Tribology International 37 (2004) 271–277

www.elsevier.com/locate/triboint

Wear of PEEK composites related to their mechanical

performances

5

Z. Zhang

, C. Breidt, L. Chang, K. Friedrich

Institute for Composite Materials (IVW GmbH), University of Kaiserslautern, Erwin Schroedinger Street 58, 67663 Kaiserslautern, Germany

Received 9 June 2003; received in revised form 18 August 2003; accepted 16 September 2003

Abstract

A series of polyetheretherketone-based composites was investigated, blended with different contents of polytetrafluoroethylene

and/or graphite, and reinforced with various amounts of short carbon fibres. The mixture of the PEEK with various fillers was
achieved by twin-screw-extruders. Thereafter, the composites were finally manufactured using an injection moulding machine.
Testing of the tribological properties of the PEEK composites was carried out on a block-on-ring apparatus. The dependence of
mechanical properties, e.g. Charpy impact resistance, fracture toughness, flexural modulus and strength, on various filler contents
of these composites was also investigated, which is believed to be of help towards a better understanding of the steps on how to
improve the composite’s wear resistance.
#

2003 Elsevier Ltd. All rights reserved.

Keywords: Tribological properties; Mechanical properties; Polyetheretherketone (PEEK); Short carbon fibre (CF); Graphite; Polytetrafluoroethylene
(PTFE)

1. Introduction

Polyetheretherketone (PEEK) is a tough semi-crys-

talline thermoplastic polymer with excellent mechanical
properties and, therefore, has been applied as matrix
material for high performance composites. Its out-
standing features, e.g. comparatively high toughness
and fatigue resistance even at elevated temperatures,
also favour the application of PEEK as an attractive
bearing material under various loading conditions. In
order to improve the friction and wear behaviour of
polymeric materials, one of the traditional concepts is
to reduce their adhesion to the counterpart material
and to enhance their hardness, stiffness and compress-
ive strength. This can be achieved quite successfully by
using special fillers. To reduce adhesion, internal lubri-
cants such as polytetrafluoroethylene (PTFE) and
graphite flakes are frequently incorporated. One of the

mechanisms of the corresponding reduction in the coef-
ficient of friction is the formation of a PTFE-transfer
film on the surface of the counterpart. Short carbon
fibres (CF) are used to increase the creep resistance and
the compressive strength of the polymer matrix system
used. Normally, the matrix should possess a high tem-
perature resistance and have a high cohesive strength.
Additional fillers that enhance the thermal conductivity
are often of great advantage, especially if effects of
temperature enhancement in the contact area are to be
avoided in order to prevent an increase in the specific
wear rate.

The influence of various kinds of reinforcements on

the wear performance of PEEK composites has been
investigated in the past two decades. As a pioneering
work, Briscoe et al.

[1]

reported an experimental study

of the friction and wear of a number of PTFE–PEEK
composites over a wide composition range. It was
their conclusion that a composition of PEEK

þ

10 wt:% PTFE could reach an optimum effect in both
frictional coefficient and wear rate. Another pioneering
work has been carried out by Voss and Friedrich

[2]

on

short glass and carbon fibre reinforced PEEK compo-

5

Partly presented in the Sixth International Tribology Confer-

ence, AUSTRIB 2002, Perth, Australia, December 2002.

Corresponding author. Tel.: [phone]; fax: [phone]-

7196.

E-mail address: [e-mail] (Z. Zhang).

0301-679X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.triboint.2003.09.005

sites. Compared to short glass fibres, carbon fibres
showed a better potential in enhancing the sliding wear
resistance. Friedrich et al.

[3]

also investigated the

effects of counterpart roughness and temperature on
the friction and wear. A comparison of the influence of
various types of short carbon fibres, i.e. PAN- or pitch-
based, was carried out by Flo¨ck et al.

