- Email: [email protected]
PA46 (aliphatic polyamide) contacts
Wear 301 (2013) 758–762
Contents lists available at SciVerse ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Running-in due to material transfer of lubricated steel/PA46 (aliphatic polyamide) contacts ¨ Matthias Scherge n, Jeanette Kramlich, Roman Bottcher, Tobias Hoppe 1 Fraunhofer IWM, Microtribology Center, W¨ ohlerstraße 11, 79108 Freiburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 August 2012 Received in revised form 5 November 2012 Accepted 9 November 2012 Available online 29 November 2012
The running-in of an aliphatic polyamide 46 (PA 46) pin versus a lubricated steel disk was studied by friction and real-time wear measurements with a pin-on-disk tribometer. It was shown that under certain boundary conditions a transfer film forms and significantly reduces friction and wear. In addition, total wear, wear rate, running-in time and coefficient of friction strongly depend on the roughness of the steel disk. In all experiments the system showed a pronounced running-in behavior, characterized by decreasing and, after a certain time, constant coefficient of friction and wear rate. When the roughness of the steel sample was high, abrasion and high friction predominated. For disks with small roughness material transfer became effective and friction and wear decreased accordingly. Material transfer was proven by radioactive labeling of the polymer and detecting its traces on the steel surface. In addition, photoelectron spectroscopy was applied to find fractions of the PA 46 amine groups in the wear track. & 2012 Elsevier B.V. All rights reserved.
Keywords: Polymer Steel Friction Wear
1. Introduction Polymers, particularly polyamides, are increasingly used in tribological applications, e.g. journal bearings. Under oil lubrication very low coefficients of friction can be achieved, e.g. 0.025–0.035 for PA 66 or 0.015–0.045 for PTFE (load 100 N, sliding velocity 1.0–2.5 m/s). However, apart from a few exceptions, the wear resistance is still insufficient [1,2]. Furthermore, the running-in behavior of most polymers in contact with a steel counter piece is unknown, thus lifetime as well as performance predictions are complicated. Most technically important polymers are partially crystalline thermoplastics, which are characterized by a high toughness, rigidity and tenacity [3]. For tribological applications aliphatic polyamides such as PA 46, PA 6 and PA 66 widely used materials. Especially the chemical structure of PA 46 offers an advantage compared to PA 6 and PA 66 [4]. Polyamides are characterized by recurring carbon amide groups in the main chain, which are responsible for the physical differences in contrast to most other polymers. The carbon amide groups with their hydrogen bonds are the reason for outstanding mechanical properties and good chemical resistance. The extremely short hydrogen bonds between the NH-group and the CO-group are responsible for the high melting point of PA 46 of 295 1C [3]. Moreover, the high degree n
Corresponding author. Tel.: þ49 721 4640 750; fax: þ49 721 4640 111. E-mail address: [email protected] (M. Scherge). 1 Current address: Robert Bosch GmbH, (GS-AM/PJ-MIC), Postfach 30 02 40, 70442 Stuttgart, Germany. 0043-1648/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.wear.2012.11.035
of crystallization provides an improved pinning of the molecules and thereby improved resistance against abrasion [5]. The mechanical performance, as well as the friction and wear behavior, can be adjusted by fillers interacting with the polyamide matrix [3]. These fillers are able to further improve the chemical and physical properties of polyamide. Even with a very low filling degree of less than 5%, nanoparticle fillers are responsible for significant changes of the composite behavior [5]. For instance, metallic powders enhance the heat resistance of polymers [4]. However, these changes come into effect only if the filler is dispersed homogeneously [6]. Agglomerated fillers with dimensions larger than 1 mm can worsen the properties of the polymer. The tribological properties of polymers sliding against steel without lubrication are mainly influenced by the low shear strength of the polymer. It is widely accepted that polymer transfer to the surface of the counter body is responsible for low friction and small wear [7]. This means that the shear plane is shifted from the steel surface into the polymer layer on top of the steel [8]. Material transfer calls for adhesion to attach fractions of the polymer to the steel surface. To understand this mechanism, results from experiments with polymers grafted to a surface can be used [9,10]. It is assumed that polar end-groups of the polymer form physical or chemical bonds with the surface. Under equilibrium conditions a polymer brush is formed which reduces the shear stress upon sliding. Usually the monomers to be grafted are dissolved in a liquid or originate from the gas phase during glow discharge. To initiate such a process in a pin-on-disk experiment a sufficient amount of energy is necessary to remove the water film
M. Scherge et al. / Wear 301 (2013) 758–762
present on every metal surface and to form a polymer film. The same applies to a lubricated steel surface, where layers of oil have to be removed prior to the attachment of polymer chains to its surface. The objective of this investigation was to understand the running-in behavior. Since we suspected that the running-in results from balancing abrasion and transfer, steel disks with different roughnesses were prepared [11]. The running-in behavior was monitored by continuous wear measurement applying the radionuclide technique (RNT) with a pin-on-disk tribometer. Autoradiography was used to detect transferred tracer atoms coming from the polymer on the steel surface. In addition chemical analysis was carried out to find traces of amine groups at the steel surface.
