Failure mode in sliding wear of PEEK based composites

Failure mode in sliding wear of PEEK based composites

Wear 301 (2013) 717–726 Contents lists available at SciVerse ScienceDirect Wear journal homepage: www.elsevier.com/locate/wear Failure mode in slid...

4MB Sizes 0 Downloads 40 Views

Wear 301 (2013) 717–726

Contents lists available at SciVerse ScienceDirect

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

Failure mode in sliding wear of PEEK based composites R. Schroeder a, F.W. Torres b, C. Binder a, A.N. Klein a, J.D.B. de Mello a,c,* a b c

Materials Laboratory—Federal University of Santa Catarina, Santa Catarina, Brazil Whirlpool S.A/Embraco Unit, Joinville, SC, Brazil Laboratory of Tribology and Materials—Federal University of Uberlˆ andia, Uberlˆ andia, Brazil

a r t i c l e i n f o

abstract

Article history: Received 31 August 2012 Received in revised form 20 November 2012 Accepted 22 November 2012 Available online 3 December 2012

Polymers and polymer composites are commonly used as solid lubricants. Polyether–ether–ketone (PEEK) is an engineering thermoplastic with a very good combination of thermal and mechanical properties. Although PEEK presents relatively high friction coefficients during unlubricated sliding, the wear rates are remarkably low. Despite this potential as a good tribological material, the high performance abilities of PEEK have not been entirely realized in practice because PEEK-ferrous metal contacts may fail by scuffing and/or two-three body abrasion for certain tribosystems. The traditional design approach to reinforce PEEK has been to use fibres for strengthening and filler particles for lubrication. In this work three materials were analysed: PEEK, carbon-fibre (CF) reinforced PEEK and PTFE þGraphite þ CF filled PEEK. A series of different tribo-tests was carried out: (i) Reciprocating linear scuffing test with 7 N incremental loading every 10 min up to 910 N; (ii) constant load (10 N, 5 h duration), reciprocating sliding tests assessing friction coefficient and wear rates of both specimen and counter body; (iii) free ball micro-abrasion tests using soft, fine abrasive particles. The wear tracks were analysed using SEM and light interferometry in order to assess the morphology of the wear tracks during scuffing, abrasion and sliding tests. The aim was to identify whether the failure mode in the sliding wear of these PEEK based materials is dominated by scuffing or by abrasive mechanisms. Unfilled PEEK exhibited very low scuffing resistance and a large wear rate during reciprocating sliding wear and abrasion, presenting a clear tendency to fail by abrasion mechanisms. CF-reinforced PEEK also presented very low scuffing resistance but higher sliding and micro abrasive wear resistance. It seemed that the presence of CF enhanced protection against abrasion by minimizing plastic deformation, although flaky debris probably originated from subsurface cracking were observed. The addition of PTFE and Graphite to the CF-reinforced PEEK produced a sharp decrease in friction coefficient, high scuffing and abrasion resistance, with an almost non-measurable wear rate in the reciprocating sliding test. The behaviour observed for PTFE þ Graphiteþ CF filled PEEK suggested the transfer of a protective tribo-layer from the composite to the counter-body and vice-versa. & 2012 Elsevier B.V. All rights reserved.

Keywords: PEEK composites Sliding wear Scuffing wear Micro-abrasion wear

1. Introduction Despite the spectacular advances in tribology since the word was coined in the mid sixties, a considerable amount of energy is still dissipated just to overcome friction even in modern mechanical systems. A nice example was recently reported by Holmberg et al. [1] concerning passengers cars: about one-third of the fuel energy is wasted to overcome friction in engine, transmission, tires and brakes. Inefficient systems like those in passenger cars are found in many other industrial sectors such as power generation, mining, petroleum, heavy transportation, etc. The impact of these systems

n Corresponding author at: Laboratory of Tribology and Materials—Federal University of Uberlˆandia, Uberlˆandia, Brazil. Tel.: þ55 3432394009; fax: þ 55 3432394273. E-mail addresses: [email protected], [email protected] (J.D.B. de Mello).

