tem and thus pushes the tribosystem further towards mixed and boundary lubrication, thereby increasing the propensity towards wear. When using hybrid systems, which use the electrical motor for parts of the vehicle operation, the engine experiences more engine starts and stops than in conventional vehicles. Start-stop operation results in increased amounts of time the engine tribosystems spend in lowvelocity operation ranges while the engine is still hot (i.e. low oil viscosity) [2], which translates to more operation in boundary or mixed lubrication regimes in systems that normally operate nearly exclusively in the hydrodynamic range. In summary, these modern trends in engine development can lead to more wear, due to more time spent in mixed Aus der Praxis für die Praxis 42 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 1 Introduction Modern engine development is focused on environmental optimization, which is mainly realized by lowering fuel consumption through reduction of friction losses or using hybrid systems, where an internal combustion engine is supplemented by an electrical motor. The reduction of friction losses can be achieved, for example, by lowering the viscosity of the engine oil [1]. However, this inherently reduces the Stribeck parameter of the sys- DOI 10.30419/ TuS-2019-0005 Evolution of wear rate, friction and roughness depending on lubrication regime Claudia Lenauer, Oliver Knaus, Thomas Wopelka, Martin Jech* Neue Fahrzeugstrategien, wie häufige Start-Stopp Vorgänge, sowie Optimierungen bezüglich Motorleistung und -effizienz, führen zu steigenden tribologischen Belastungen indem zum Beispiel Lager längere Betriebszeiten im Mischreibungsbereich oder gar Grenzreibungsbereich verbringen. Um die Auswirkung des Reibregimes auf Lagerverschleiß grundsätzlich zu untersuchen, wurde eine Serie von Versuchen durchgeführt, bei welchen die Versuchsparameter derart ausgesucht wurden, dass sie (so gut wie experimentell machbar) in Grenz- und Mischreibung, sowie Hydrodynamik liegen. Reibung und Verschleiß wurden kontinuierlich während der Versuche gemessen, während die Änderungen der Topographie im Grenzreibungsbereich mit einer Serie kürzerer Versuche erfasst wurden. Die Zeit bis steady-state erreicht wurde war stark abhängig von den Belastungsparametern und zeigt einen ähnlichen Trend für Reibung und Verschleiß. Die Einlaufzeit bezüglich Oberflächenbeschaffenheit war dagegen deutlich kürzer. Schlüsselwörter Lagerschalen, Reibregime, Einlaufverschleiß, radioaktive Tracer, Rauheit, Mischreibung New engine strategies, such as frequent start-stop operation and increasingly severe loading conditions due to efforts to optimize engine performance and efficiency, lengthen the time the bearings spend in mixed or even boundary lubrication regimes. To investigate how the lubrication regime affects bearing bush wear, a series of tests were performed at distinct positions on the Stribeck curve, representing, as closely as possible, the boundary, mixed and hydrodynamic lubrication regimes. Both friction and wear were measured throughout the test, while the evolution of surface roughness was determined for the boundary condition in a series of shorter tests. The absolute amount of wear, as well as the running-in time needed to reach steady-state behaviour, depends strongly on the preset lubrication regimes with the running-in time showing a similar trend for friction and wear. The running-in time for the surface roughness is considerably shorter than for either friction or wear. Keywords Bearing bushes, lubrication regime, running-in wear, radioactive tracers, mixed lubrication Kurzfassung Abstract * Dipl.-Ing. Claudia Lenauer, Orcid-ID: https: / / orcid.org/ 0000-0001-9917-8658 Mag. Dr. Thomas Wopelka, Dipl.Ing. Dr. Martin Jech, Orcid-ID: https: / / orcid.org/ 0000-0002-9565-5792 AC 2 T research GmbH, 2700 Wiener Neustadt, Austria Dipl.-Ing. Oliver Knaus AVL List GmbH, 8020 Graz, Austria T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 42 or boundary lubrication regimes. Therefore, the amount of wear of a tribosystem depending on the lubrication conditions needs to be investigated, including transient effects such as running-in [3]. Blau [4] writes that running-in is a result of many simultaneous processes occurring in the contact, including surface roughness alterations, changes in surface composition and microstructure. Running-in of friction and wear are therefore very complex phenomena that are not necessarily simultaneous and can manifest very differently depending on the contact geometry, loading conditions, materials and lubrication. Therefore, friction, wear and surface topography - including their evolution - are investigated experimentally for a specific system - bearing bushes - in order to throw some light on the tribological processes. The system was simplified as far as possible to reduce the number of (still complex) phenomena by choosing a system with unidirectional movement, easily measurable surface structure, constant loading conditions (the only changes are those due to wear), and a non-additivated lubricant (to eliminate complex tribochemistry and tribofilm formation). 