Morphology and microstructure of polyamide 46 wear debris and transfer film: In relation to wear mechanisms

Page 1

Wear 265 (2008) 1100–1105

Contents lists available at ScienceDirect

Wear

journa l homepage: www.e lsev ier .com/ locate /wear

Morphology and microstructure of polyamide 46 wear debrisand transfer film: In relation to wear mechanisms

ent o

ostru), Fo

the phl thectionad twowearnsfers of

Peihong Cong, Fei Xiang, Xujun Liu, Tongsheng Li ∗

Key Laboratory of Molecular Engineering of Polymers of Ministry of Education, DepartmFudan University, Shanghai 200433, China

a r t i c l e i n f o

Article history:Received 26 January 2007Received in revised form 15 February 2008Accepted 10 March 2008Available online 25 April 2008

Keywords:FrictionMicrostructurePolyamide 46 (PA46)Wear debrisTransfer film

a b s t r a c t

The morphology and micrelectron microscopy (SEM(XRD) for understandingmethod was used to peemolecules happened direworn surface. PA46 formewhich related to differentthe crystal state in the traThe main wear mechanism

1. Introduction

The application of self-lubricating polymer-matrix compositesin tribological fields are tending to be more and more with the

development of new technology, especially in aviation and spacefields. Through the years, many fundamental tribological investiga-tions of various polymers have been carried out for understandingthe tribological characteristics of polymers and their composites[1–5]. In most practical systems, the sliding contact often occursbetween a polymer and a metal. This is a preferred arrangementbecause of the incompatibility of the mating materials as wellas the good heat transfer characteristics of the metallic material.Wear debris, or wear particles, are inevitably generated when thematerials rub against each other. Their shape, color, dimension,composition, quantity, morphological features and structural char-acteristics are closely correlated to the wear mode of materials. Onthe other hand, a transfer film often forms on the mating metallicmaterials. Factually, wear debris and transfer film serve as mediacarrying extensive information on the friction and wear process.

The investigation of wear debris has been considered importantfor the research of polymer tribology [6–9]. Usually, the followingthree aspects are considered: (1) forming process of wear debris;(2) changes of the wear debris in the contact interface and (3) rolesof the wear debris. Many aspects of the role of the wear debris are

∗ Corresponding author. Fax: +86 21 65640293.E-mail address: [email protected] (T. Li).

0043-1648/$ – see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.wear.2008.03.004

f Macromolecular Science,

cture of PA46 wear debris and transfer film were studied by using scanningurier transform infrared spectrometer (FT-IR) and X-ray diffractometerysical and chemical changes of PA46 caused by friction. An acid-eroding

transfer film from the mated steel ring surface. It was found that PA46l arrange along the friction direction both in the transfer film and on thedifferent shapes of wear debris: belt-like and spiral rod-like wear debris,mechanisms of PA46 during sliding process. Compared to the PA46 bulk,film and wear debris decreased and the amorphous structure increased.PA46 against steel were suggested to be adhesive wear and melting flow.

© 2008 Elsevier B.V. All rights reserved.

found to be common to different species of materials under differ-ent sliding environments. The wear mechanisms and the cause tothe wear can been effectively deduced through an examination ofthe wear debris.

As for the role of transfer films, it is widely believed that thetransfer films provide shielding of the soft polymer surface from thehard metal asperities [10–13]. As most polymers are self-lubricating

materials, the transfer film of polymers can act as a lubricant so thatthe friction coefficient is much lower as compared to that betweenmetal and metal. In these cases, the selectivity of the transfer filmsand the stability and thickness of these films during repetitivesliding govern the tribological behaviors of the tribosystem. Verycomplex tribochemical phenomena were also observed to occuron the frictional interface, which related to the friction and wearproperties [11,12].

Polyamide (PA) is a commonly used material in many tri-bological applications because of its perfect balance of variousmechanical properties and many complementary attributes. It hasbeen reported that PA has superior wear resistance sliding againsta steel counterface relative to other polymers [14]. The frictioncoefficient of PA was affected greatly by load, sliding speed andtemperature, and the friction coefficient values as high as 1.0–2.0under certain rubbing conditions were reported [14]. For reductionthe friction of PA, some polymers such as high density polyethylene(HDPE) were often blended with PA. The tribological behaviors ofthe PA polyblends have widely been investigated [15–19]. However,little has been paid on the wear and transfer behaviors of pure PAbulk.