0043-1648/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.wear.2012.11.055

involves not only economic losses but many others factors; they also contribute to an increase in environment degradation, e.g., with increasing amounts of CO2 emissions and over consumption of natural raw materials. Therefore, any action towards a reduction in friction and wear losses may have a large impact on society. Solid lubrication has been considered one of the most promising means to achieve better tribological behaviour in these critical modern systems [2]. The so-called ‘‘self-lubricating materials’’ have been used in industry for many decades, mainly in applications where fluid lubrication is not feasible (high vacuum and temperature, high speeds and loads, tight clearances, etc.) [3]. Even in lubricated contacts where hydrodynamic lubrication is supposed to be effective, the use of self-lubricating materials demonstrated a beneficial synergistic effect [4]. Therefore, tribologists tend to investigate the lubricity behaviour of unexplored or even of new materials for certain tribosystems. Particularly, polymers tribology has become a mature

718

R. Schroeder et al. / Wear 301 (2013) 717–726

technology and gained much attention mainly as a replacement material for steels [5]. It is already known that polymers can be rarely used in practice. On the other hand, polymer composites capable of resisting higher temperatures, speed, load and environment effects are competitive materials for metals or ceramics. Polyether-ether-ketone (PEEK) composites are certainly some of the most promising materials in current polymer tribology [6]. The traditional design approach of using fibres to reinforce and additive particles to improve lubrication has proved to be efficient in PEEK composites development [7]. At the end of the eighties Voss and Friedrich [8] showed that the wear behaviour of PEEK could be improved by short-glass and carbon fibre (CF) additions. The best results were obtained by using carbon fibres. Recently [9], the morphology of such carbon fibres added to PEEK was studied: long woven CF-reinforced PEEK exhibited both better friction coefficient and wear properties than short CF-reinforced composite. Ceramic particles were also successfully investigated as possible reinforcements for PEEK composites [10,11]. Like papers concerning the reinforcement of PEEK for strengthening many others focused on using fillers for lubrication, in particular PEEK/PTFE composites [12–16]. Although presenting reduced wear properties, Polytetrafluoroethylene (PTFE) is a well know solid lubricant which is recognized as being a good combination of thermal and chemical resistance. Therefore PTFE additions to the tough and stable PEEK produce a reasonable self-lubricating behaviour for certain blends. As an example, Bijwe [15] investigated the optimal fraction (up to 30%) of PTFE in PEEK; results showed an interesting decrease in friction coefficient and wear rate with increasing PTFE contents. In a different approach, Vail et al. [17] used high tenacity expanded polytetrafluoroethylene (ePTFE) filaments as both fibre reinforcement and reservoir for solid lubricants in PEEK. The wear rates and friction coefficient obtained from the ePTFE filled composites were better than conventionally filled PTFE–PEEK composites. Besides the many works performed on reinforcements and fillers, the systemic character of tribological behaviour has also been addressed in many investigations using PEEK composites. Some examples treated the quality of the counter face surface [18], the effect of cryogenic or elevated temperature [19,20] and also the influence of various gas environments [21]. Despite being a good tribological material, the high performance abilities of PEEK have not been entirely realized in practice because PEEK-ferrous metal contacts may fail by scuffing and/or two–three body abrasion for certain tribosystems. Despite the high volume of published work, the failure mode of the composites during sliding is still unclear and this is the main goal of our investigation. In this work, three materials based on Polyether-ether-ketone (unfilled PEEK and two PEEK composites) were submitted to a series of wear test configurations in order to assess their failure mode, in particular, aiming to address whether their failure mode in sliding leaned towards scuffing or towards abrasion; in addition to determine how the additives (fibres and solid lubricants) affect the tribological behaviour of PEEK under different wear configurations.

2. Materials and methods Three different commercially available proprietary materials were selected based on their characteristics: (i) unfilled PEEK (PEEK) which was used as a reference; (ii) a wear grade, carbon fibre reinforced PEEK (Composite A) and; (iii) a self-lubricating grade 10%PTFEþ10%Graphiteþ10%CF filled PEEK (Composite B).

Table 1 Mechanical and thermal properties of the materials used in this work. Material properties

PEEK

Composite A

Composite B

Tensile strength (MPa) Tensile elongation (%) Tensile modulus (GPa) Melting point (1C) Glass transition (1C) Thermal conductivity (W m  1 1C  1)