2. Experimental procedure 2.1. Sample preparation In order to apply continuous wear measurement using the radionuclide technique, PA 46 was filled with 9% Fe2O3-nanoparticles during extrusion to serve as tracer after weak radioactive irradiation, see Section 2.2. The nanoparticles were added to the extruder in liquid form. To prevent coagulation, the nanoparticles were functionalized before immersion into the liquid. Functionalization is a process by which a chemical species can be made to react or not to react in a certain way by activating certain sites on a surface or a molecule. Anti-coagulation in this case was done by adding a suitable non-polar group. After extrusion the received pellets were molded into a bar as base material for the pins and for accompanying analysis such as the determination of tensile strength, Young’s modulus and the measurement of the glass temperature. The bars have a tensile strength of 73.65 MPa, a Young’s modulus of 2.98 GPa and a glass temperature of 40–50 1C. PA tablets with a diameter of 5 mm and a thickness of 4 mm were machined without visibly changing the original surface. All tablets were conditioned under ambient temperature and humidity. Polyamide is known to incorporate water into its amorphous regions. This is particularly the case for PA 46 with a nominal water consumption of about 12%. The water interacts with the microstructure and lowers the glass transition temperature [3]. This behavior influences both friction and wear. To analyze the arrangement of the nanoparticles in the polymer matrix, scanning electron microscopy (SEM) was used. Fig. 1 shows an SEM image of PA 46 with 9% nanoparticles, indicating
759
that the particles (white spots) are not agglomerated and distributed uniformly. The disks were made of steel (100Cr6) with a diameter of 200 mm and a thickness of 10 mm. The topography of the steel disks was achieved through a process of super-finish using a band tool (Supfina). By varying finishing times and grain size of the abrasives, disks with different average roughness were received (disk 1—0.08 mm; disk 2—0.14 mm; disk 3—0.27 mm). Roughness was determined with a confocal microscope (Sensofar). 2.2. Instrumentation Continuous friction and wear measurements were carried out with a pin-on-disk tribometer (Basalt SST, Tetra GmbH). The tribometer covered a range between 0 and 1000 rpm. Depending on the distance between pin and the center of the disk, sliding velocities up to 10 m/s were possible. The maximum normal force that can be applied to the pin was 1000 N. Both normal and tangential forces were recorded throughout the experiment. To measure wear, the oil circuit was connected to an RNT device (Zyklotron AG), see Fig. 2. To obtain a pin, a PA tablet was attached to a holder allowing a self-adjusting contact with the steel disk. Particles developed by wear were removed from the disk by the oil flow and were carried to the RNT device via oil circuit. Since the polymer was homogenously filled with Fe2O3-nanoparticles (see Fig. 1), the activation of a fraction of the iron atoms with thermal neutrons ensured a homogeneously labeled sample. In contrast to conventional cyclotron activation utilizing a focused beam of deuterons to generate nuclides, the activation in a stream of thermal neutron was a gentle way to obtain nuclides without weakening the polymer matrix. The radiation process delivered Fe-59 nuclides. The more nuclides arrive at the detector, the more wear is present. Through calibration, activity was related to the wear mass. One of the benefits of the RNT is its high resolution of a few _ micrograms per liter oil. From the wear mass mw the wear rate w was calculated by dividing the wear depth dw by the duration of the experiment t. _ ¼ w
dw mw ¼ t Art
ð1Þ
A is the contact area of the pin and r is the density of PA 46. Detailed information about the RNT method can be found in [12,13]. 2.3. Lubricant For the experiments two types of oil with similar viscosity were used to obtain optimum results. For wear analysis fullyformulated oils are preferable since their additivation suppresses sedimentation of wear particles. Therefore all wear tests were
Fig. 1. SEM micrograph of PA 46 with 9% nanoparticles.
Fig. 2. Pin-on-disk tribometer connected to a radionuclide wear measurement system.
M. Scherge et al. / Wear 301 (2013) 758–762
2.4. Test procedure To become familiar with the tribological properties of the system, first Stribeck curves were recorded to obtain the coefficient of friction (COF) as function of sliding velocity times viscosity divided by the contact pressure. After passing a parameter field with different normal forces and sliding velocities, areas of boundary and hydrodynamic lubrication were identified. All sample pairings were then subjected to constant normal force and sliding velocity to investigate the running-in behavior in the boundary lubrication regime. All recorded friction data were averaged over a period of 10 s. The oil temperature was set to 60 1C. In addition to the parameter fields, tests with the nitrogen-free reference oil were performed to analyze the steel disks with respect to the anticipated polymer transfer. Since the oil did not contain a significant percentage of nitrogen, nitrogen in the wear track would indicate the deposition of amine groups coming from the polymer.