[4]

. Lu and Frie-

drich

[5]

studied systematically the sliding friction and

wear of a number of PTFE–PEEK compositions and
short CF reinforced PEEK composites against smooth
steel counterparts. The conclusion drawn from this
study was that an optimum range for PTFE in PEEK
is 10–20 vol.%, whereas 15–25 vol.% is the favourable
filler content scope for short CF. A combination of 10
vol.% of PTFE and 10 vol.% of short CF, with
additional inclusion of 10 vol.% of graphite as a fur-
ther internal lubricant, was expected from this research
as an optimum composition with an overall good fric-
tion and wear performance within broad loading and
temperature ranges

[5]

. Some companies also supply

wear resistant PEEK composites on market with each
10% ‘‘weight’’ contents of short carbon fibre, PTFE
and graphite. However, systemic investigations are still
required to approve this purpose based on the varia-
tions of filler contents.

To continue these efforts, the composition of PEEK

þ

vol:% PTFE

þ 10 vol:% graphite þ 10 vol:% CF

was

selected as a benchmark in the present study. A ser-
ies of PEEK-based composites was considered, blen-
ded with different contents of PTFE and/or graphite,
and reinforced with various amounts of short CF.
The mixture of the PEEK with various fillers was
achieved

by

twin-screw-extruders.

Thereafter,

the

composites were finally manufactured by injection
moulding. Tribological properties were carried out on
a block-on-ring apparatus. The dependence of various
mechanical properties, e.g. Charpy impact resistance,
fracture toughness, flexural modulus and strength, on
different filler contents of these composites was also
investigated, which is believed to be of help towards
a better understanding of the steps how to improve
the composite’s wear resistance.

2. Experimental details

2.1. Materials and compounding

The PEEK matrix (PEEK 450G) was supplied by

Victrex, UK. The pitch-based short carbon fibre (Kur-
eha M-2007S), PTFE (Dyneon 9207) and graphite
flakes (Superior 9039) were selected as fillers. The mix-
ture of the PEEK with various fillers was achieved by
twin-screw-extruders at about 395

v

C with standard

screw configurations, but various screw diameters of
25, 27 or 30 mm, respectively, as well as a screw aspect

ratio of around L=D

¼ 40. For comparison purpose,

some compositions have been extruded using all four
twin-screw-extruders

offered

by

various

German

extrusion companies.

Fig. 1

gives a light microscopic

photograph of the granule of the benchmark compo-
sition, in which the distribution of various reinforce-
ments in the matrix is satisfactory.

The composites were finally manufactured using an

ARBURG

ALLROUNDER

injection

moulding

machine at a barrel temperature of 395

v

C. The injec-

tion pressure was kept constant at 1400 bar, the mould
temperature was fixed at 60

v

C, and a constant injec-

tion speed was applied for all composites.

A comparison of the same composition extruded by

different extrusion lines showed that the influence of
the various extruder configurations used could be
neglected, since quite similar mechanical and tribologi-
cal properties were achieved in all cases. These results
will be published somewhere else.

Table 1

summarises

all the compositions of ten PEEK composites, as well
as the plain PEEK, considered in this study.

2.2. Wear test

In order to determine the tribological properties of

PEEK composites, unlubricated sliding wear tests were
carried out on a block-on-ring apparatus

[3]

designed

and constructed at the IVW. A hardened and polished
carbon steel ring (German Standard 100Cr6) with a
diameter of 60 mm and an initial surface roughness of
Ra

¼ 0:1 lm served as counterpart. The ring was slid

against the composite specimens for 20 h under ambi-
ent conditions. The rotational speed and the pressure
were kept constant at 1 m/s and 1 MPa, respectively.
The specimen’s mass loss after experiments, Dm, was

Fig. 1.

Light microscope of granule of PEEK

þ 10 vol:% PTFEþ

10 vol:% graphite

þ 10 vol:% CF.

272

Z. Zhang et al. / Tribology International 37 (2004) 271–277

measured, and the specific wear rate _

W

W s of the material

was calculated by using the equation,

_

W

W

s

¼

Dm

qF

N

L

ðmm

3

=

Nm

Þ

in which q is the density of the specimen, F

N

is the nor-

mal load applied on the specimen during sliding, and L
is the total sliding distance. The inverse of the wear
rate is usually referred to as the wear resistance of a
material. Six specimens of each composition were mea-
sured. The specific wear rates for various materials
given in this paper are each an average value of these
six experimental results, with an error scatter of the
maximum absolute measuring error.