110 25 3.90 343 143 0.29

180 1.9 18 343 143 1.30

140 2.2 12.5 343 143 0.87

Table 1 presents nominal mechanical and thermal properties of the studied materials. All materials were produced by injection moulding (100  100  3 mm plate). A series of different tribo-tests, schematically presented in Fig. 1, were carried out: 1. Reciprocating linear constant load sliding tests; 2. Reciprocating linear scuffing tests; 3. Free-ball micro-abrasion tests. The reciprocating linear sliding tests Fig. 1a, were conducted in a CETR universal tribometer UMT. In the constant load mode, it was configured in a ball-on-plate mode, where the plate was the test-material and the counter body an AISI 52100 steel (|: 10 mm) ball. The test had duration of 5 h under a constant load of 10.0 N, with a reciprocation frequency of 2 Hz and stroke of 10 mm. It was conducted in air, without any fluid lubricant, in controlled humidity (50%) and at room temperature (2273 1C). The results are the average of, at least, 5 tests for each material. These experiments intended to access friction coefficient and wear rates of specimen and counter-bodies. The wear rates were calculated using wear volumes measured by using white-light interferometry (Zygo New View 7200). The second series of tests was carried out in order to investigate the scuffing behaviour of the three materials Fig. 1a. In this case, the scuffing resistance was defined as the extension of the lubricious regime (m o0.2). The tests used the same tribological parameters imposed to that of the sliding test. By increasing the normal load in increments of 7 N at 10 min interval the scuffing resistance was determined. In this study, the scuffing resistance was defined as the work (N m) at which the value of the friction coefficient first rose above 0.20. Again, the results were the average of 5 tests. Abrasive wear tests were carried out in a CSM CaloWears ‘free ball’ micro-abrasion tester [22,23]. In this method Fig. 1b, a sphere of radius R is rotated against a specimen in the presence of a slurry of fine abrasive particles. The geometry of the wear scar is assumed to reproduce the spherical geometry of the ball, and the wear coefficient k was calculated by using Eq. (1) [22]: 4



pb 64R  S  N

)

for

b|R

ð1Þ

where K is the coefficient of wear (mm3/N m); b the surface chordal diameter of the crater, S the sliding distance, N the normal load applied and R the radius of the ball. Hard martensitic steel bearing balls (AISI 52100, 20 mm in diameter) were used as a counter-body. They were ultrasonically cleaned before each test and induced a normal load of 0.2270.01 N, monitored by a load cell. The drive shaft rotation was kept constant at 150 rpm. Abrasive slurries of silica (SiO2) particles suspended in distilled water were used, at a concentration of 75 g of abrasive per cubic centimetre of water. The slurries were agitated continuously throughout each test to prevent the settling of the abrasive particles.

R. Schroeder et al. / Wear 301 (2013) 717–726

719

Fig. 1. Schematic drawing of the wear tests. (a) Reciprocating linear sliding (left scuffing tests—right constant load tests. (b) Free-ball micro abrasion.

0.40 PEEK

0.35 Friction Coefficient

A peristaltic pump fed slurry on to the top of the ball at a rate of approximately three drops per minute. In order to determine the permanent wear regime, sequential tests at a range of test times (3, 6, 9, 12, 15 and 18 min were carried out on the specimen at the same location. Wear tracks were analysed by scanning electron microscopy (SEM—Jeol) and by white-light interferometry in order to understand the wear mechanisms occurring in each tribosystem.

3. Results and discussions

Composite A

0.30 0.25 0.20 0.15

Composite B 0.10 0.05

3.1. Reciprocating sliding tests

0.00 0

2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 Test time (s) Fig. 2. Typical evolutions of the friction coefficient.

0.40 0.35 Friction Coefficient

Fig. 2 presents typical evolutions of the friction coefficient with test time for the studied materials. The friction coefficient of all samples rapidly reaches the steady-state condition and remains in such mode during the whole test (5 h). In order to avoid any influence of the running in period on friction behaviour the average friction coefficient was calculated only in the permanent regime (test time higher than 5000 s) and are presented in Fig. 3. The tribological behaviour in pure sliding was drastically influenced by the composition and properties of each material. Unfilled PEEK presented a relative high friction coefficient (m ¼0.3470.01) and this was modified by the addition of reinforcing fibres and solid lubricant particles. For CF reinforced PEEK (Composite A), the friction coefficient decreased a little to m ¼0.2970.01, whereas when filled with CF fibre þGraphiteþPTFE (Composite B) the friction coefficient sharply decreased to m ¼0.0970.01. Fig. 4 shows the wear volume variation of all studied materials after the reciprocating sliding tests. Wear was evidently different between the unfilled PEEK and the two composites. Wear rates were strongly influenced by the material composition Fig. 5. Composite B presented a wear rate (2.15 mm3 N  1 m  1 10  7)

0.30 0.25 0.20 0.15

0.34 0.29

0.10 0.05

0.09

0.00 PEEK

Composite A

Fig. 3. Average friction coefficient.

Composite B

720

R. Schroeder et al. / Wear 301 (2013) 717–726

Fig. 4. Typical wear scars after 5 h of reciprocating sliding tests.

Fig. 5. Wear rate in sliding tests.

three orders of magnitude lower than PEEK (1370.0 mm3 N  1 m  1 10  7) and even one order of magnitude lower than composite A (22.10 mm3 N  1 m  1 10  7) which in turn had a two orders of magnitude lower wear rate when compared to unfilled PEEK. The hard steel ball was very little affected by sliding tests. In fact, it was impossible to quantify any expressive and accurate wear scar within the 5 h test.