3. Results 3.1. Friction and wear tests Fig. 3 shows the friction coefficients for disk 1, 2 and 3 as function of sliding velocity divided by the contact pressure. Viscosity was treated as constant and omitted. It was shown that the largest differences between the friction curves appeared in the boundary lubrication regime. Under hydrodynamic lubrication (v/p 40.6) all curves behave similarly. Steel disks with larger roughness show more pronounced boundary lubrication due to increased local stresses. Based on these findings the coefficient of friction and the wear _ were analyzed for a constant normal force of 300 N rate w
0.10 0.09 0.08 COF Ra = 0.27µm
0.07
COF Ra = 0.14 µm
COF
0.06
COF Ra = 0.08 µm
0.05 0.04
disk 1 2 3
0.12 0.11 0.10 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00
COF 0.02 0.04 0.09
dw/dt [µm/h] 0.04 0.6 6
10
400
8
3
4 2
1
2 0 0
2
4
6
8
10
12
14
200
Wear [µm]
300 6
Wear [mg]
performed with FUCHS TITAN GT1 5W30 LongLife III (ACEA A3). To analyze the transfer process, however, a nitrogen-free oil (Fuchs PAO) was used. This oil contains mainly tridecane, undecane and heptadecane. The contents of nitrogen is negligibly small (0.83 wt% with a standard deviation of 0.38 wt%) and originates presumably from the intake of air during chemical analysis (Thermo Flash EA measurement). Thus, traces of nitrogen in the wear track should originate from the PA 46 amine groups.
COF
760
100
0
16
Time [h] Fig. 4. Friction and wear as function of time. The dotted curves represent the wear behavior. The error of the coefficient of friction of experiment and repetition is 0.01. The error of wear is smaller than the size of the symbol.
(8 15 MPa) and a sliding velocity of 1 m/s. This corresponds to a v/p ratio of 0.07 which clearly leads to boundary lubrication. Fig. 4 shows the coefficient of friction and the total wear as function of time and roughness of the steel disks. All tested pairings showed a distinct running-in behavior similar to lubricated metal–metal contacts. In addition to the pronounced running-in behavior friction and wear show a clear dependence on the roughness of the steel disk. For disk 3 (highest roughness) the level of total wear is about a hundred times higher than for metal–metal contacts operated under similar stressing conditions. With decreasing roughness both friction, total wear and wear rate decrease. All initial coefficients of friction start at values of about 0.1. Shortly after the start of experiment the coefficients of friction significantly drop, which can be attributed to the topographical running-in. The disk with the lowest roughness achieves a wear rate of 0.04 mm/h and a coefficient of friction of 0.02 at the end of the experiment. All other wear rates and friction data are shown as inset in Fig. 4. 3.2. Topography of PA and steel The topography of the PA samples was examined before and after tribological testing. The initial topography of the polymer surface, measured with a confocal microscope, is determined by the roughness of the mold, see Fig. 5. After the tests the surfaces show slightly lower roughness, however, scattered grooves which determine the peak to valley roughness are still visible. With increasing roughness of the steel disk the number of asperities on the PA surface increases. Since PA 46 is rather soft, the tribologically induced pin roughness was attained shortly after the start of the experiment. A start-stop test showed that already after 5 min roughness had adapted to the new conditions. For the steel disks no change in roughness was measured. The roughness profiles were measured perpendicular to the friction tracks. 3.3. Autoradiography
0.03 0.02 0.01 0.00 0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
v/p [m/(MPa*s)] Fig. 3. Stribeck curves. The error of the coefficient of friction of experiment and repetition is 0.01.
The assignment of this technique is aimed at the detection of radioactive traces at the surface of the steel disk. By means of a gamma spectrometer with an array of highly-sensitive germanium detectors even the very low Fe-59 activity was detected by adjusting the exposure time. Fig. 6 shows on the left hand side the trace of nuclides on disk 1 (low roughness). On the right hand side disk 3 (high roughness) is shown. On both disks traces of the pin material were detected. Although no quantitative characterization
M. Scherge et al. / Wear 301 (2013) 758–762
761
Fig. 5. Topography of PA 46 before (top) and after (bottom) tribological stressing.
Fig. 6. Autoradiography measurements of the disks. Left—disk 1. Right—disk 3. The width of the wear track is 5 mm.
was possible, the blackening of the traces gives an estimate of the amount of the material transfer. Fig. 6 shows that the transferred material does not cover the steel surface uniformly. We assume that only the tops of the asperities are covered. Disk 3 has a more than three times higher roughness than disk 1. Thus, the changes in blackening are clearer to be seen. Since autoradiography only shows the traces of the iron particles, chemical analysis to find nitrogen was added in the next step. 3.4. Chemical analysis Photoelectron spectroscopy depth profiles were taken inside and outside of the wear track. Fig. 7 shows a comparison of the element distributions down to a depth of 100 nm. In addition to nitrogen the concentrations of iron, oxygen and carbon (C/CHx)
Fig. 7. XPS depth profiles.
