Friction and wear behaviour of polymers in liquid hydrogen

Friction and wear behaviour of polymers in liquid hydrogen

Cryogenics 93 (2018) 1–6 Contents lists available at ScienceDirect Cryogenics journal homepage: Short communicat...

3MB Sizes 0 Downloads 24 Views

Cryogenics 93 (2018) 1–6

Contents lists available at ScienceDirect

Cryogenics journal homepage:

Short communication

Friction and wear behaviour of polymers in liquid hydrogen ⁎


Géraldine Theiler , Thomas Gradt Bundesanstalt für Materialforschung und -prüfung (BAM), Division 6.3 Macrotribology and Wear Protection, Unter den Eichen 44-46, D-12203 Berlin, Germany



Keywords: Polymers Friction Wear Hydrogen Cryogenic temperature

The tribological behaviour of polymer composites were investigated in liquid hydrogen at −253 °C and compared with previous results obtained in gaseous hydrogen at ambient temperature. Experiments reveal that the friction and wear mechanisms in cryogenic conditions are related to the low temperature properties of polymers and to the formation or inhibition of the transfer film. For chemically stable polymers such as PEEK, that does not transfer in hydrogen gas, the tribological behaviour is improved in cryogenic liquids compared to ambient temperature. For tribo-reactive materials and graphite filled polymers that form a homogenous lubricating film in gaseous hydrogen, the sliding performance declines under cryogenic condition. The influence of hydrogen temperatures is discussed in this paper in relation to material compositions and transfer film formation.

1. Introduction The development of hydrogen technologies is a key strategy to reduce greenhouse gas emission worldwide. One of the leading countries is Japan that first launched hydrogen-powered fuel cell cars [1] and aims to realize by 2020 a “Hydrogen Society” for the upcoming Olympics in Tokyo. Another leader in this field is Germany, that aims to reduce greenhouse gas emissions by at least 80 percent by 2050 compared to 1990 [2]. Power to Gas is a particularly promising system solution, in which hydrogen and methane from renewable electricity can be used equally in mobility, industrial, heat supply and electricity generation applications [2]. In the near future, a total of 400 hydrogen refueling stations are to ensure nationwide coverage across Germany by 2023 [3]. The booming of hydrogen applications implies new material combinations for cost-effective constructions that insure high safety and reliabilities in distribution and delivery. In tribosystems, polymers are used in joints, compression equipment, and valves as seals, rider rings or seats. Therefore, many efforts are being made to investigate material compatibility and performances in hydrogen [4,5], including tribological properties [6–11]. Author’s previous study on polyimides at ambient temperature showed that the chemical structure of the polymer as well as the counterpart’s roughness have a major influence on their tribological behaviour [10]. It was also observed that the lubricity of graphite was found to be more effective in hydrogen than in moist air. A further study on the sliding behaviour of PEEK composites indicated that the formation and adhesion of a thin and homogeneous transfer film strongly depends on the environment [11].

Corresponding author. E-mail address: [email protected] (G. Theiler). Received 17 July 2017; Received in revised form 12 March 2018; Accepted 7 May 2018 Available online 08 May 2018 0011-2275/ © 2018 Elsevier Ltd. All rights reserved.

Many studies report on the tribological properties of materials in cryogenic liquids [12–18], fewer in vacuum or inert gas [19–21]. Attempts to find an empirical relationship between the friction force and the temperature were made in the past decades and summarised in [22]. Most experiments showed lower friction and wear in LN2 compared to room temperature, attributed to the decreasing deformation. Concerning the friction and wear behaviour of polymers in LH2, only few studies are available [9,17,18,23]. The aim of this work is to compare tribological processes of polymer materials in liquid hydrogen at −253 °C with the ones observed in gaseous hydrogen at ambient temperature. 2. Experimental details The materials investigated in this study were co-polyimides based on benzophenonetetracarboxylic dianhydride and pyromellitic dianhydride (PI1, PI2) and PEEK polymers described in [10] and [11], respectively. Pure and graphite filled PI sintered composites were provided by Ensinger Sintimid GmbH. PEEK materials (Victrex 450G) were prepared by injection molding by the Institute for Composite Materials (IVW, Kaiserslautern). Material compositions and nomenclature are presented in Table 1. The sliding performances of these composites in various environments at ambient temperature have been previously reported in [10,11]. Shore D hardness was measured in air (∼22 °C) and liquid nitrogen (−196 °C) with a test setup corresponding to EN ISO 868. Pin-on-disc tests were performed in the cryogenic tribometer (CT2) described in [15] at 0.2 m/s and 1 m/s under 50 N. Experiments were conducted in high purity hydrogen gas at room temperature and in high