2.3. Mechanical properties and SEM fractography

Instrumented Charpy impact testing was performed

on a CEAST pendulum impact testing machine accord-
ing to DIN-ISO-179-2. Unnotched specimens with
dimensions of 80

10 4 mm

3

(with the distance

between the supports of 62 mm) were fractured by the
impact mass at a speed of 2.9 m/s and an impact input
energy of 4 J.

A Zwick universal testing machine was applied to

investigate the flexural modulus and strength under
3-point bending conditions. The specimens had dimen-
sions of 100

10 4 mm

3

, and DIN-ISO-178 was

applied. The test speed was kept constant at 2 mm/
min. Fracture toughness was measured using the com-
pact tension (CT) specimen geometry. The specimens
were cut in the dimensions of 60

60 4 mm

3

, and

pre-notches were made according to the standard of
DIN-ISO-13586. A constant test speed of 5 mm/min
was used during the measurements. All mechanical
and wear measurements were carried out at room tem-
perature.

A JEOL-5400 scanning electron microscopy (SEM)

helped in analysing the worn surfaces of PEEK compo-

sites after the tribological investigations mentioned
above.

3. Results and discussion

3.1. Enhancement on wear resistance

PTFE is one of the commercially used reinforce-

ments that could reduce the frictional coefficient, and,
due to this fact, sometimes also the wear rate of poly-
meric composites. The unique tribological properties of
PTFE are due to its peculiar molecular and morpho-
logical structure. One of the mechanisms of the corre-
sponding reduction in the coefficient of friction is that
PTFE could easily form a third-body transfer film
when sliding against steel counterparts

[6]

. PTFE itself

has a relatively low wear resistance due to its soft nat-
ure, which can be of a disadvantage when used as a
single material in some practical applications.

Graphite is also a potential candidate of reinforce-

ments which could form a transfer film on the sliding
counterpart. As one of the three forms of carbon,
graphite has a layer structure (carbon layer) in which
the atoms are arranged in a hexagonal unit cell within
each layer

[7]

. These layers are linked by weak van der

Waals bonds, which may be easily broken by shear
force under sliding conditions. Especially within the
carbon layer, graphite is quite stiff, and therefore, its
transfer film also exhibits a stiff behaviour compared to
that of PTFE.

Carbon fibres have a preferred orientation such that

the carbon layers are preferentially parallel to the fibre
axis. As a result, carbon fibres are mechanically strong
and electrically and thermally conducting along the
fibre axis

[8]

. Short CF have been reported to show a

better wear performance in PEEK matrix composites
compared to that of short glass fibres

[2]

. Therefore, it

has been selected to increase the creep resistance and
the compressive strength in the present study.

In the present study, the best effect derives from a

combination of all three kinds of traditional reinforce-
ments in the PEEK matrix. The dependence of the spe-
cific wear rate on the PEEK matrix volume content is
given in

Fig. 2

. The symbols represent the measured

values along with the experimental scatter. One matrix
content may cover more than one measuring point
because various kinds of reinforcements were involved,
e.g. compositions of nos. 3, 5, 6 and 11. It becomes
clear that the higher the total filler content is, the lower
is the specific wear rate of the PEEK composites. The
specific

wear

rate

has

been

reduced

from

6

10

6

mm

3

=

Nm of the plain PEEK to the range of

4 8

10

7

mm

3

=

Nm with a filler content of 20–40

vol.%.

Table 1
Material compositions of PEEK composites

No. of
composition

Matrix
(vol.%)

PTFE
(vol.%)

Graphite
(vol.%)

Short CF
(vol.%)

1

100

0

0

0

2

70

10

10

10

3

65

15

5

15

4

75

5

5

15

5

65

15

10

10

6

65

10

10

15

7

80

0

5

15

8

80

5

0

15

9

60

10

10

20

10

77.8

5.55

5.55

11.1

11

65

5

5

25

Z. Zhang et al. / Tribology International 37 (2004) 271–277

273

In the current stage of specimens listed in

Table 1

,

composition no. 9 exhibits the best wear resistance of
about 4:1

10

7

mm

3

=

Nm with the highest filler con-

tent of 40 vol.%. Specimen no. 8 is also a potential can-
didate for further consideration, since it has a quite
low filler content of 20 vol.%, and at the same time a
relatively low specific wear rate of 5:6

10

7

mm

3

=

Nm.

