually grow, join, cross each other
and meet the surface until wear debris, including spalls, are
detached after a certain number of stressing.
5. Conclusions
A widespread interest in plastics has grown in the mid-
20th century due to the features of their structure, specific
mechanical behavior, and considerable possibility to change
the polymer properties. But creep of polymers, strong
dependence of their properties on temperature, low heat
conductivity, and sensitivity to the environment often posed
numerous problems. Extensive studies took years to outline
the field of modern engineering in which the plastics can be
applied as tribological materials, more commonly in the
form coatings and solid lubricants. The latter are used either
in pure form or in composite and laminated structures [55].
Thin polymer films, e.g. self-assembled monolayers formed
by chemosorption and physical adsorption of organic
molecules (polymers) are prospective boundary lubricants
in the fast-growing area of memory storage devices,
microelectromechanical systems and other precision mech-
anisms [56,57].
Further progress in the field of friction and wear of
polymers and their composites should be based on solving a
number of important problems which will allow us to
establish the fine mechanisms which occur in the working
920 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
surfaces of friction pairs. It appears important to study the
structural changes, which at molecular level in the surface
layers, and to investigate tribo-chemical reactions. It is also
important to find methods of controlled regulation over the
structure and frictional properties of polymers, based on
physical premises and concepts.
Here, it should be noted that the modern devices (AFM,
SFA, and so on) allow one to solve the numerous problems
connected with friction and wear of polymers.
It is clear that the progress in engineering will provide a
lot of new opportunities in applications of plastics, so
research in their mechanical and tribological behavior will
be a challenging and fruitful field of science and technology.
Acknowledgements
This work was partially funded under INTAS grant 99-
0671 and SCOPES Project No. 7BYPJ065579. The authors
are thankful to Prof. V. Tsukruk (Iowa State University) and
N. Spencer (ETH-Zurich) for specimens covered with
polymeric and organic SAM coating which was used at
adhesion measurements.
References
[1] Briscoe BJ. Interfacial friction of polymer composites. General
fundamental principles. In: Klaus F, editor. Friction and wear of
polymer composites. Amsterdam: Elsevier; 1986. p. 25–59.
[2] Briscoe BJ. Friction of organic polymers. In: Singer IL, Pollok HM,
editors. Fundamental of friction: macroscopic and microscopic
processes. Dordrecht: Kluwer Academic Publishers; 1992. p. 167–82.
[3] Bowden FP, Tabor D. Friction and lubrication of solids. Oxford:
Clarendon Press; 1964 p. 544.
[4] Kragelskii IV. Friction and wear. Elmsford: Pergamon Press; 1982
p. 458.
[5] Buckley DH. Surface effects in adhesion, friction, wear, and
lubrication. Amsterdam: Elsevier; 1981 p. 631.
[6] Bely VA, Sviridenok AI, Petrokovets MI, Savkin VG. Friction
and wear in polymer-based materials. Oxford: Pergamon Press;
1982 p. 416.
[7] Godet M. The body approach. A mechanical view of wear. Wear
1984;100:437–52.
[8] Deryagin BV, Krotova NA, Smilga VP. Adhesion of solids. New
York: Consultant Bureau; 1978.
[9] Johnson KL, Kendall K, Roberts AD. Surface energy and the contact
of elastic solids. Proc R Soc 1971;A324:301–13.
[10] Deryagin BV, Muller VM, Toporov YuP. Effect of contact
deformation on the adhesion of particles. J Colloid Interf Sci 1975;
53:314–26.
[11] Johnson KL, Greenwood JA. An adhesion map for the contact of
elastic spheres. J Colloid Interf Sci 1997;192:326–33.
[12] Myshkin NK, Petrokovets MI, Chizhik SA. Simulation of real contact
in tribology. Tribol Int 1998;31:79–86.
[13] Israelachvili JN. Intermolecular and surface forces. London:
Academic Press; 1991.
[14] Gibson CT, Watson GS, Mapledoram LD, Kondo H, Myhra S.
Characterisation of organic thin films by atomic force microscopy—
application of force vs. distance analysis and other modes. Appl Surf
Sci 1999;144–145:618–22.
[15] Israelachvili JN. Adhesion forces between surfaces in liquids and
condensable vapours. Surf Sci Rep 1992;14:109–59.
[16] Myshkin NK, Grigoriev AYa, Dubravin AM, Komkov OYu,
Spencer ND, Tosatti M. Experimental equipment for interfacial
force and friction measurements of micro scale samples. Tribology
and lubrication engineering. In: 14th international colloquium
tribology, Esslingen, Germany, 2004. p. 73–8.
[17] Myshkin NK, Kovalev AV, Kovaleva IN, Grigoriev AYa. Phenom-
enological model of adhesion contact. In: Proceedings of tribology of
surface layers and coatings, Prague, Czech Republic, 2004. p. 3–12.
[18] Briscoe BJ, Fiori L, Pelillo E. Nano-indentation of polymeric
surfaces. J Phys D: Appl Phys 1998;31:2395–405.
[19] Shulga H, Kovalev A, Myshkin N, Tsukruk VV. Some aspects of
AFM nanomechanical probing of surface polymer films. Eur Polym J
2004;40:949.
[20] Kovalev A, Shulga H, Lemieux M, Myshkin N, Tsukruk VV.
Nanomechanical probing of layered nanoscale polymer films with
atomic force microscopy. J Mater Res 2004;19(3):716–28.