2 Methods 2.1 Samples CuZn25Al5 bearing bushes were tested against 100Cr6 counterparts. The bearing bushes have an inner diameter of 30 mm, a length of 20 mm and were cut in half in axial direction to provide two samples per bush. The bearing shaft counterparts have a diameter of 29.5 mm and, additionally, a convexity in axial direction of 100 µm. The convexity ensures that the contact (and wear) zone is located in the centre of the bearing bush and edge-running effects are eliminated. NEXBASE® 3043 (see Table 1 for specifications), a synthetic base oil without additives, was selected to lubricate the contact, to minimize effects attributable to lubricant chemistry. The low viscosity at test temperature (20-23 °C) moves the transition point from the mixed to the hydrodynamic lubrication regime towards higher speeds (compared to higher viscosity engine oils). This opens up a wider range of possible operational parameters during which the system is in the mixed lubrication regime. 2.2 Tribometer Set-up Tribological testing was carried out on a modified precision bearing test rig (SLPG - Sinterlager Prüfgerät [5]) with unidirectional sliding. The bearing shaft is rotated, while the bearing bush is mounted in a sample holder connected to the torque measurement, which is used to determine the friction moment. The load is applied to the bearing bush via a load arm. The lu- Aus der Praxis für die Praxis 43 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 DOI 10.30419/ TuS-2019-0005 Viscosity at 100°C [mm²/ s] 4.3 Viscosity Index 122 Viscosity at 40°C [mm²/ s] 20 Density at 15°C [g/ cm³] 837 Viscosity at 20°C - 23°C [mm²/ s] * 46 - 40 Table 1: Technical Data for NEXBASE 3043; * Viscosity at test temperature calculated with the Ubbelohde-Walther equation. Figure 1: Stribeck curve for 100 and 150 N: full curve (top) with blue shaded box marking excerpt which is used to show the location of chosen loading conditions shown as Starting Conditions 1-3 (bottom) T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 43 running-in, additional shorter tests were performed for the most severe loading condition (150 N, 75 rpm). Three tests each were run for 10 and 30 minutes. 2.3 Surface Topography Analysis The surface topography of the bearing bush samples were investigated before and after testing using a confocal microscope. An area of 1.2 mm x 1.2 mm in the centre of the wear zone was examined. Due to the production process the surface shows grooves oriented in circumferential direction of the bushes - along the sliding direction. Therefore, linear roughness parameters were determined in axial direction of the bushes. Roughness parameters in circumferential direction were below 100 nm (around the resolution limit of the lens) and statistically significant differences before and after testing could not be determined. 2.4 Phenomenological Models The wear curves measured during the experiments showed a pronounced running-in, followed by a steady-state phase. Therefore the wear curves were fitted with an exponential function (similar to Barwell [8] and Kumar [9]) (1) which has been shown to be a good model for this kind of transient wear behaviour and results in a wear rate of (2) The last term represents the constant wear rate c w , while the ‘intensity’ of the running-in wear is represented by a w (i.e. the higher the value of a w , the more pronounced the running-in). The parameter b w describes the duration of running-in, the inverse of which is the time constant τ w of the exponential wear curve. At t = 3 τ w , the amount of wear has reached 95 % of a w . A fit function for the coefficient of friction, similar to the wear rate, was chosen to be able to compare the runningin behaviour of the wear and friction curves. In Eq. 3 the friction coefficient µ(t) is described by three fit parameters a f , b f and c f : (3) Again, the parameter b f is the inverse of the time constant τ f of the friction coefficient and the friction coefficient is within 95 % of its steady-state value after t = 3 τ f ,. Hanief et al. [10] use a mathematically similar model for the evolution of roughness during ! "# $ % & ! 1 ' exp! '( & "## ) * & " , +& +- $ %. & exp! '( & "# ) * & . / ! "# $ % 0 exp2'( 0 "3 ) * 0 Aus der Praxis für die Praxis 44 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 bricant is provided continuously from above the contact. Since the shaft centre and attitude can shift due to wear [2], the offset in the friction torque caused by this shift was measured for each experiment and the friction coefficient corrected accordingly. Wear of the bearing bushes was measured continuously throughout the test using the RIC (radio-isotope concentration) method [6,7]. An area in the centre of the CuZn25Al5 bearing bushes, where the tribological contact is anticipated, was activated using the thin layer activation technique [6], which results in the radioactive isotope Zn65 being present in a known concentration in this area. During the tests, wear particles originating from the activated area contain Zn65 and are transported to the wear measurement device by the lubricant circuit. Several pre-tests with varying loads (30-200 N) and speed ramps (0-2000 rpm, which equals 0-3.14 m/ s for this geometry, at 4 minutes per ramp) were carried out to determine the optimal normal loads for the main test matrix. Two loads that simultaneously produce measurable friction and wear, and are low enough to produce only a few large wear particles (and thus are not yet in the abrasive wear regime) were chosen: 100 and 150 N, which results in initial