were recorded showing increased values inside the wear track. Considering the nitrogen concentration as indicator of the film thickness, the polymer build-up has a thickness of about 160 nm, see Fig. 8. The depth scale was calibrated using a SiO2 normal. That means that minor deviations can occur with respect to the PA46 film thickness. Due to limits in resolution, concentrations below 0.5% have to be treated as noise. Therefore, the nitrogen concentration outside the wear track was not considered for a sputter depth larger than about 10 nm. We assume that the nitrogen contents up to a depth of 10 nm originates from the intake of air during sample preparation. The depth profiles raise the question if carbon and nitrogen were intermixed with the iron matrix. However, as mentioned in Section 3.1, the roughness of the steel disks did not change during the experiment, suggesting that only the tops of the asperities interacted with the polymer and became covered with the transfer film. The XPS ‘‘sees’’ both
762
M. Scherge et al. / Wear 301 (2013) 758–762
process is not established, the tribological system remains in abrasion. After deposition of a certain amount of polymer onto the steel surface, friction drops. This process is retarded when a rough steel surface abrades the polymer pin. Then the equilibrium between polymer deposition and removal adjusts at large wear rate and high friction. The effect of material transfer has its biggest impact in the boundary lubrication regime, since the deposited polymer reduces the shear resistance of the system. When the tribological system enters hydrodynamic lubrication, transfer effects should loose significance.
5. Conclusions
Fig. 8. Atomic concentration of nitrogen in the wear track and outside of the wear track.
peaks and valleys and returns a combined signal of iron and carbon. Due to sputtering the transfer layer is thinned and the carbon signal decreases.
4. Discussion In this section the running-in behavior will be discussed. In the presented experiments the most probable causes for a running-in can be the temporal decrease in roughness leading to hydrodynamics or the build-up of a polymer transfer film. When roughness decreases as function of runtime, the real area of contact increases and the contact pressure decreases. Starting with boundary lubrication, friction should decrease until hydrodynamics kicks in. Parallel to decreasing friction, the wear rate should decrease until it becomes zero due to separation of both solids by the formed oil film. However, Fig. 4 shows that at the end of the experiment wear is still present. Even the lowest wear rate of 50 nm per hour indicates that there must be a direct mechanical interaction between polymer and steel. Thus, the gradual transition into hydrodynamic lubrication rules out as cause for running-in. The second possible path for the running-in is the build-up of the transfer film. As shown in the previous sections both radiography and XPS depth profiles suggested that a polymer layer is formed on top of the asperities. The material transfer seems to be connected to the roughness of the disk. More wear particles were measured by the radionuclide technique for the rougher disk, which also correlates to higher coefficients of friction. In addition, low roughness correlates with low total wear, small wear rate and small friction. This means that as long as the material transfer
Continuous friction and wear measurement revealed that the tribological system of PA 46 and steel shows a pronounced running-in behavior. By choosing an appropriate roughness of the steel sample, i.e., Ra o0.1 mm, low friction and small wear rates can be achieved. The necessary pre-condition is the formation of a stable transfer film changing the tribological properties of the steel surface. Under these conditions PA 46 is a promising candidate for highly-stressed mechanical applications such as journal bearings in combustion engines.
References [1] Z. Zhang, Q. Xue, W. Liu, W. Shen, Friction and wear behaviors of several polymers under oil-lubricated conditions, Journal of Applied Polymer Science 68 (1998) 2175–2182. [2] M. Palabiyik, S. Bahadur, Tribological studies of polyamide 6 and high-density polyethylene blends filled with PTFE and copper oxide and reinforced with short glass fibers, Wear 253 (2002) 369–376. [3] L. Bottenbruch, R. Binsack, [Hrsg.]: Polyamide 3/4, VDI-Book, Hanser, 1998. [4] P. Elsner, P. Eyerer, T. Hirth, Polymer Engineering, VDI-Book, Springer, Berlin, 2008. [5] W. Michaeli, A. Elas, B. Rothe, Polyamid 6-schichtsilikat-Nanocompound hergestellt durch In-situ polymerisation, NanoTechnologie 1 (2008) 44–48. [6] Y. Liang, X. Xia, Synthesis and performances of Fe2O3/PA-6 nanocomposite fiber, Materials Letters 61 (14–15) (2007) 3269–3272. [7] P. Cong, F. Xiang, X. Liu, T. Li, Morphology and microstructure of polyamide 46 wear debris and transfer film: in relation to wear mechanisms, Wear 265 (2008) 1100–1105. ¨ [8] T. Kreer, M.H. Muser, K. Binder, Frictional drag between polymer bearing surfaces, Computer Physics Communications 147 (2002) 358–361. [9] K. Kato, E. Uchida, E. Kang, Y. Uyama, Y. Ikada, Polymer surface with graft chains, Progress in Polymer Science 28 (2003) 209–259. [10] S.V. Kotomin, N.N. Avdeev, Macromodel approach to experimental study of polymer brushes rubbing, Wear 222 (1998) 21–28. [11] V. Quaglini, P. Dubini, D. Ferroni, C. Poggi, Influence of counterface roughness on friction properties of engineering plastics for bearing applications, Materials and Design 30 (2009) 1650–1658. ¨ [12] M. Scherge, K. Pohlmann, A. Gerve´, Wear measurement using radionuclidetechnique (RNT), Wear 254 (9) (2003) 801–817. [13] F. Ditro´i, S. Taka´cs, F. Ta´rka´nyi, M. Reichel, M. Scherge, A. Gerve´, Thin layer activation of large areas for wear study, Wear 261 (11–12) (2006) 1397–1401.