Page 2

r 265

P. Cong et al. / Wea

In this paper, the morphology and microstructure of PA46 weardebris and transfer film were studied in detail for understandingthe wear mechanisms of PA46 bulk. An acid-eroding method wasused to peel the transfer film from the mated steel ring surface. Thetribochemical effects of PA46 bulk were also discussed based on theFTIR analysis results.

2. Experiment part

2.1. Sample preparation

PA46 with a density of 1.18 g/cm3 and a melting temperatureof 295 ◦C was supplied by DSM Co. Ltd. Before processing, thePA46 pellets were dried at 75 ◦C for 24 h in a vacuum oven, andthen the dried PA46 pellets were injection molded using a conven-tional injection-molding machine (SZ-30, Chinese Shanghai PlasticMachinery Works) equipped with a standard test mould. The tem-peratures maintained in the two zones of the barrel were 300 ◦Cand 300 ◦C, while the mould was at 25 ◦C.

2.2. Friction and wear tests

Friction and wear tests were conducted on a ring-on-block (M-200, Chinese Xuanhua Machinery Works) according to GB 3960-83.Fig. 1 shows the contact schematic diagram of the frictional pairs.The block specimen was PA46, with sizes of 30 mm × 7 mm × 6 mm.The mated ring specimen was 45# carbon structure steel with aninner diameter of 16 mm, an outer diameter of 40 mm and a width

of 10 mm. Before each test, both surfaces were polished with metal-lographic abrasive paper. Surface roughness Ra of the block and ringspecimens were about 0.20 �m and 0.10 �m, respectively. Prior tofriction tests, the specimens were cleaned in acetone for 15 min byusing an ultrasonic bath, and dried in air for at least 30 min. The fric-tion tests were performed at a speed of 0.42 m/s and a normal loadof 196 N under ambient conditions (temperature: 20 ± 3 ◦C, relativehumidity: 50 ± 10%). The sliding time was 120 min.

2.3. Wash method of PA46 transfer film formed on steel ringsurface

PA46 transfer film formed on steel ring surface was peeled fromthe steel ring and washed using the following process. Firstly, thesteel ring after the friction test was cleaned with acetone/formicacid (30/70) mixed solution using an ultrasonic bath for 15 min.Then, the obtained suspending solution was washed by dilutedH2SO4 for three times, and filtrated to obtain peeled transfer film.Finally, the transfer film was washed to pH 7 by deionized water,and dried in a vacuum oven to obtain clean transfer film for analysisobjectives.

Fig. 1. Contact schematic diagram of the frictional couples.

(2008) 1100–1105 1101

2.4. Analysis methods

The morphologies of PA46 worn surface, transfer film and weardebris were observed using stereomicroscopy and scanning elec-tronic microscopy (SEM, JSM-5600LV, Japan JEOL Co.). All thepolymer surfaces and debris were sputter-coated with a thin layerof gold–palladium alloy prior to SEM observation.

The chemical structure of PA46 bulk, transfer film and weardebris was analyzed by means of Fourier transform infrared spec-trometer (FT-IR, ESR Nicolet, NEXUS 470). The samples for FTIRanalysis were prepared using squashed method.

X-ray diffraction spectra of PA46 bulk, transfer film and weardebris were obtained with X-ray diffractometer (XRD, X’ PertPRO, Philips Analytical Instrument) using Cu K� radiation (40 kV,40 mA).

3. Results and discussion

3.1. Tribological property

Fig. 2 shows a typical variation of friction coefficient of PA46with sliding time. The initial friction coefficient of PA46 was as lowas 0.28. It increased to 0.86 when sliding about 10 min, and thendecreased to 0.73 after 20 min sliding, and remained the steady-state friction till the friction test was stopped. It can be seen that thevariation of friction coefficient of PA46 during sliding process canbe divided into two stages: run-into and steady stages. This frictionphenomenon is similar to most of the polymers, which is caused

by the removal of surface contamination layers and the changesof actual contact area. The friction coefficient of PA46 at steady-state was higher than that of PA6 (about 0.55 [12]) and PA66 (about0.67 [17]). This result may be due to the strong hydrogen bondingin PA46. It was observed that the steel ring surface changed intodeep purple and was very hot after sliding test, which indicatedthat a great number of frictional heat was produced during slidingprocess.