Hertzian contact stresses of 58 and 66 MPa, respectively. For each of these two loads, a 24 h test with ramped speeds (0-2000 rpm, 4 min per ramp) was performed to determine the Stribeck curve for these test conditions. The Stribeck curves changed during the 24 hrs due to initial running-in phase but stabilized throughout the test. The Stribeck curves at the end of the tests (to minimize possible differences due to running-in) were then used to determine the constant speeds used in the main test matrix. Three positions on the Stribeck curve were chosen as starting conditions (SC) for the main tests: one as close as possible to the boundary lubrication regime (SC1), one near the transition between the mixed (SC3) and the hydrodynamic regime and the third between the others (SC2) representing the mixed lubrication regime. The velocities for both loads was chosen so that for each starting condition, the Stribeck parameters have the same value for both loads (Table 2). Three tests with test durations of 120 min. were performed for each starting condition and the test schedule was randomized. To investigate the roughness changes due to DOI 10.30419/ TuS-2019-0005 Stribeck parameter: Starting Condition 1 2.38e-05 Starting Condition 2 4.29e-05 Starting Condition 3 6.67e-05 100 N 50 rpm 90 rpm 140 rpm 150 N 75 rpm 135 rpm 210 rpm Table 2: Main test matrix, showing chosen speeds for each load T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 44 running-in (notation and order of terms adjusted to the conventions in the present article), where the surface roughness follows the same kind of exponential decay as the wear rate and the friction coefficient: (4) 3 Results 3.1 Wear Wear of the bearing bushes was measured using the RIC method; Figure 2 shows the wear curves for all 2 h-tests. The tests with a load of 100 N are shown on the left and those with 150 N on the right. The wear results from one test (SC 2, 150N) had to be excluded, due to misalignment of the samples leading to a contact area/ worn area lying far off-centre. The lowest amount of wear was observed for SC 3 (top graphs), while the more severe loading conditions show successively more wear. Due to the variabilities in tribological testing, there is a con- 4! "# $ % 5 exp! '( 5 "# ) * 5 . siderable amount of scattering, especially for the lower load and lower speeds. However, the trend is clear in that the total amount of wear is considerably higher for tests run at lower speeds (i.e. more severe conditions) - for both 100 N and 150 N load. SC 3, which is closest to the hydrodynamic regime, shows no discernible increase in wear after a short initial running-in phase. Running-in is higher for the conditions closer to boundary lubrication. Another noticeable difference between the wear curves for lower and higher speeds is the amount of time needed to reach the steadystate part of the wear curve. The tests closer to the hydrodynamic regime have a shorter run-in. 3.2 Friction The changes of the friction coefficient over the duration of the tests is shown in Figure 3 for 100 N and 150 N. For SC 3 (the highest speeds), the friction coefficient stabilizes at a low value very quickly. While similarly low values are reached for the starting conditions with Aus der Praxis für die Praxis 45 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 DOI 10.30419/ TuS-2019-0005 Figure 2: Wear for 100N tests (left) and 150 N (right); top (blue): SC 3; middle (green): SC 2; bottom (red): SC 1 T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 45 sures that the roughness parameters represent the surface topography of the majority of the worn area without being skewed by any deeper scratches. The 3D image of the initial surface of the bearing bushes and the related profile in axial direction is shown in (Figure 4, top left). The sample surface had grooves in circumferential direction with plateau regions in between. During the friction process these plateau regions will be part of the contact area, the bottom of the grooves will not be in contact with the surface of the counterpart until the structure is worn to a high extent. Depending on the starting condition, the plateaus of this surface structure are worn to a greater or lesser extent during the sliding process (Figure 4, centre, right). It is difficult to find a single roughness parameter that can describe the changes in structure and roughness observed for these surfaces. Several profile roughness and surface parameters were considered and analysed, but proved to be no more informative than the amplitude parameter Ra in axial direction. In order to aid comparisons Aus der Praxis für die Praxis 46 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 lower speeds, the transitions occur later the lower the speed. It can also be seen that the scattering within each friction coefficient curve is lower for higher speeds. The spikes that can be seen in some of the curves are attributed to wear particles passing through the contact. 