The reinforcement of the PEEK, e.g., with fibres (Composite A), increased its resistance to plastic flow, thus diminishing stick-slip and resulting in lower friction coefficients. This is in agreement with the results presented by Lu and Friedrich [24] for low temperature experiments. However, the friction coefficient induced by Composite A is still high (0.29) in particular for selflubricating applications, especially for oil-free systems. On the other hand, the addition of carbon fibres and solid lubricants to PEEK (composite B) promoted a significant reduction in the friction coefficient. This may be associated with a protective tribo-layer formed at the sliding interface which may be explained in terms of transfer mechanisms from the polymeric specimen to counter-body and vice-versa. The transferring mechanism in polymeric composites has already been widely discussed mainly in terms of two complementary effects [25,26]: (i) the protective effect against the metallic asperities found in the metallic ball which shift the wear mode from severe to mild and (ii) the marked reduction in the friction coefficients which diminishes the energy availability to cause plastic deformation in cases where the tribo-layer has lubricating effects. Fig. 6 shows the early steps of the formation of the tribo-layer, particularly when loose PTFE particles in the interface adhere to the specimen Fig. 6a and to the counter body Fig. 6b. As the PTFE particles are homogeneously dispersed in the volume of the

R. Schroeder et al. / Wear 301 (2013) 717–726

721

Fig. 6. PTFE debris originating the protective tribo-layer. (a) PTFE wear debris. (b) Counter body wear scar. (c) EDX spectra of region 1 in Fig. 6a. (d) EDX spectra of region 2 in Fig. 6a.

composite B, there is a continuous self-replenishment or resupply of solid lubricants to the contact, thus keeping the friction coefficient highly stable at low values. Fig. 7 summarizes the main aspects of the wear scars after the sliding tests. The low magnification images (left column) give an idea of the magnitude of the wear whereas the high magnification images focus on the wear mechanisms. Due to low resistance to plastic deformation PEEK presented high levels of wear as can be seen on the very large wear track Fig. 7a. The detailed analysis (Fig. 7a and b) revealed that smearing in the earlier stage of sliding (arrows Fig. 7a) and ploughing constituted the dominant mechanism in surface modifications, with typical grooves aligned in the direction of sliding mainly in the middle of the track. According to Zhang et al. [27] frictional heat generated during sliding provoked an increase of the contact temperature. Thus, the deterioration of mechanical properties such as hardness and shearing strength occur as a result of a rise in interfacial temperature which produces the high plastic flow. The grooves probably originate from the cutting effect of the hard asperities of the counter face and also from third-body polymeric debris that slide on the contact. Improvement of polymer debris hardness can occur during the reciprocating sliding process as a result of work hardening caused by repeated plastic deformation [27]. The debris is surely harder than the specimen and can act as an abrasive grain in the following sliding process. Composite A had a considerably small scar width Fig. 7c. Abrasive wear was not noticeable and the wear mechanism changed to the production of flaky debris. Composite B exhibited a wear track even smaller Fig. 7e. The wear scars were very smooth and almost indistinguishable by SEM. The carbon fibres clearly protruding in the virgin surfaces Fig. 7f, were almost indistinguishable inside the wear scars.

The decreasing of two orders of magnitude in the wear rate of composite A when compared to PEEK may be attributed to the higher strength obtained with CF addition. In other words, microploughing was largely inhibited in this material, but still with some surface damage persisted as it can be seen in Fig. 7d. The production of flaky debris instead of grooving indicates that the wear mode has changed to probably micro-cracking/surface fatigue. The surface fatigue behaviour of PEEK composites was recently investigated by Avanzini et al. [28] in severe rolling contact experiments and similar production of flaky debris from subsurface micro-cracking was also identified. However, despite presenting a brittle failure mode, it is important to emphasize that the wear resistance is greatly improved compared to the unfilled PEEK by the addition of Carbon fibres. Additionally, Greco et al. [9] also revealed that with fibre addition, the counter-body wear is to some extent increased as a consequence of the fracture of fibres and abrasive particles formation. However, as already mentioned, it was impossible to quantify any accurate wear scar in the balls within the 5 h test. The explanation for that may be linked to the low speed (0.4 m/s) and nominal contact pressure (PEEK¼0.07 GPa/Composite A ¼0.2 GPa/Composite B¼0.05 GPa, i.e., a 10 N normal load applied to a 10 mm diameter ball) applied in the present tests versus the high speeds (65 m/s) and loads used by Greco et al. [9]. It is reasonable to suppose that the tribo-layer formed on the surface of composite B and transferred to the counter-body during sliding protects the surface of the composite against the abrasive action of the hard asperities of the steel ball, thus contributing to the minimization of the abrasive wear. Furthermore, the reduction of the friction coefficient by the action of a PTFE tribo-layer led to less frictional energy dissipation, which is associated with less heat in the contact and less energy available for plastic deformation and/or crack propagation. The synergistic

722

R. Schroeder et al. / Wear 301 (2013) 717–726

Wear scar

Wear scar

Wear scar Wear scar

Fig. 7. Typical wear scars seen under low (left) and high (right) magnification: (a) and (b) PEEK, (c) and (d) Composite A, (e) and (f) Composite B.