3.2. Worn surfaces

Fig. 3

presents a comparison of worn surfaces of

(a, b) plain PEEK and (c, d) PEEK composite with
each 10 vol.% of PTFE, graphite and short CF. Plain
PEEK shows a river-like wave worn surface and the
plastic flow is obviously recognised, which may due to
the tough nature of plain PEEK. The situation is differ-
ent for PEEK composite that a lake-like smooth sur-
face is observed, which is believed may be due to
governed by the wear mechanism of short carbon fibre
thinning, cracking and peeling-off

[9]

. Reinforced by

short CF and modified by graphite and PTFE as inter-
nal lubricants, the wear mechanisms of PEEK compo-
sites become even more complicated. Nevertheless, the
PTFE and graphite reduce the adhesion between the
materials and counterparts, for this reason decrease the
frictional coefficient as well, and the short CF improves
the hardness and creep resistance, which finally lead to
a significant improvement of the wear resistance.

3.3. Correlations between tribological and mechanical
properties

Figs. 4 and 5

show the dependences of the specific

wear rate on the flexural modulus and strength,
respectively. The modulus has been enhanced about
50–120% by various amounts of fillers. It is clear that
the elastic CF and graphite contribute to the improve-
ment significantly. However, the viscoelastic PTFE
does not contribute to the modulus and reduces the
strength as well. The increase of the modulus does have

Fig. 2.

Dependence of the specific wear rate on the matrix volume

content in PEEK composites.

Fig. 3.

Worn surfaces of (a, b) plain PEEK and (c, d) PEEK composite with each 10 vol.% of PTFE, graphite and short CF.

274

Z. Zhang et al. / Tribology International 37 (2004) 271–277

a positive contribution to the wear resistance as shown
in

Fig. 4

. Higher modulus indicates lower wear rate

generally. The situation is different for the strength. As
shown in

Fig. 5

, the wear rate does not show a strong

dependence on the flexural strength. One composition
(no. 8) that is kept on a similar level of strength as
plain PEEK, however, exhibits a significant improve-
ment of wear resistance.

Figs. 6 and 7

mirror the results of the impact resist-

ance and the fracture toughness. The toughness of
plain PEEK is very striking. Even with a heavy impact
mass under an input energy of 15 J, the unnotched spe-
cimens could not be broken during the impact test;
therefore, no impact data of plain PEEK are given in

Fig. 6

. In general, all fillers reduce the impact resist-

Fig. 4.

Dependence of the specific wear rate on the flexural modulus

of PEEK composites.

Fig. 5.

Dependence of the specific wear rate on the flexural strength

of PEEK composites.

Fig. 6.

Dependence of the specific wear rate on the impact resist-

ance of PEEK composites.

Fig. 7.

Dependence of the specific wear rate on the fracture tough-

ness of PEEK composites.

Z. Zhang et al. / Tribology International 37 (2004) 271–277

275

ance

and

fracture

toughness

extensively

[10,11]

.

Regarding the toughness, the PEEK composition no. 8
maintains the best results in both the impact strength
and the fracture toughness, which is due to the viscoe-
lastic behaviour of PTFE.

Fig. 8

presents the dependence of the wear rate on

the density, which also confirms that the material com-
position plays a key role in wear performance.

3.4. Fractography

Fig. 9

gives various examples of fractographies of the

benchmark composition (no. 2) after (a) the Charpy
impact test, (b) the 3-point-bending measurement and
(c,d) the CT fracture toughness experiment. The PEEK
matrix exhibits a ductile fracture behaviour, which gov-
erns the crack propagation under static loading as
shown in

Fig. 9c,d

. However, the fracture surface

obtained by the Charpy measurement is completely dif-
ferent for same composite (

Fig. 9a

). The PEEK matrix

shows brittle fracture behaviour due to the high strain
rate, and the impact resistance of composites is likely
domain of the fibre cracking and pull-out.