3.3 Topography Visual inspection of the worn surfaces after tribological testing revealed a small number of scratches indicative of abrasive wear particles on the majority of samples. Scratches, contrary to the grooves of the initial surface structure, are easily visible to the naked eye. Generally, there were fewer (or even none) scratches for SC 3 and slightly more for SC 1. The surfaces with a higher number of scratches and/ or particularly pronounced scratches were those where the friction coefficient curves exhibited spikes and/ or high amount of scattering. Surface topography measurements on post-test samples were performed in the worn area, while avoiding scratches. Avoidance of scratches - by deliberately taking images of surface areas with no/ low amount of scratches - en- DOI 10.30419/ TuS-2019-0005 Figure 3: Coefficient of friction for 100N tests (left) and 150 N (right); top (blue): SC 3; middle (green): SC 2; bottom (red): SC 1 T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 46 with literature, where Ra is generally widely used in tribology, this parameter was chosen for Figure 4 and Figure 5. For the present groove structure of the surface, the Ra parameter will describe the depth of the grooves rather than the roughness on the plateau regions. Thus it can be used for estimating the change in groove depths due to the wear process in the experiments. For a qualitative impression of all surface features, the reader is referred to the surface topography images and profiles in Figure 4 and Figure 5. Aus der Praxis für die Praxis 47 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 DOI 10.30419/ TuS-2019-0005 Figure 4: Top left: Initial surface topography (~1.2 x 1.2 mm 2 ); Arrows mark sliding direction (circular arrow) and axis of rotation (straight arrow); and profile in axial direction. Centre column: surface topographies after tests performed under starting conditions SC1-3 with the corresponding profiles to the right. Bottom left: Ra values for initial surface (left green bar) and after testing with starting conditions SC1-3. Tested surfaces are larger scale (z-axis: ±2 µm) than initial surface (z-axis: ±4 µm) Figure 5: left: Ra of the initial surface and after 10, 30 and 120 minutes of tribological testing, with topographical images illustrating the surface structures; right: x-axis profiles of the initial surface and after 10, 30 and 120 minutes of tribological testing (top to bottom). Tested surfaces are larger scale (z-axis: ±2 µm) than initial surface (z-axis: ±4 µm) T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 47 brication (e.g. [3,12]), which is confirmed by a closer look at the experimentally measured Stribeck curve (Figure 1), which shows a minimum coefficient of friction of between 0.001 and 0.02. Therefore all tests but one seem to reach hydrodynamic lubrication during the 2 h test duration, if only the friction levels are taken into account. Regarding wear, the one test that did not reach hydrodynamic lubrication exhibited a notably higher steady state wear rate (~370 nm/ h) than other SC 1 tests. Even for the tests that reach hydrodynamic lubrication (according to the friction results), non-negligible steady-state wear rates (differing according to starting condition) were measured. The other tests with SC 1 had a mean steadystate wear rate of 85 ± 24 nm/ h, one test with SC2 had a wear rate of 60 nm/ h with the others averaging 11 ± 9 nm/ h, and tests with SC3 had wear rates under 4 nm/ h. Since hydrodynamic lubrication would, by definition, involve fully separated surfaces, wear measured after reaching this state leads to the assumption that wear particles are present - produced during running-in - and are responsible for abrasive wear to some degree. Both wear (Figure 2) and friction curves (Figure 3) were fitted with their respective models (equations 1 & 2), which yielded the values for the steady-state wear rates mentioned in the above, as well as the running-in duration and the ‘intensity’ of the running-in. The trend for the running-in durations is very similar for both friction and wear. The length of the running-in period depends on the starting condition - the experiments starting closer to boundary lubrication resulted in longer running-in durations than those closer to the hydrodynamic regime. The individual tests reveal an even stronger relationship between the running-in durations for wear and friction - as can be seen in Figure 6 (left), those test with longer running-in durations for wear also have longer runningin periods for friction. At some point however, the running-in durations for wear no longer rise appreciably, while friction running-in is still not concluded. The amount of running-in is also closely related for the friction and wear curves (Figure 6, right). Again, the higher the amount of running-in wear, the higher the starting coefficient of friction. This levels off for friction at around a f = 0.08-0.09, while the running-in wear intensity still rises. When looking at the changes in roughness, however, the surface topography seems to be run-in after 30 minutes. This is quite surprising, considering the wear curves show that running-in takes around 100 minutes to reach 95 % completion (for this starting condition SC 1), while the topography analysis suggests no substantial changes between 30 and 120 min. This leads to the conclusion that, for this tribosystem, running-in encompasses more than just topographical changes (i.e. the deformation of asperities and local removal of material described in [13]). Tribofilm formation is another process often associated with running-in [7], but in the present experiments, the lubricant contained no additives and therefore Aus der Praxis für die Praxis 48 