0.35

80 PEEK

0.25

Composite A

0.20

60

Normal Load

50 40

0.15

30

0.10

20

Composite B

0.05

Normal Load ( N )

Friction coefficient

0.30

composites can be markedly influenced by the ratio of each polymer in the blend and therefore the behaviour reported here cannot be postulated for all PEEK/PTFE composites.

70

10

0.00 0

50

100 150 Sliding distance (m)

200

0 250

Fig. 8. Friction coefficient behaviour during the scuffing tests.

beneficial effect of all these aspects ensure a very low wear rate ((2.1570.43)  10  7 mm3/N m) and very smooth wear tracks, as seen in Fig. 7d and e. It is important to emphasize that, as pointed out earlier by Burris and Sawyer [16], the wear mechanism in PEEK/PTFE

3.2. Reciprocating linear scuffing tests Fig. 8 shows typical evolution of the friction coefficient and the applied normal load with sliding distance for the three studied materials submitted to reciprocating linear scuffing test. Unfilled PEEK presented a high friction coefficient (around m ¼0.30) from the beginning of the test i.e., it had very low scuffing resistance. Composite A showed some lubricity capability in the very beginning of the experiment but already in the second/third load steps, the frictional forces increased markedly. In a completely different manner, Composite B showed a very low friction coefficient (around m ¼0.09) during the whole experiment, which guaranteed higher scuffing resistance. Typical aspects of the wear tracks are presented in Fig. 9. PEEK presented a large wear scar with grooves typical of abrasive wear mechanism Fig. 9a. Composite A Fig. 9b, although showing equivalent thickness of the wear track, presented a quite different aspect: Abrasive wear was not noticeable, and the wear mechanism

R. Schroeder et al. / Wear 301 (2013) 717–726

Wear scar

723

Wear scar

Wear scar

Fig. 9. SEM images of wear tracks formed during reciprocating scuffing tests: (a) PEEK, (b) Composite A, (c) Composite B.

Scuffing Resistance (N.m)

100.000

10.000

1.000 16834

100 1105 10 15 1 PEEK

Composite A

Composite B

Fig. 10. Scuffing resistance of PEEK composites.

changed to the production of flaky debris. Composite B exhibited a small wear track Fig. 9c. The wear scars were very smooth and almost indistinguishable by SEM. The values of the scuffing resistance for each test were computed. At least five tests were performed for each material, and an average scuffing resistance was thus calculated. The results are summarized in Fig. 10. The scuffing resistance of composite B presented very small dispersion and was greater (16834 752 N m), by about three orders of magnitude when compared to the scuffing resistance of PEEK (15711 N m) and one order of magnitude when compared to composite A (11057759 N m). Typical sliding grooves from micro-ploughing were observed in the wear track of PEEK Fig. 11a. This is in agreement with the hypothesis of smearing in the genesis of the degradation followed by two/three-body abrasion, just as observed in pure sliding. It seems that a similar failure mode was acting in both constant and incremental loading experiments which can be inputted to

the intense elastic/plastic deformation on polymer/metal contacts associated with the large adhesive forces. These forces originated from the growth of the real area of contact caused by the high level of plastic deformation The increase in strength promoted by the addition of carbon fibres in composite A reduced the intensity of elastic/plastic deformation as well as the growth of the real area of contact. Again, an important formation of flaky debris, typical of subsurface micro-cracking, was evident in the worn surfaces Fig. 11b. As the stresses imposed in scuffing tests were greater than those acting in sliding tests, the associated damage was also more severe, leading to large scars and fractures, but did not change significantly the micro-mechanism of failure by micro-cracking/ surface fatigue acting in pure sliding. On the other hand, composite B exhibited an extremely high scuffing resistance. In fact the friction coefficient remained stable at low values up to the end of the incremental loading test (91 N) which corresponds to maximum load allowed by the tribometer used to perform the tests. The tribo-layer apparently resisted the increased load and there was no evidence of failure either by scuffing or by abrasive wear. The wear tracks kept very smooth, a consequence of the flattening of the surface topography Fig. 11c. The figure also shows the abundant presence of PTFE ‘‘islands’’. It looks like the combination of solid lubricants (PTFEþGraphite) and CF-fibres play a major role in this tribosystem. The great durability of the tribo-layer is certainly a combination of the continuous lubricant self-replenishment mechanism associated with the large bearing capacity of the fibre reinforced matrix. 3.3. Free-ball micro-abrasion tests The evolution of the wear coefficient with test time is illustrated in Fig. 12. This figure shows the results presented by unfilled PEEK when abraded by silica. The results are typical of all the tested conditions. Wear coefficients were obtained in four sets (PEEK 1, PEEK2, PEEK3 and PEEK 4) of 6 tests each. The wear