Contents lists available at SciVerse ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Running-in due to material transfer of lubricated steel/PA46 (aliphatic polyamide) contacts ¨ Matthias Scherge n, Jeanette Kramlich, Roman Bottcher, Tobias Hoppe 1 Fraunhofer IWM, Microtribology Center, W¨ ohlerstraße 11, 79108 Freiburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 August 2012 Received in revised form 5 November 2012 Accepted 9 November 2012 Available online 29 November 2012
The running-in of an aliphatic polyamide 46 (PA 46) pin versus a lubricated steel disk was studied by friction and real-time wear measurements with a pin-on-disk tribometer. It was shown that under certain boundary conditions a transfer film forms and significantly reduces friction and wear. In addition, total wear, wear rate, running-in time and coefficient of friction strongly depend on the roughness of the steel disk. In all experiments the system showed a pronounced running-in behavior, characterized by decreasing and, after a certain time, constant coefficient of friction and wear rate. When the roughness of the steel sample was high, abrasion and high friction predominated. For disks with small roughness material transfer became effective and friction and wear decreased accordingly. Material transfer was proven by radioactive labeling of the polymer and detecting its traces on the steel surface. In addition, photoelectron spectroscopy was applied to find fractions of the PA 46 amine groups in the wear track. & 2012 Elsevier B.V. All rights reserved.
Keywords: Polymer Steel Friction Wear
1. Introduction Polymers, particularly polyamides, are increasingly used in tribological applications, e.g. journal bearings. Under oil lubrication very low coefficients of friction can be achieved, e.g. 0.025–0.035 for PA 66 or 0.015–0.045 for PTFE (load 100 N, sliding velocity 1.0–2.5 m/s). However, apart from a few exceptions, the wear resistance is still insufficient [1,2]. Furthermore, the running-in behavior of most polymers in contact with a steel counter piece is unknown, thus lifetime as well as performance predictions are complicated. Most technically important polymers are partially crystalline thermoplastics, which are characterized by a high toughness, rigidity and tenacity [3]. For tribological applications aliphatic polyamides such as PA 46, PA 6 and PA 66 widely used materials. Especially the chemical structure of PA 46 offers an advantage compared to PA 6 and PA 66 [4]. Polyamides are characterized by recurring carbon amide groups in the main chain, which are responsible for the physical differences in contrast to most other polymers. The carbon amide groups with their hydrogen bonds are the reason for outstanding mechanical properties and good chemical resistance. The extremely short hydrogen bonds between the NH-group and the CO-group are responsible for the high melting point of PA 46 of 295 1C [3]. Moreover, the high degree n
Corresponding author. Tel.: þ49 721 4640 750; fax: þ49 721 4640 111. E-mail address: [email protected] (M. Scherge). 1 Current address: Robert Bosch GmbH, (GS-AM/PJ-MIC), Postfach 30 02 40, 70442 Stuttgart, Germany. 0043-1648/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.wear.2012.11.035
of crystallization provides an improved pinning of the molecules and thereby improved resistance against abrasion [5]. The mechanical performance, as well as the friction and wear behavior, can be adjusted by fillers interacting with the polyamide matrix [3]. These fillers are able to further improve the chemical and physical properties of polyamide. Even with a very low filling degree of less than 5%, nanoparticle fillers are responsible for significant changes of the composite behavior [5]. For instance, metallic powders enhance the heat resistance of polymers [4]. However, these changes come into effect only if the filler is dispersed homogeneously [6]. Agglomerated fillers with dimensions larger than 1 mm can worsen the properties of the polymer. The tribological properties of polymers sliding against steel without lubrication are mainly influenced by the low shear strength of the polymer. It is widely accepted that polymer transfer to the surface of the counter body is responsible for low friction and small wear [7]. This means that the shear plane is shifted from the steel surface into the polymer layer on top of the steel [8]. Material transfer calls for adhesion to attach fractions of the polymer to the steel surface. To understand this mechanism, results from experiments with polymers grafted to a surface can be used [9,10]. It is assumed that polar end-groups of the polymer form physical or chemical bonds with the surface. Under equilibrium conditions a polymer brush is formed which reduces the shear stress upon sliding. Usually the monomers to be grafted are dissolved in a liquid or originate from the gas phase during glow discharge. To initiate such a process in a pin-on-disk experiment a sufficient amount of energy is necessary to remove the water film
M. Scherge et al. / Wear 301 (2013) 758–762
present on every metal surface and to form a polymer film. The same applies to a lubricated steel surface, where layers of oil have to be removed prior to the attachment of polymer chains to its surface. The objective of this investigation was to understand the running-in behavior. Since we suspected that the running-in results from balancing abrasion and transfer, steel disks with different roughnesses were prepared [11]. The running-in behavior was monitored by continuous wear measurement applying the radionuclide technique (RNT) with a pin-on-disk tribometer. Autoradiography was used to detect transferred tracer atoms coming from the polymer on the steel surface. In addition chemical analysis was carried out to find traces of amine groups at the steel surface.