Cryogenics 93 (2018) 1–6

G. Theiler, T. Gradt

purity LH2 at −253 °C (H2O < 5 ppm; O2 < 2 ppm). Before testing, the cryogenic chamber was purged 3 times with high purity N2 to eliminate any environmental contamination. The contact area of the polymer composite is 4 × 4 mm2. Steel discs (AISI 52100, Ra = 0.22 µm) were used as counterbody. The wear of the composites was determined by the total weight loss after 5000 m sliding. At least two (in cryogen liquid) or four (at room temperature) experimental values were available to calculate the average friction coefficient and wear rate. Error bars indicate the variations of the friction or wear measurements. After sliding, worn surfaces of the polymer composites and discs were examined by means of optical microscope (Keyence, VHX-500) and scanning electron microscope (SEM, Zeiss Supra™40) equipped with energy dispersive X-ray spectrometer (Thermo NSS). Polymer samples were thinly coated with gold or platin before SEM analysis. Micro-ATR-IR analyses (attenuated total reflection) (Hyperion 3000, Bruker) were performed on the counterface with a germanium crystal. The wavelength range was set from 4000 cm−1 to 500 cm−1.

Table 1 List of materials and nomenclature. PI1 PI1Gr

unfilled PI1 PI1 + 15 wt% graphite


unfilled PI2 PI2 + 15 wt% graphite


unfilled PEEK PEEK + 15 wt% graphite

3. Results and discussion 3.1. Unfilled polymers Fig. 1. Hardness measurement in air (∼22 °C) and in liquid nitrogen (−196 °C).

Fig. 1 shows the results of the hardness tests. Each value is the arithmetic mean of 5 indents. As expected, the hardness of the polymers

Fig. 2. Progression of the friction coefficient of (a) PI1 and (b) PEEK in hydrogen at ambient and cryogenic temperature, at v = 0.2 m/s or 1 m/s.

Fig. 3. Friction (a) and wear rate (b) of unfilled polymers at ambient and cryogenic temperature in hydrogen at v = 0.2 m/s and 1 m/s. 2

Cryogenics 93 (2018) 1–6

G. Theiler, T. Gradt

Fig. 4. Optical microscopy images of the discs after tests in H2 (1) and in LH2 (2) against (a) PI1, (b) PI2 and (c) PEEK at 0.2 m/s and against (d) PEEK at 1 m/s.

unfilled polymers are presented in Fig. 4. Two tendencies are seen accordingly to the friction and wear behaviour. At room temperature in H2, a transfer film is built up with PI1, while PI2 and PEEK polymers hardly transfer onto the disc: Plate-like polymer debris is observed at the surface of the disc. These are matrix particles related to adhesive wear of pure polymer. For PI1 material, a transfer film is formed at ambient temperature in H2, which is associated with low friction and wear. The transfer mechanisms at room temperature were discussed by means of FT-IR spectroscopy analyses in [10] and [11] for PIs and PEEK materials, respectively. It could be shown that hydrogen environment promotes reactions between metal and the benzophenonetetracarboxylic dianhydride based polyimide (PI1) resulting in imide ring opening. On the other hand, no reactions were detected with PI2, due to its better chemical stability [10]. Like PI2, FT-IR analyses of the PEEK transfer did not detect any triboreactions nor transfer film after tests in hydrogen at ambient temperature [11]. Only isolated particles are present on the disc surface. In LH2, the transfer and friction mechanisms are different from the