4. Conclusions

In conclusion, the wear resistance of PEEK can be

significantly enhanced by the use of various reinforce-
ments (in particular, short carbon fibres, graphite
flakes and PTFE particles), but at the cost of a deterio-
ration of some other mechanical properties in some
degrees, e.g. toughness and strength. Various fillers
exhibit different routes to an improved wear resistance
in the composites. The lowest wear rate has been con-
firmed with a composition of PEEK

þ 10 vol:% PTFEþ

vol:% graphite

þ 20 vol:% CF under a sliding speed of

1 m/s and contact pressure of 1 MPa in the present
study.

The wear rates of PEEK composites show slightly

dependence on the material composites, modulus, den-

Fig. 8.

Dependence of the specific wear rate on the density of

PEEK composites.

Fig. 9.

Fractography of PEEK

þ 10 vol:% PTFE þ 10 vol:% graphite þ 10 vol:% CF after (a) Charpy impact test; (b) flexural bending test; and

(c) and (d) fracture toughness test.

276

Z. Zhang et al. / Tribology International 37 (2004) 271–277

sity and impact resistance, opposite to the flexural
properties and toughness, which are not so influential.
The results are different from that of neat polymers
either under an abrasive wear situation where the
Ratner–Lancaster correlation

[12,13]

can be applied, or

an equation used by Friedrich

[14]

to correlate the

erosive wear rate of polymers with the quotient of their
hardness to fracture energy.

Acknowledgements

The project was supported by the Alexander von

Humboldt

Sofja

Kovalevskaja

Award

program,

financed by the German Federal Ministry of Education
and Research (BMBF) within the German govern-
ment’s ‘‘ZIP’’ program for investment in the future.

References

[1] Briscoe BJ, Yao LH, Stolarski TA. The friction and wear of

PTFE–PEEK composites: an initial appraisal of the optimum
composition. Wear 1986;108:357–74.

[2] Voss H, Friedrich K. On the wear behaviour of short-fibre-rein-

forced PEEK composites. Wear 1987;116:1–18.

[3] Friedrich K, Kargar-Kocsis J, Lu ZP. Effects of steel counterface

roughness and temperature on the friction and wear of

PEEK composites under dry sliding conditions. Wear;148:
235–47.

[4] Flo¨ck J, Friedrich K, Yuan Q. On the friction and wear behav-

iour of PAN- and pitch-carbon fibre reinforced PEEK compo-
sites. Wear 1999;225–229:304–11.

[5] Lu ZP, Friedrich K. On sliding friction and wear of PEEK and

its composites. Wear 1995;181–183:624–31.

[6] Tanaka K. Effects of various fillers on the friction and wear of

PTFE-based composites. In: Friedrich K, editor. Friction and
Wear of Polymer Composites. Pipes RB, editor. Composite
materials Series, vol. 1. Amsterdam, The Netherlands: Elsevier;
1986, p. 137–74.

[7] Chung

DDL.

Graphite.

Journal

of

Materials

Science

2002;37:1475–89.

[8] Chand S. Carbon fibers for composites. Journal of Materials

Science 2000;35:1303–13.

[9] Stachowiak GW, Batchelor AW. Engineering Tribology, 2nd ed.

Jordan Hill, Oxford, UK: Butterworth-Heinemann; 2001.

[10] Karger-Kocsis J, Friedrich K. Effect of temperature and strain

rate on fracture toughness of PEEK and its short glass fiber
composites. Polymer 1986;27:1753.

[11] Friedrich K, Walter R, Voss H, Karger-Kocsis J. Effect of short

fiber reinforcement on fatigue crack propagation and fracture of
PEEK matrix composites. Composites 1986;17:205.

[12] Ratner SN, Farberoua II, Radyukeuich OV, Lure EG. Corre-

lation between wear resistance of plastics and other mechanical
properties. In: James DI, editor. Abrasion of rubber. London:
MacLaren; 1967, p. 145–54.

[13] Lancaster JK. Friction and wear. In: Jenkins AD, editor. Poly-

mer science and material science handbook. London: North
Holland Publishing Co.; 1972 [Chapter 14].

[14] Friedrich K. Erosive wear of polymer surfaces by steel ball blast-

ing. Journal of Materials Science 1986;21:3317–32.

Z. Zhang et al. / Tribology International 37 (2004) 271–277

277