3.2. Morphology of PA46 wear debris

The formation of a large number of wear debris was alsoobserved during sliding process. Fig. 3 gives the SEM micrographs oftwo kinds of typical wear debris. Most wear debris showed belt-likeshape (Fig. 3a) with a width of 200–300 �m and a length of 2–3 mm.The long belt-like wear debris accumulated as folding belt. Somerod-like wear debris (Fig. 3b) with a diameter of 30–80 �m and alength of 100–300 �m were also collected. The surface of the rod-like wear debris was rough, and the section of the rod-like weardebris showed obviously spiral pattern from the SEM micrographat high magnification.

Fig. 2. A typical variation of friction coefficient of PA46 vs. sliding time.

Page 3

1102 P. Cong et al. / Wear 265 (2008) 1100–1105

Fig. 3. SEM micrographs of PA46 wear debris. (a) B

3.3. Morphology of PA46 worn surface

PA46 worn surface was eroded using formic acid for observationof the change of wear surface. Fig. 4 shows the morphology of thePA46 worn surface before and after eroded that was obtained bystereomicroscopy. It was observed that the worn surface was com-posed of rough and smooth strip regions, both the strip regions werealong the sliding direction. The rough strip region was peeled offfrom the bulk as long belt. This phenomenon indicated that the fric-tion caused the changes of physical or chemical structure of PA46surface layer.

Fig. 4. Morphology of PA46 worn surfaces befo

elt-like debris and (b) spiral red-like debris.

3.4. Transfer film analysis

SEM micrographs of the steel counterface sliding against PA46bulk before and after eroded with formic acid are given in Fig. 5.It was observed that PA46 formed a ‘lumpy’, discontinuous film(Fig. 5a), indicating that PA46 was easy to transfer on the steel coun-terface to form transfer film. This result may be due to the polarPA46 molecules, which can adhere to the counterface stronger ormay react with the mated steel surface.

The steel ring that covered with the discontinuous transfer filmswas treated in formic acid using an ultrasonic bath for 15 min.

re (a) and after (b) erode by formic acid.

Page 4

P. Cong et al. / Wear 265 (2008) 1100–1105 1103

Fig. 5. SEM micrographs of PA46 transfer film on steel ring

Some transfer films survived after the cleaning process, the belt-liketransfer films which were parallel to the sliding direction attachingto the steel surface were observed (Fig. 5b) after being eroded withformic acid.

It is known that formic acid is a kind of solvent for the freechains of PA-type polymers. Therefore, most of the PA46 trans-fer film was removed because of the soluble effect of formic acid.However, many belt-like PA46 transfer films survived after beingeroded. This result inferred that PA46 molecules oriented in thetransfer film, and the oriented transfer film had different solubility

Fig. 6. SEM micrographs of peeled PA46 transfer film. (

surface before (a) and after (b) eroded by formic acid.

with the unoriented PA46. In our earlier study, a highly orientedPA46 transfer film was observed when a steel ring rubbed againstPA46/HDPE polyblend [19].

It was also observed that a large number of PA46 transfer filmwas peeled from the steel ring surface by formic acid. The peeledtransfer films were washed using the method given in Section 2.3.Fig. 6 gives the SEM micrograph of the transfer film. The thin belt-like transfer films with width of 5–50 �m and length more than300 �m were observed. They exhibited different morphology fea-ture compared to the wear debris shown in Fig. 3, indicating that

a) Low magnification and (b) high magnification.

Page 5

1104 P. Cong et al. / Wear 265 (2008) 1100–1105

a), transfer film (b) and wear debris (c).

bulk. These bands stem from the bending and wagging vibration ofmethylene sequences, and the changes of these bands relate to thetransition of crystal structure. All the above changes reflected thatfriction led to the change of PA46 crystal structure.

In the range of 1100–800 cm−1: the bands at around 1087 cm−1,1047 cm−1 and 879 cm−1 originated from the skeleton vibration

Fig. 7. FT-IR spectra of PA46 bulk (

some PA46 molecules in transfer film were oriented along the fric-tion direction.

3.5. FTIR and XRD analysis results

The washed transfer film and wear debris were analyzed usingFTIR and XRD to clarify the changes caused by friction. Figs. 7 and 8show the obtained FTIR and XRD spectra, respectively. For compar-ison, the results of PA46 bulk were also shown in the correspondingfigures.