2. Experimental procedure 2.1. Sample preparation In order to apply continuous wear measurement using the radionuclide technique, PA 46 was filled with 9% Fe2O3-nanoparticles during extrusion to serve as tracer after weak radioactive irradiation, see Section 2.2. The nanoparticles were added to the extruder in liquid form. To prevent coagulation, the nanoparticles were functionalized before immersion into the liquid. Functionalization is a process by which a chemical species can be made to react or not to react in a certain way by activating certain sites on a surface or a molecule. Anti-coagulation in this case was done by adding a suitable non-polar group. After extrusion the received pellets were molded into a bar as base material for the pins and for accompanying analysis such as the determination of tensile strength, Young’s modulus and the measurement of the glass temperature. The bars have a tensile strength of 73.65 MPa, a Young’s modulus of 2.98 GPa and a glass temperature of 40–50 1C. PA tablets with a diameter of 5 mm and a thickness of 4 mm were machined without visibly changing the original surface. All tablets were conditioned under ambient temperature and humidity. Polyamide is known to incorporate water into its amorphous regions. This is particularly the case for PA 46 with a nominal water consumption of about 12%. The water interacts with the microstructure and lowers the glass transition temperature [3]. This behavior influences both friction and wear. To analyze the arrangement of the nanoparticles in the polymer matrix, scanning electron microscopy (SEM) was used. Fig. 1 shows an SEM image of PA 46 with 9% nanoparticles, indicating
759
that the particles (white spots) are not agglomerated and distributed uniformly. The disks were made of steel (100Cr6) with a diameter of 200 mm and a thickness of 10 mm. The topography of the steel disks was achieved through a process of super-finish using a band tool (Supfina). By varying finishing times and grain size of the abrasives, disks with different average roughness were received (disk 1—0.08 mm; disk 2—0.14 mm; disk 3—0.27 mm). Roughness was determined with a confocal microscope (Sensofar). 2.2. Instrumentation Continuous friction and wear measurements were carried out with a pin-on-disk tribometer (Basalt SST, Tetra GmbH). The tribometer covered a range between 0 and 1000 rpm. Depending on the distance between pin and the center of the disk, sliding velocities up to 10 m/s were possible. The maximum normal force that can be applied to the pin was 1000 N. Both normal and tangential forces were recorded throughout the experiment. To measure wear, the oil circuit was connected to an RNT device (Zyklotron AG), see Fig. 2. To obtain a pin, a PA tablet was attached to a holder allowing a self-adjusting contact with the steel disk. Particles developed by wear were removed from the disk by the oil flow and were carried to the RNT device via oil circuit. Since the polymer was homogenously filled with Fe2O3-nanoparticles (see Fig. 1), the activation of a fraction of the iron atoms with thermal neutrons ensured a homogeneously labeled sample. In contrast to conventional cyclotron activation utilizing a focused beam of deuterons to generate nuclides, the activation in a stream of thermal neutron was a gentle way to obtain nuclides without weakening the polymer matrix. The radiation process delivered Fe-59 nuclides. The more nuclides arrive at the detector, the more wear is present. Through calibration, activity was related to the wear mass. One of the benefits of the RNT is its high resolution of a few _ micrograms per liter oil. From the wear mass mw the wear rate w was calculated by dividing the wear depth dw by the duration of the experiment t. _ ¼ w
dw mw ¼ t Art
ð1Þ
A is the contact area of the pin and r is the density of PA 46. Detailed information about the RNT method can be found in [12,13]. 2.3. Lubricant For the experiments two types of oil with similar viscosity were used to obtain optimum results. For wear analysis fullyformulated oils are preferable since their additivation suppresses sedimentation of wear particles. Therefore all wear tests were
Fig. 1. SEM micrograph of PA 46 with 9% nanoparticles.
Fig. 2. Pin-on-disk tribometer connected to a radionuclide wear measurement system.
M. Scherge et al. / Wear 301 (2013) 758–762
2.4. Test procedure To become familiar with the tribological properties of the system, first Stribeck curves were recorded to obtain the coefficient of friction (COF) as function of sliding velocity times viscosity divided by the contact pressure. After passing a parameter field with different normal forces and sliding velocities, areas of boundary and hydrodynamic lubrication were identified. All sample pairings were then subjected to constant normal force and sliding velocity to investigate the running-in behavior in the boundary lubrication regime. All recorded friction data were averaged over a period of 10 s. The oil temperature was set to 60 1C. In addition to the parameter fields, tests with the nitrogen-free reference oil were performed to analyze the steel disks with respect to the anticipated polymer transfer. Since the oil did not contain a significant percentage of nitrogen, nitrogen in the wear track would indicate the deposition of amine groups coming from the polymer.