[21] Moore DF. The friction and lubrication of elastomers. Oxford:
Pergamon Press; 1972 p. 256.
[22] Bartenev GM, Lavrentev VV. Friction and wear of polymers.
Amsterdam: Elsevier; 1981 p. 212.
[23] Hutchings IM. Tribology: friction and wear of engineering materials.
London: Edward Arnold; 1992 p. 273.
[24] Yamaguchi Y. Tribology of plastic materials. Amsterdam: Elsevier;
1990 p. 362.
[25] Myshkin NK, Petrokovets MI. Mechanical behavior of plastics:
surface properties and tribology. In: Totten George E, Liang Hong,
editors. Mechanical tribology, materials, characterization, and
applications. New York: Marcel Dekker; 2004. p. 57–94. Chapter 3.
[26] Petrokovets MI. Effect of temperature on real contact area of rough
surfaces. J Friction Wear 1999;20(2):1–6.
[27] Barber JR. Indentation of a semi-infinite solid by a hot sphere. Int
J Mech Sci 1973;15:813–9.
[28] Greenwood JA, Williamson JBP. Contact of nominally flat surfaces.
Proc R Soc 1966;A295:300–19.
[29] Greenwood JF, Tripp JH. The elastic contact of rough spheres. ASME
J Appl Mech 1967;34:153–9.
[30] Myshkin NK, Petrokovets MI, Chizhik SA. Basic problems in contact
characterization at nanolevel. Tribol Int 1999;32:379–85.
[31] Shooter K, Tabor D. The frictional properties of plastics. Proc R Soc
1952;B65:661.
[32] Bowers RC, Clinton WC, Zisman WA. Frictional behavior of
polyethylene, polytetrafluorethylene and halogenated derivatives.
Lubrication Eng 1953;9:204–9.
[33] Shooter K, Thomas RH. Frictional properties of some plastics.
Research 1952;2:533–9.
[34] Rees BL. Static friction of bulk polymers over a temperature range.
Research 1957;10:331–8.
[35] Schallamach A. The load dependence of rubber friction. Proc Phys
Soc 1952;65B(393):658–61.
[36] Oloffson B, Gralben N. Measurement of friction between single fibers.
Text Res J 1947;17:488–96.
[37] Flom DG, Porile NT. Effect of temperature and high-speed sliding on
the friction of teflon on teflon. Nature 1955;175:682–5.
[38] Flom DG, Porile NT. Friction of teflon sliding on teflon. J Appl Phys
1955;26:1080–92.
[39] White NS. Small oil-free bearings. J Res Nat Bur Stand 1956;57:
185–9.
[40] Schallamach A. The velocity and temperature dependence of rubber
friction. Proc Phys Soc 1955;B66:1161–9.
[41] Milz WC, Sargent LE. Frictional characteristic of plastics. Lubrica-
tion Eng 1955;11:313–7.
[42] Tanaka K. Kinetic friction and dynamic elastic contact behavior of
polymers. Wear 1984;100:243–62.
[43] Fort T. Adsorption and boundary friction of polymer surfaces. J Phys
Chem 1962;66:1136–43.
 N.K. Myshkin et al. / Tribology International 38 (2005) 910–921
[44] Vinogradov GV, Bartenev GM, Elkin AI, Mikhailov VK. Effect of
temperature on friction and adhesion of crystalline polumers. Wear
1970;16:358–68.
[45] Ludema KC, Tabor D. The friction and viscoelastic properties of
polymeric solids. Wear 1966;9:329–48.
[46] King RT, Tabor D. The effect of temperature on the mechanical
properties and the friction of plastics. Proc Phys Soc 1953;B66:
728–37.
[47] Lancaster JK. Relationships between the wear of polymers and
their mechanical properties. Proc Inst Mech Eng 1968–1969;183:
98–106.
[48] Blau PJ. Friction and wear transitions of materials. New York: Noyes
Publication; 1989.
[49] Myshkin NK, Kim CK, Petrokovets MI. Introduction to tribology.
Seoul: Cheong Moon Gak; 1997.
[50] Ratner SB, Farberova II, Radyukevich OV, Lure EG. Correlation
between the wear resistance of plastics and other mechanical
properties. Sov Plast 1964;7:37–40 [in Russian].
921
[51] Makinson KR, Tabor D. The friction and transfer of polytetrafluor-
oethylene. Proc R Soc 1964;A281:49–61.
[52] Sviridenok AI, Bely VA, Smurugov VA, Savkin VG. A study of
transfer in frictional interaction of polymers. Wear 1973;25:301–8.
[53] Tanaka K, Uchiyama Y, Toyooka S. The mechanism of wear of
PTFE. Wear 1973;23:153–72.
[54] Jain VK, Bahadur S. Surface topography changes in polymer–metal
sliding Proceedings of international conference on wear of materials,
Dearborn 1979 p. 581–8.
[55] Pratt GC. Bearing materials. In: Bever MB, editor. Encyclopedia of
materials science and engineering. Oxford: Pergamon Press; 1986. p.
281–8.
[56] Tsukruk VV, Nguen T, Lemieux M, Hazel J, Weber WN,
Shevchenko VV, et al. Tribological properties of modified MEMS
surfaces. In: Bhushan B, editor. Tribology issues and opportunities in
MEMS. Dodrecht: Kluwer Academic Publishers; 1998. p. 607–14.
[57] Bhushan B. Tribology and mechanics of magnetic storage devices.
2nd ed. New York: IEEE Press; 1996 p. 1125.