increased at cryogenic temperature, by 7%, 9% or 13% for PI1, PI2 and PEEK, respectively. The friction behaviour of PI1 and PEEK is illustrated in Fig. 2. The effect of the velocity and temperature clearly depends on the polymer material. With increasing temperature and velocity, the friction behaviour significantly improves for PI1 from 0.5 at v = 0.2 m/s in LH2 to 0.05 in H2. On the contrary, the friction behaviour of unfilled PEEK drastically deteriorates from 0.1 to 0.7 with increasing temperature and sliding velocity. The average friction coefficients of the unfilled polymers against steel are presented in Fig. 3a. While the friction of PI1 drastically increases in LH2 compared to H2, lower friction values are obtained for pure PI2 and PEEK. Similar effects are observed on the wear as shown in Fig. 3b: the wear rate of PI2 and PEEK are much reduced in LH2 compared to ambient temperature, while it increases for PI1. The effect of the velocities on the wear rate is similar (Fig. 3b). Increasing the velocity decreases the wear rate of PI1 in LH2 and increases that of PEEK at both temperatures. Optical microscopy images of the steel counterparts tested against

Fig. 5. Optical microscopy of the polymer pin after tests against (a) PI2 and (b) PEEK at 0.2 m/s in H2 (1) and in LH2 (2). 3

Cryogenics 93 (2018) 1–6

G. Theiler, T. Gradt

Fig. 6. SEM images of the counterface (a, b) and the polymer pin (c, d) after tests against PEEK at 0.2 m/s in H2 (a, c) and in LH2 (b, d).

one in H2. For unfilled PEEK and PI2, lower friction coefficients and wear rates are obtained in LH2. This is in accordance with previous works [16,21,22] which suggested that as the Young’s modulus increases at low temperatures, the deformation component of friction decreases as well as the real contact area and therefore the adhesion term of the friction coefficient. Looking at the polymer pins (Fig. 5), abrasive ploughing is observed at the worn surface of PEEK and PI2 in LH2, while at ambient temperature, plastic deformation occurs at the surface of the pin. SEM images also show the contrast between the surface of the PEEK pins and the corresponding counterbody after tests in H2 and LH2 (Fig. 6). Abrasive wear in LH2 produces powdery debris at the surface of the disc. This transfer adheres mechanically to the counterface and protects the polymer pin from further wearing. Therefore, for chemically stable PEEK and PI2, the tribological behaviour is improved under cryogenic conditions compared to ambient temperature due to a change in the wear mechanism from adhesive at ambient temperature to abrasive at low temperature. Different tribological mechanisms appear with chemically active PI1, which forms a transfer film at ambient temperature in H2. In LH2,

no transfer is observed at the surface of the discs tested (Fig. 4), leading to higher friction and wear rate. This can be attributed to the higher shear strength of polymers at cryogenic temperature. This theory was supported by Gardos [24], who suggested that the friction of PTFE should be higher at low temperatures due to large “lumpy” transfer film particles adhering to the counterface [22]. This theory, however, was not validated for PEEK and PI2, despite their higher shear strength at cryogenic temperature. ATR-IR analyses of the PI1 transfer on the disc after H2 and LH2 experiments are presented in Fig. 7. The large absorption band at 3360 cm−1, associated with OeH and NeH stretching is present in both H2 and LH2 conditions. The imide carbonyl stretching at 1719 cm−1, however, is clearly reduced at cryogenic temperature. The loss of C]O group is associated with a small increase of the aromatic peak (C]C). This suggests that during the friction process in LH2, the stiff PI1 molecules undergo mechanical destruction, while thermal activated chemical reactions occur in H2 gas at ambient temperature. The lack of transfer layer in LH2 with PI1 produced a significant increase of the friction and wear in LH2. By increasing the velocity, two tendencies are observed again depending on the polymer: against PEEK, polymer filaments are seen at

Fig. 7. ATR-IR spectra of the transfer film or particles after tests against PI1 in H2 and in LH2, respectively.

Fig. 8. Progression of the friction coefficient of PEEKGr in hydrogen at ambient and cryogenic temperatures, at v = 0.2 m/s or 1 m/s. 4

Cryogenics 93 (2018) 1–6

G. Theiler, T. Gradt

Fig. 9. Friction (a) and wear rate (b) of graphite filled polymers at ambient and cryogenic temperature in hydrogen at v = 0.2 m/s and 1 m/s.