3. Results 3.1. Friction and wear tests Fig. 3 shows the friction coefficients for disk 1, 2 and 3 as function of sliding velocity divided by the contact pressure. Viscosity was treated as constant and omitted. It was shown that the largest differences between the friction curves appeared in the boundary lubrication regime. Under hydrodynamic lubrication (v/p 40.6) all curves behave similarly. Steel disks with larger roughness show more pronounced boundary lubrication due to increased local stresses. Based on these findings the coefficient of friction and the wear _ were analyzed for a constant normal force of 300 N rate w
0.10 0.09 0.08 COF Ra = 0.27µm
0.07
COF Ra = 0.14 µm
COF
0.06
COF Ra = 0.08 µm
0.05 0.04
disk 1 2 3
0.12 0.11 0.10 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00
COF 0.02 0.04 0.09
dw/dt [µm/h] 0.04 0.6 6
10
400
8
3
4 2
1
2 0 0
2
4
6
8
10
12
14
200
Wear [µm]
300 6
Wear [mg]
performed with FUCHS TITAN GT1 5W30 LongLife III (ACEA A3). To analyze the transfer process, however, a nitrogen-free oil (Fuchs PAO) was used. This oil contains mainly tridecane, undecane and heptadecane. The contents of nitrogen is negligibly small (0.83 wt% with a standard deviation of 0.38 wt%) and originates presumably from the intake of air during chemical analysis (Thermo Flash EA measurement). Thus, traces of nitrogen in the wear track should originate from the PA 46 amine groups.
COF
760
100
0
16
Time [h] Fig. 4. Friction and wear as function of time. The dotted curves represent the wear behavior. The error of the coefficient of friction of experiment and repetition is 0.01. The error of wear is smaller than the size of the symbol.
(8 15 MPa) and a sliding velocity of 1 m/s. This corresponds to a v/p ratio of 0.07 which clearly leads to boundary lubrication. Fig. 4 shows the coefficient of friction and the total wear as function of time and roughness of the steel disks. All tested pairings showed a distinct running-in behavior similar to lubricated metal–metal contacts. In addition to the pronounced running-in behavior friction and wear show a clear dependence on the roughness of the steel disk. For disk 3 (highest roughness) the level of total wear is about a hundred times higher than for metal–metal contacts operated under similar stressing conditions. With decreasing roughness both friction, total wear and wear rate decrease. All initial coefficients of friction start at values of about 0.1. Shortly after the start of experiment the coefficients of friction significantly drop, which can be attributed to the topographical running-in. The disk with the lowest roughness achieves a wear rate of 0.04 mm/h and a coefficient of friction of 0.02 at the end of the experiment. All other wear rates and friction data are shown as inset in Fig. 4. 3.2. Topography of PA and steel The topography of the PA samples was examined before and after tribological testing. The initial topography of the polymer surface, measured with a confocal microscope, is determined by the roughness of the mold, see Fig. 5. After the tests the surfaces show slightly lower roughness, however, scattered grooves which determine the peak to valley roughness are still visible. With increasing roughness of the steel disk the number of asperities on the PA surface increases. Since PA 46 is rather soft, the tribologically induced pin roughness was attained shortly after the start of the experiment. A start-stop test showed that already after 5 min roughness had adapted to the new conditions. For the steel disks no change in roughness was measured. The roughness profiles were measured perpendicular to the friction tracks. 3.3. Autoradiography
0.03 0.02 0.01 0.00 0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
v/p [m/(MPa*s)] Fig. 3. Stribeck curves. The error of the coefficient of friction of experiment and repetition is 0.01.
The assignment of this technique is aimed at the detection of radioactive traces at the surface of the steel disk. By means of a gamma spectrometer with an array of highly-sensitive germanium detectors even the very low Fe-59 activity was detected by adjusting the exposure time. Fig. 6 shows on the left hand side the trace of nuclides on disk 1 (low roughness). On the right hand side disk 3 (high roughness) is shown. On both disks traces of the pin material were detected. Although no quantitative characterization
M. Scherge et al. / Wear 301 (2013) 758–762
761
Fig. 5. Topography of PA 46 before (top) and after (bottom) tribological stressing.
Fig. 6. Autoradiography measurements of the disks. Left—disk 1. Right—disk 3. The width of the wear track is 5 mm.
was possible, the blackening of the traces gives an estimate of the amount of the material transfer. Fig. 6 shows that the transferred material does not cover the steel surface uniformly. We assume that only the tops of the asperities are covered. Disk 3 has a more than three times higher roughness than disk 1. Thus, the changes in blackening are clearer to be seen. Since autoradiography only shows the traces of the iron particles, chemical analysis to find nitrogen was added in the next step. 3.4. Chemical analysis Photoelectron spectroscopy depth profiles were taken inside and outside of the wear track. Fig. 7 shows a comparison of the element distributions down to a depth of 100 nm. In addition to nitrogen the concentrations of iron, oxygen and carbon (C/CHx)
Fig. 7. XPS depth profiles.