724

R. Schroeder et al. / Wear 301 (2013) 717–726

Fig. 11. Wear mechanisms acting on scuffing mode: (a) PEEK, (b) Composite A and (c) Composite B.

Fig. 12. Wear coefficient evolution with test time during micro abrasion tests for Composite B.

Fig. 13. Micro abrasion wear coefficients.

coefficient was evolved in a random manner with the test time up to 12 min. After that, the dispersion decreased and the evolution of the wear coefficient became roughly constant. In this paper, all the analyses were done in the permanent wear regime, which was represented by an averaged wear coefficient (k), obtained by the averaging of the results relative to the test times 12, 15 and 18 min, e.g., this coefficient is representative of an average of 12 tests carried out in each sample. Fig. 13 summarizes the results of free-ball micro-abrasion tests. Again, the smallest abrasive wear coefficients were presented by Composite B (k¼(1.19 70.09)  10  15 m3 N  1 m  1) whereas unfilled PEEK showed the highest (k¼(40.00 7 5.00)  10  15 m3 N  1 m  1). Composite A presented a wear coefficient (k¼8.5070.50  10  15 m3 N  1 m  1) 7 times higher than composite B but 5 times lower than PEEK which in turn had a wear coefficient one order of magnitude higher than composite B. Fig. 14 shows typical aspects of the abraded surfaces and illustrates the wear mechanism acting in accord with the material. A grooving wear mechanism similar to a two-body mechanism was found to be dominant for PEEK Fig. 14a. This process occurs in the micro-scale abrasion test when a significant proportion of the particles on the specimen/ball interface slide without rolling, thus producing a series of fine parallel grooves in the specimen surface [29]. Moreover, high magnification showed intense debris formation associated with plastic deformation Fig. 14b. Wear scars associated with composite A also presented the aligned grooves typical of sliding abrasion Fig. 14c, with formation of very fine debris Fig. 14d. In a different way, composite B presented very shallow and smooth wear scars with almost no generation of debris. Moreover, the randomly oriented carbon fibres were visible in the images of the wear tracks of both composites. In addition, reservoirs of PTFE (clear rounded regions) were clearly exposed on the wear track of Composite B. Unfilled PEEK, again, presented very low wear resistance against abrasive particles, particularly in terms of micro-ploughing. However,

R. Schroeder et al. / Wear 301 (2013) 717–726

725

Fig. 14. SEM images of wear tracks formed during the micro-abrasion experiments.

the large production of debris seen on the worn surface (Fig. 14b) was not typical for pure micro-ploughing, particularly if it is assumed that the material loss should be ideally zero [30]. It is reasonable to suppose that there are transitions in abrasive wear mode, probably caused by superimposition of interactions [31,32]. In fact, the continuous sliding of soft abrasives over the PEEK surface and the interaction of grooves may, at some moment, change the wear mechanism from pure plastic deformation to brittle fracture, thus increasing debris formation. Da Silva et al. [33] showed this superimposition of interactions in ductile abrasive wear mode of metals, and it is reasonable to suppose that the same variation in wear mechanisms may occur in the abrasion of polymers. In the case of composite A, the presence of grooves was not so evident indicating a wear mechanism transition, probably from micro-ploughing to a more fracture dependent mechanism, as a consequence of the increase in strength promoted by the addition of carbon fibres as already related. Reinforced by Carbon fibre and modified by graphite and PTFE as internal lubricants, the wear mechanisms of composite B became even more complicated. Despite presenting shallow and smooth wear scars with almost no generation of debris, a few tiny

grooves were still present in the wear scar of composite B (Fig. 14e). The better performance of composite B was a little surprising because both graphite and PTFE are very well known for their low mechanical resistance. In theory increasing the content of solid lubricants induces a reduction in global abrasive wear resistance. According to Zhang et al. [27], PTFE and graphite also reduce the adhesion between the materials and counterparts. For this reason PTFE and graphite additions decrease the frictional coefficient as well, and the Carbon fibre improves the hardness and creep resistance, which finally lead to a significant improvement of the wear resistance.