All the characteristic absorption bands of the amide groupswere present in Fig. 7: at round 3299 cm−1 (hydrogen-bonded N–Hstretching), at round 3066 cm−1 (overtone of amide II), at round1637 cm−1 (C O, stretching), at round 1541 cm−1 (C–N stretch-ing), at round 690 cm−1 (NH-out-of-plane bending) and at round580 cm−1 (C O out-of-plane bending). The strong absorption bandsat 2945 cm−1 and 2874 cm−1 belong to the CH2 segments in themain chain of PA46 [20,21]. Comparing the FTIR spectra of PA46transfer film and wear debris to PA46 bulk, it was clearly seenthat the wavenumber and intensity of the characteristic absorp-tion bands of amide groups and CH2 segments changed. The bands

were divided into the ranges of 3500–3000 cm−1, 1500–1100 cm−1

and 1100–800 cm−1 and discussed in detail as following:In the range of 3500–3000 cm−1: the intensity of the peak at

around 3299 cm−1, which is associated directly with the strength ofhydrogen bond decreased in the PA46 transfer film and wear debris,indicating that hydrogen-bonded strength in the PA46 transferfilm and wear debris was weaker than that in the PA46 bulk. Theintensity of a wide adsorption band at around 3421 cm−1, whichoriginated from the free NH groups obviously increased, implyingthat some of the hydrogen bonds broke and the free NH groupsexisted. This result was in agreement with the peak change ataround 3299 cm−1.

In the range of 1500–1100 cm−1: the intensity of the peaks ataround 1478 cm−1 and 1416 cm−1 that are attributed to the scis-soring vibration of the CH2 unit next to the CO and NH groups,respectively decreased in the transfer film and wear debris. On theother hand, both the bands shift to the high wavenumber side. Allthese changes could be due to the increase of the strengthening ofthe liberation motion and the twisting of the CH2–amide bonds[22,23]. At the same time, the peaks at 1363 cm−1, 1282 cm−1,1200 cm−1 and 1140 cm−1 decreased in the wear debris than in the

Fig. 8. XRD spectra of PA46 bulk (a), transfer film (b) and wear debris (c).

Page 6

r 265

[

[

[13] S. Bahadur, The development of transfer layers and their role in polymer tribol-ogy, Wear 245 (2000) 92–99.

P. Cong et al. / Wea

of methylene units, which have a relationship with the amor-phous phase. The band at around 942 cm−1 belongs to “crystallinebands”. Comparing the spectrum of the wear debris to the bulk, itcan be seen that the band at around 942 cm−1 became dramati-cally weaker. In the contrary, the intensity of the bands at around1087 cm−1, 1047 cm−1 and 879 cm−1 markedly increased. Thesechanges confirmed that the friction resulted in the change of molec-ular arrangement, the amorphous phase in the wear debris wasmore than in the bulk.

The XRD spectra of PA46 bulk, transfer film and wear debrisare presented in Fig. 8. A wide peak, reflecting amorphous phasewas obtained for wear debris. For the transfer film, crystal peaks at� = 20◦ and � = 23◦, which were similar to the bulk were detected.However, the peaks intensity was weaker than that in the bulk. Theanalysis results indicated that the content of crystal phase in thetransfer film was less than that in the bulk, almost no crystal phaseexistence in the wear debris.

3.6. Discussion

It is well known that the flash temperature at the real contactarea is much higher than the average temperature at the slidinginterface. When shearing occurs within the bulk of one of the spec-imens rather than at the interface, it is the property of the softermaterial that determines the magnitude of the friction and it isthe softer material that is transferred to the harder [24]. In thiscase, a great number of PA46 was transferred onto the steel ringsurface during sliding process. From all the obtained informationsynthesized through the surface analyses, it was reasonable to inferthat the wear process of PA46 was as follows: (1) the PA46 surfacelayer happened melting flow because of the occurrence of greatfrictional heat, belt-like wear debris was formed by friction shear-ing force, meanwhile the transfer film was formed on the matedsteel ring; (2) the transfer film was peeled off by friction shearing

force, and resulted in the formation of spiral rod-like wear debris.The process of (1) and (2) repeated during sliding process, andhence the mixed wear particles with different morphologies wereobserved.

The transfer film undergoes repeated melting flow and coolprocesses during whole sliding process, and caused the crystalstructure changed. Compared to the PA46 bulk, the wear debriscontained more amorphous structure. PA46 molecules happenedto orient along the friction direction, either on the worn surface orin the transfer film.

4. Conclusions

The morphology and microstructure of PA46 wear debris andtransfer film were studied by means of SEM observation, FTIR andXRD analysis. The following conclusions can be drawn:

(1) PA46 molecules happened orientation along the friction direc-tion both in the transfer film and on the sliding surface.