were recorded showing increased values inside the wear track. Considering the nitrogen concentration as indicator of the film thickness, the polymer build-up has a thickness of about 160 nm, see Fig. 8. The depth scale was calibrated using a SiO2 normal. That means that minor deviations can occur with respect to the PA46 film thickness. Due to limits in resolution, concentrations below 0.5% have to be treated as noise. Therefore, the nitrogen concentration outside the wear track was not considered for a sputter depth larger than about 10 nm. We assume that the nitrogen contents up to a depth of 10 nm originates from the intake of air during sample preparation. The depth profiles raise the question if carbon and nitrogen were intermixed with the iron matrix. However, as mentioned in Section 3.1, the roughness of the steel disks did not change during the experiment, suggesting that only the tops of the asperities interacted with the polymer and became covered with the transfer film. The XPS ‘‘sees’’ both
762
M. Scherge et al. / Wear 301 (2013) 758–762
process is not established, the tribological system remains in abrasion. After deposition of a certain amount of polymer onto the steel surface, friction drops. This process is retarded when a rough steel surface abrades the polymer pin. Then the equilibrium between polymer deposition and removal adjusts at large wear rate and high friction. The effect of material transfer has its biggest impact in the boundary lubrication regime, since the deposited polymer reduces the shear resistance of the system. When the tribological system enters hydrodynamic lubrication, transfer effects should loose significance.
5. Conclusions
Fig. 8. Atomic concentration of nitrogen in the wear track and outside of the wear track.
peaks and valleys and returns a combined signal of iron and carbon. Due to sputtering the transfer layer is thinned and the carbon signal decreases.
4. Discussion In this section the running-in behavior will be discussed. In the presented experiments the most probable causes for a running-in can be the temporal decrease in roughness leading to hydrodynamics or the build-up of a polymer transfer film. When roughness decreases as function of runtime, the real area of contact increases and the contact pressure decreases. Starting with boundary lubrication, friction should decrease until hydrodynamics kicks in. Parallel to decreasing friction, the wear rate should decrease until it becomes zero due to separation of both solids by the formed oil film. However, Fig. 4 shows that at the end of the experiment wear is still present. Even the lowest wear rate of 50 nm per hour indicates that there must be a direct mechanical interaction between polymer and steel. Thus, the gradual transition into hydrodynamic lubrication rules out as cause for running-in. The second possible path for the running-in is the build-up of the transfer film. As shown in the previous sections both radiography and XPS depth profiles suggested that a polymer layer is formed on top of the asperities. The material transfer seems to be connected to the roughness of the disk. More wear particles were measured by the radionuclide technique for the rougher disk, which also correlates to higher coefficients of friction. In addition, low roughness correlates with low total wear, small wear rate and small friction. This means that as long as the material transfer
Continuous friction and wear measurement revealed that the tribological system of PA 46 and steel shows a pronounced running-in behavior. By choosing an appropriate roughness of the steel sample, i.e., Ra o0.1 mm, low friction and small wear rates can be achieved. The necessary pre-condition is the formation of a stable transfer film changing the tribological properties of the steel surface. Under these conditions PA 46 is a promising candidate for highly-stressed mechanical applications such as journal bearings in combustion engines.
References [1] Z. Zhang, Q. Xue, W. Liu, W. Shen, Friction and wear behaviors of several polymers under oil-lubricated conditions, Journal of Applied Polymer Science 68 (1998) 2175–2182. [2] M. Palabiyik, S. Bahadur, Tribological studies of polyamide 6 and high-density polyethylene blends filled with PTFE and copper oxide and reinforced with short glass fibers, Wear 253 (2002) 369–376. [3] L. Bottenbruch, R. Binsack, [Hrsg.]: Polyamide 3/4, VDI-Book, Hanser, 1998. [4] P. Elsner, P. Eyerer, T. Hirth, Polymer Engineering, VDI-Book, Springer, Berlin, 2008. [5] W. Michaeli, A. Elas, B. Rothe, Polyamid 6-schichtsilikat-Nanocompound hergestellt durch In-situ polymerisation, NanoTechnologie 1 (2008) 44–48. [6] Y. Liang, X. Xia, Synthesis and performances of Fe2O3/PA-6 nanocomposite fiber, Materials Letters 61 (14–15) (2007) 3269–3272. [7] P. Cong, F. Xiang, X. Liu, T. Li, Morphology and microstructure of polyamide 46 wear debris and transfer film: in relation to wear mechanisms, Wear 265 (2008) 1100–1105. ¨ [8] T. Kreer, M.H. Muser, K. Binder, Frictional drag between polymer bearing surfaces, Computer Physics Communications 147 (2002) 358–361. [9] K. Kato, E. Uchida, E. Kang, Y. Uyama, Y. Ikada, Polymer surface with graft chains, Progress in Polymer Science 28 (2003) 209–259. [10] S.V. Kotomin, N.N. Avdeev, Macromodel approach to experimental study of polymer brushes rubbing, Wear 222 (1998) 21–28. [11] V. Quaglini, P. Dubini, D. Ferroni, C. Poggi, Influence of counterface roughness on friction properties of engineering plastics for bearing applications, Materials and Design 30 (2009) 1650–1658. ¨ [12] M. Scherge, K. Pohlmann, A. Gerve´, Wear measurement using radionuclidetechnique (RNT), Wear 254 (9) (2003) 801–817. [13] F. Ditro´i, S. Taka´cs, F. Ta´rka´nyi, M. Reichel, M. Scherge, A. Gerve´, Thin layer activation of large areas for wear study, Wear 261 (11–12) (2006) 1397–1401.