Fig. 10. Optical microscopy of the transfer discs after tests against PEEKGr at 0.2 m/s in H2 (a) and LH2 (b).

Surface analyses of the discs are presented in Fig. 10 for PEEKGr. A homogeneous film is observed in gaseous hydrogen at room temperature (Fig. 10a). As described in [10,11], the saturation of graphite dangling bonds produces an efficient lubricating film on the reduced steel disc, leading to low friction and wear rate. In LH2, however, the friction increases for all graphite filled composites. Optical microscopy of the counterface indicated, indeed, that the transfer film is not as homogenous in LH2 as it is in H2 (Fig. 10b). It can be also deduced from the SEM images of the pin that graphite release is prevented at low temperature (Fig. 11). Despite the lack of smooth lubricating film, which leads usually to higher wear, the influence of LH2 moderately affects the wear rate of graphite filled polymers, particularly for PI2Gr and PEEKGr. This is attributed to the increased hardness of the polymer matrix, but also to the iron-rich surface of the pin detected by EDX (Fig. 11), which protects the pin from wearing out.

the surface of the disc in H2 (Fig. 4d1). Due to high contact temperature, softening and deformation of the polymer occur at 1 m/s, inducing higher friction and wear rate values. In LH2, the degradation of the polymer is reduced due to a much better cooling effect of the environment. In contrary, stable low friction is achieved faster at 1 m/s than at 0.2 m/s for PI1 in H2. In this case, triboreactions are accelerated at higher velocities. The friction coefficient decreases with increasing sliding velocities. 3.2. Graphite filled polymers The progression of the friction behaviour of graphite filled PEEK is illustrated in Fig. 8. In hydrogen gas at ambient temperature, the sliding velocity does not affect the friction behaviour. At both sliding speeds, the steady state friction is reached after a running-in distance of about 1000 m. In LH2, only a small transition appears at 0.2 m/s, while the friction stays unchanged (at 0.3) at 1 m/s. The average friction and wear rate of graphite filled composites are shown in Fig. 9. The addition of graphite leads to a significant reduction of friction and wear rate in H2 comparing to the low performance of pure PEEK and PI2 (presented in Fig. 3). In LH2, however, the friction of all graphite filled polymers increases compared to H2 conditions, while the influence of the temperature affects slightly the wear rate depending on the composites. These effects are, however, much moderate comparing to the large variation observed for pure polymers. Furthermore, the wear rate of graphite filled PEEK is not affected by the sliding speed in H2 nor in LH2.

4. Conclusion The tribological characteristics of pure and graphite filled polymers were investigated in LH2 (−253 °C) and compared to the results obtained in gaseous hydrogen at ambient temperatures. Based on the current results, the following conclusions can be drawn: – The friction and wear mechanisms are different in LH2 from those in H2 and depend on the polymer structure, low temperature properties and transfer film formation or inhibition. 5

Cryogenics 93 (2018) 1–6

G. Theiler, T. Gradt

Fig. 11. SEM images and EDX maps of the polymer pin after tests against PEEKGr at 0.2 m/s in (a) H2 and in (b) LH2.

– For chemically stable polymers such as PEEK, the tribological behaviour is improved under cryogenic conditions compared to ambient temperature. The wear mechanism changes from adhesive at ambient temperature to abrasive at low temperature. – For tribo-reactive materials like PI1, thermal activated chemical reactions occur in hydrogen at ambient temperature, leading to the formation of a transfer film that promotes low friction and low wear. In cryogenic conditions, however, the sliding performance of this polymer are deteriorated due to the lack of transfer film formation. – Similarly, for graphite filled polymers, that show excellent sliding properties in gaseous hydrogen due to the formation of a homogenous lubricating film, the transfer mechanisms are impeded at cryogenic temperature, leading to poorer friction performance.



[10] [11] [12] [13] [14] [15]



The authors gratefully acknowledge financial supports from the Deutsche Forschungsgemeinschaft (DFG), under the project-no. GR1002/10-1.