4. Concluding remarks The tribological behaviour, summarized in Table 2, was drastically influenced by the composition and properties of the material. Unfilled PEEK presented a relative high friction coefficient and this was modified by the addition of reinforcing fibres and solid lubricant particles. For CF reinforced PEEK (Composite A), the

726

R. Schroeder et al. / Wear 301 (2013) 717–726

Table 2 Summary of the tribological behaviour. Sample

Friction coefficient

Wear rate (mm3 N  1 m  1)  10  7

Scuffing resistance (N m)

Abrasive wear coefficient k (m3 N  1 m  1)  10  15

PEEK Composite A Composite B

0.34 70.01 0.29 70.01 0.09 70.01

1370.0 7 90 22.1 7 2.2 2.17 0.4

157 11 11057 759 168347 52

40.0 75.0 8.5 70.5 1.2 70.1

friction coefficient slightly decreased, whereas when filled with CF fibreþ GraphiteþPTFE (Composite B) the friction coefficient sharply decreased. In all studied configurations (sliding, scuffing and microabrasion) wear rates were also strongly influenced by the material composition. Composite B presented the best performance and unfilled PEEK presented the worst. In general the difference was in the order of magnitude range. The hard steel ball was very little affected and it was impossible to identify any expressive wear scar in it. Wear mechanisms and failure mode were much more dependent on material composition than on test configuration and severity. Plastic deformation, in particular sliding grooves typical for micro-ploughing, was always observed in the wear track of PEEK. This was due to the intense elastic/plastic deformation in polymer/metal contacts associated with the large adhesive forces. In the case of composite A, the presence of grooves was not so evident indicating a wear mechanism transition, probably from micro-ploughing to a more fracture dependent mechanism, as a consequence of the increase in strength promoted by the addition of carbon fibres. Reinforced by Carbon fibre and modified by graphite and PTFE as internal lubricants, the wear mechanisms acting on composite B became even more complicated. Despite presenting shallow and smooth wear scars, a consequence of the flattening of the surface topography, with almost no generation of debris, the wear scar always showed an abundant presence of PTFE reservoirs. These reservoirs, accordingly to the literature, give rise to a strong protective tribo-layer. The great durability of the tribo-layer is certainly a combination of the continuous lubricant self-replenishment mechanism associated with the large bearing capacity of the fibre reinforced matrix. Finally, the role of graphite particles was not discussed in the present study, but the question is of fundamental importance. This topic is still under investigation and will hopefully be clarified by future work.

[6] [7]

[8] [9]

[10] [11]

[12]

[13] [14]

[15]

[16] [17]

[18]

[19]

[20]

[21]

[22] [23]

Acknowledgements

[27]

The authors wish to thank the Laboratory of Electronic Microscopy (LCME) of the Federal University of Santa Catarina for carrying out SEM analysis. The authors also thank CNPq and Capes/Proex for financial support.

[24] [25] [26] [28]

References [29] [1] K. Holmberg, P. Anderson, A. Erdemir, Global energy consumption due to friction in passenger cars, Tribology International 47 (2012) 221–234. [2] C. Donnet, A. Erdemir, Historical developments and new trends in tribological and solid lubricant coatings, Surface and Coatings Technology 180–181 (2004) 76–84. [3] K. Miyoshi, Solid Lubrication: Fundamentals and Applications, Marcel Dekker Inc., New York, 2001, pp. 10–13. [4] A. Erdemir, R.A. Erck, G.R. Fenske, H. Hong, Solid/liquid lubrication of ceramics at elevated temperature, Wear 203/204 (1996) 588–594. [5] B.J. Briscoe, S.K. Sinha, Tribological applications of polymers and their composites: past, present and future prospects, in: B.J Briscoe (Ed.), Tribology of

[30] [31] [32]

[33]