(2) PA46 formed two kinds of wear debris with different mor-phology: belt-like and spiral rod-like wear debris. The former

[

[

[

[

[

[

[

[

[

[[

(2008) 1100–1105 1105

was formed by friction shearing force because of the frictionheat at sliding interface. The latter was suggested to comefrom the transfer film that was peeled off by friction shearingforce.

(3) Compared to the PA46 bulk, the crystal state in the transferfilm and wear debris decreased and the amorphous structureincreased.

(4) The main wear mechanisms of PA46 against steel were adhesivewear and melting flow.

References

[1] K. Tanaka, T. Miyata, Studies on the friction and transfer of semi-crystallinepolymers, Wear 41 (1977) 383–389.

[2] H. Voss, J.H. Magill, K. Friedrich, On the wear behavior of doubly-oriented poly-mers. Nylon 6, J. Appl. Polym. Sci. 33 (1987) 1745–1761.

[3] M.K. Kar, S. Bahadur, An investigation of the temperature rise in polymer–metalsliding, Wear 82 (1982) 81–92.

[4] B.H. Stuart, Tribological studies of poly(etherether ketone) blends, Tribol. Int.31 (1998) 647–651.

[5] J. Hanchi, N.S. Eiss Jr., Tribological behavior of a thermotropic liquid crystallinepolymer and in situ composites base on their blends under dry sliding condi-tions at elevated temperature, Wear 200 (1996) 105–121.

[6] M.G. Jacko, P.H.S. Tsang, S.K. Rhee, Wear debris compaction and friction filmformation of polymer composites, Wear 133 (1989) 23–28.

[7] H.M. Hawthorne, On the role of interfacial debris morphology in a conformingcontact tribosystem, Wear 149 (1991) 169–185.

[8] A.L. Galvin, J.L. Tipper, E. Ingham, J. Fisher, Nanometre size debris generatedfrom crosslinked and non-crosslinked ultra high molecular weight polyethy-lene in artificial joints, Wear 259 (2005) 977–983.

[9] M.Q. Zhang, Z.P. Lu, K. Friedrich, On the wear debris of polyetheretherke-tone: fractal dimensions in relation to wear mechanism, Tribol. Int. 30 (1997)87–102.

10] S.H. Rhee, K.C. Ludema, Mechanisms of formation of polymeric transfer films,Wear 46 (1978) 231–240.

[11] J. Gao, J. Zhao, H. Dang, Molecule structure variation in friction of stainlesssteel/PTFE and its composite, J. Appl. Polym. Sci. 36 (1988) 73–85.

12] J. Gao, Tribochemical effects in formation of polymer transfer film, Wear 245(2000) 100–106.

14] M. Watanabe, H. Yamaguchi, The friction and wear properties of nylon, Wear110 (1986) 379–388.

15] C.C. Chen, E. Fontan, K. Min, J.L. White, An investigation of instability of phasemorphology of blends of nylons with polyethylenes and polystyrenes andeffects of compatibilizing agent, Polym. Eng. Sci. 28 (1988) 69–80.

16] M. Palabiyik, S. Bahadur, Mechanical and tribological properties of polyamide6 and high density polyethylene polyblends with and without compatibilizer,Wear 246 (2000) 149–158.

17] Z.B. Chen, T.S. Li, Y.L. Yang, Y. Zhang, S.Q. Lai, The effect of phase structure onthe tribological properties of PA66/HDPE blends, Macromol. Mater. Eng. 289(2004) 662–671.

18] M. Palabiyik, S. Bahadur, Tribological studies of polyamide 6 and high-densitypolyethylene blends filled with PTFE and copper oxide and reinforced withshort glass fibers, Wear 253 (2002) 369–376.

19] F. Xiang, T. Li, X. Liu, P. Cong, R. Lv, Effect of compatibilizer on the transfersand tribological behaviors of PA46/HDPE polyblends, Tribol. Lett. 22 (2006)279–288.

20] X. Cui, D. Yan, Preparation, characterization and crystalline transitions of odd-oven polyamides 11, 12 and 11, 10, Eur. Polym. J. 41 (2005) 863–870.

21] X. Cui, W. Li, D. Yan, A study of the crystalline transitions of polyamides X 18,Polym. Int. 53 (2004) 2031–2037.

22] J. Jakes, S. Krimm, A valence force field for the amide group, Spectrochim. Acta27A (1971) 19–34.

23] R. Zbinden, Infrared Spectroscopy of High Polymer, Academic, New York, 1964.24] F.P. Bowden, D. Taber, The Friction and Lubrication of Solids, vol. I, Clarendon

Press, Oxford, 1950.