[1] 2020.html. [2] German Energy Agency (Dena) < power-to-gas-strategy-platform.html > . [3] [4] [5] Yamabe J, Nishimura S. Hydrogen-induced degradation of rubber seals. In: Gangloff RP, Somerday BP, editors. Gaseous hydrogen embrittlement of materials in energy technologies. Woodhead Publishing; 2012. p. 769. [6] Sawae Y, Okada K, Morita T, Kurono Y, Sugimura J. Low wear of carbon fiber filled PTFE in gaseous hydrogen, In: Nordtrib 16 nordic symposium on tribology, Aarhus, Denmark. [7] Nakashima K, Yamaguti A, Kurono Y, Sawae Y, Murakami T, Sugimura J, et al. Effect of high-pressure hydrogen exposure on wear of polytetrafluoroethylen sliding

[20] [21] [22]




against stainless steel. Proc Inst Mech Eng, Part J-J Eng Tribol 2010;224(J3):285–92. Iwai T, Higashiyama T, Ida K, Shokaku Y. Friction and wear of rubber and PTFE in hydrogen atmosphere, In: Proc. world tribology congress 2009, Kyoto, Japan. p. B1–124. Theiler G, Gradt T. Polymer composites for tribological applications in hydrogen environment. In: Proc. 2nd int. conf. on hydrogen safety, San Sebastian, Spain, 11–13 Sept. 2007. Theiler G, Gradt T. Tribological characteristics of polyimide composites in hydrogen environment. Tribol Int 2015;92:162–71. Elsevier Ltd. Theiler G, Gradt T. Environmental effects on the sliding behaviour of PEEK composites. Wear 2016;368–369:278–86. Glaeser WA, Kissel JW, Snediker DK. Wear mechanisms of polymers at cryogenic temperatures. Polym Sci Technol 1974;5B:651–62. Michael PC, Rabinovich E, Iwasa Y. Friction and wear of polymeric materials at 293, 77 and 4.2 K. Cryogenics 1991;31:695–704. Bozet JL. Type of wear for the pair Ti6Al4V/PCTE in ambient air and in liquid nitrogen. Wear 1993;162–164:1025–8. Hübner W, Gradt T, Schneider T, Börner H. Tribological behaviour of materials at cryogenic temperatures. Wear 1998;216:150–9. Theiler G, Hübner W, Gradt T, Klein P, Friedrich K. Friction and wear of PTFE composites at cryogenic temperatures. Tribol Int 2002;35:449–58. Friedrich K, Theiler G, Klein P. Polymer composites for tribological applications in a range between liquid helium and room temperature. In: Sinha SK, Briscoe BJ, editors. Polymer tribology. London: Imperial College Press; 2008. p. 375–413. Zhang Z, Klein P, Theiler G, Hübner W. Sliding wear of polymer composites in liquid hydrogen media. J Mater Sci 2004;39(9):2989–95. Cook NK, Burris DL, Dickrell PL, Sawyer WG. Cryogenic friction behaviour of PTFE based solid lubricant composites. Tribol Lett 2005;20(2):109–13. Burton JC, Taborek P, Rutledge JE. Temperature dependence of friction under cryogenic conditions in vacuum. Tribol Lett August 2006;23(2). Wang Q, Zheng F, Wang T. Tribological properties of polymers PI, PTFE and PEEK at cryogenic temperature in vacuum. Cryogenics 2016;75:19–25. Theiler G, Gradt T. Friction and wear of polymer materials at cryogenic temperatures. In: Kalia S, Fu SY, editors. Polymers at cryogenic temperatures. Springer; 2013. p. 41–58. Wissander DW, Ludwig LP, Johnson RL. Wear and friction of various polymer laminates in liquid nitrogen and in liquid hydrogen. 1968, Nasa Technical report; 1966. Gardos MN. Self-lubricating composites for extreme environmental conditions. In: Friedrich K, editor. Friction and wear of polymer composites, Pipes RB, Series editor. Composites materials series, vol. 1. Amsterdam: Elsevier; 1986. p. 397–447.