Polymeric Nanocomposites, Elsevier, Amsterdam, 2008, pp. 1–14, Tribology and Interface Engineering Series 55. K. Friedrich, Z. Lu, A.M. Hager, Recent advances in polymer composites’ tribology, Wear 190 (1995) 139–144. K. Friedrich, Z. Zhang, A.K. Schlarb, Effects of various fillers on the sliding wear of polymer composites, Composites Science and Technology 65 (2005) 2329–2343. H. Voss, K. Friedrich, On the wear behaviour of short-fibre-reinforced PEEK composites, Wear 116 (1987) 1–18. A.C. Greco, R. Erck, O. Ajayi, G. Fenske, Effect of reinforcement morphology on high-speed sliding friction and wear of PEEK polymers, Wear 271 (2011) 2222–2229. G. Pan, Q. Guo, W. Zhang, A. Tian, Fretting wear behaviours of nanometer Al2O3 and SiO2 reinforced PEEK composites, Wear 266 (2009) 1208–1215. G. Lin, G. Xie, G. Sui, R. Yang, Hybrid effect of nanoparticles with carbon fibers on the mechanical and wear properties of polymer composites, Composites: Part B 43 (2012) 44–49. B.J. Briscoe, L.H. Yao, T.A. Stolarsky, The friction and wear of poly(tetrafluoroethylene)–poly(ether-ether-ketone) composites: an initial appraisal of the optimum composition, Wear 108 (1986) 357–375. B.H. Stuart, Tribological studies of poly(ether ether ketone) blends, Tribology International 31 (11) (1998) 647–651. Q. Wang, Q. Xue, W. Liu, J. Chen, The friction and wear characteristics of nanometer sic and polytetrafluoroethylene filled polyetheretherketone, Wear 243 (2000) 140–146. J. Bijwe, S. Sen, A. Ghosh, Influence of PTFE content in PEEK–PTFE blends on mechanical properties and tribo-performance in various wear modes, Wear 258 (10) (2005) 1536–1542. D.L. Burris, W.G. Sawyer, A low friction and ultra low wear rate PEEK/PTFE composite, Wear 261 (2006) 410–418. J.R. Vail, B.A. Krick, K.R. Marchman, W.G. Sawyer, Polytetrafluoroethylene (PTFE) fibre reinforced polyetheretherketone (PEEK) composites, Wear 270 (2011) 737–741. D.M. Elliott, J. Fisher, D.T. Clark, Effect of counterface surface roughness and its evolution on the wear and friction of PEEK and PEEK-bonded carbon fibre composites on stainless steel, Wear 217 (1998) 288–296. N.L. McCook, D.L. Burris, P.L. Dickrell, W.G. Sawyer, Cryogenic friction behavior of PTFE based solid lubricant composites, Tribology Letters 20 (2) (2005) 109–113. L. Mu, X. Feng, J. Zhu, H. Wang, Q. Sun, Y. Shi, X. Lu, Comparative Study of tribological properties of different fibers reinforced PTFE/PEEK composites at elevated temperatures, Tribology Transactions 53 (2010) 189–194. N.L. McCook, M.A. Hamilton, D.L. Burris, W.G. Sawyer, Tribological results of PEEK nanocomposites in dry sliding against 440C in various gas environments, Wear 262 (2007) 1511–1515. R.I Trezona, I.M. Hutchings, Three-body abrasive wear testing of soft materials, Wear 233/235 (1999) 209–221. K.L. Rutherford, I.M. Hutchings, Theory and application of a microscale abrasive wear test, Journal of Testing and Evaluation 25 (1997) 250–260. G. Zhang, H. Liao, H. Li, C. Mateus, J.-M. Bordes, C. Coddet, On dry sliding friction and wear behavior of PEEK and PEEK/SiC-composite coatings, Wear 260 (2006) 594–600. Z.P. Lu, K. Friedrich, On the sliding friction and wear of PEEK and its composites, Wear 181 (1995) 624–631/183 181 (1995) 624–631. S. Bahadur, The development of transfer layers and their role in polymer tribology, Wear 245 (2000) 92–99. G. Jintang, Tribochemical effects in formation of polymer transfer film, Wear 245 (2000) 100–106.  C. Petrogalli, Wear and rolling contact A. Avanzini, G. Donzella, A. Mazzu, fatigue of PEEK and PEEK composites, Tribology International 57 (2013) 22–30. J.D.B. De Mello, W.M. Silva Ju´nior, W. M., R. BINDER, Abrasive wear of steamtreated sintered iron, Wear 258 (2005) 166–177. K.L. Zum Gahr, Microstructure and wear of materials, Tribology Series 10, Elsevier Science Publishers B.V., Amsterdam, 1987, p. 137. J.A. Williams, Y. Xie, The generation of wear surfaces by the interaction of parallel grooves, Wear 155 (1992) 363–379. W.M. da Silva, H.L. Costa, J.D.B. de Mello, Transitions in abrasive wear mechanisms: effect of the superimposition of interactions, Wear 271 (5–6) (2011) 977–986. R.I. Trezona, D.N. Allsopp, I.M. Hutchings, Transitions between two-body and three-body abrasive wear: influence of test conditions in the microscale abrasive wear test, Wear 225–229 (1999) 205–214.