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 The Ra parameters are shown in Figure 4 (bottom left) and it is clear that tests under all starting conditions reduce the Ra value considerably compared to the initial value. SC 1, the starting condition closest to boundary lubrication, leads to the lowest Ra values after the test and thus the highest wear which is the same trend as measured with the RIC method. Quantitative differences can be attributed to the different definitions of “wear depth” - the depth corresponding to the average material loss over the worn area for the RIC wear depth and the flattened grooves (which includes plastic deformation) in the centre of the wear scar for the topographic description of wear. This effect was also observed in [11], but is exacerbated in the present study because of the surface structure of the bearing bushes and the use of Ra as a proxy for wear. Figure 5 shows the evolution of the Ra roughness parameter for the 150 N SC 1 tests, where shorter tests with durations of 10 and 30 minutes were performed. The Ra values for the initial surface (t = 0), after 10 min, 30 min and 120 min of testing are graphed, along with a representative image of the surface topography at each point. The profiles normal to the sliding direction are shown on the right. To show an appropriate amount of detail, tested surfaces are larger scale (z-axis: ±2 µm) than the initial surface (z-axis: ±4 µm). The roughness is already reduced considerably after only 10 minutes due to the wear process leading to a depth reduction of the grooves, while the grooves are still clearly defined. The Ra value has been reduced to 582±44 nm from the initial value of 1463±23 nm. After 30 minutes the profile is mostly flat and most of the grooves are gone, even though the lines of the former structure are still visible. The Ra value is very small: only 123±16 nm. After 120 minutes, the Ra value is slightly higher at 192±44 nm. The slightly higher roughness can be explained by the wear process - after the initial structure (i.e. the grooves) has been removed and the surface has been smoothened during running-in, the wear particles produced by the wear process will occasionally lead to scratches increasing the roughness again. Like friction and wear rate, the evolution of roughness also seems to follow an exponential decay function (see Eq. (5) and [10]), however, the fitted exponential function does not represent the measured values well - therefore no fitted curve is shown in Figure 5 and no running-in time constant is given. 4 Discussion Both coefficient of friction and wear rate have running-in periods, after which the respective values stabilize at low steady-state levels, which is a sign that the tests achieve good conformability [2]. All tests resulted in an average coefficient of friction below 0.01 at t = 100-115 min, with the sole exception of a single test (100 N, SC 1, µ (t = 110-115 min) = 0.029). Coefficient of friction levels below 0.01 are usually associated with hydrodynamic lu- DOI 10.30419/ TuS-2019-0005 T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 48 the formation of an additive-based tribo-film can be excluded. Formation of oxides or mechanical mixing of the near-surface layers may be another possibility, but beyond the scope of this paper. 5 Conclusion The experimental investigation of the evolution of wear, friction and surface topography of bearing bushes tested at starting conditions representing near-boundary, mixed and near-hydrodynamic lubrication conditions showed that: • Regarding friction, most tests reach levels associated with hydrodynamic lubrication (independent of starting conditions), but measurable steady-state wear rates were found and differed according to starting conditions. Wear still occurring under hydrodynamic lubrication conditions can be attributed to wear particles produced during running-in leading to abrasive wear. • Starting conditions strongly influence both intensity and duration of running-in - regarding friction and wear. • When comparing the intensity and duration of runningin for friction versus wear, both follow a more or less linear trend up to a certain point. Then higher runningin intensity for wear is no longer accompanied by a similar rise in friction running-in intensity, and, conversely, the longer running-in duration for friction is not followed by a corresponding increase in running-in durations for wear. • The running-in duration for the surface topography tested under near-boundary conditions is less than 30 minutes, considerably shorter than the running-in duration for these conditions for both friction and wear. Therefore, the running-in of this particular system is governed by processes other than mere removal of asperities. These processes, for example mechanically-induced subsurface material structure changes, will be the subject of future investigations into this topic. 6 Acknowledgements This work was funded by the Austrian COMET Program (Project K2 XTribology, no. 824187 / 849109) and carried out at the “Excellence Centre of Tribology” and in cooperation with AVL List GmbH. The authors would like to thank Christoph Haslehner for running the tribometer experiments. References [1] C. KNAUDER, H. ALLMAIER, D. SANDER, S. REICH, T. SAMS: Analysis of the Journal Bearing Friction Losses in a Heavy-Duty Diesel Engine. Lubricants 2015; 3: 142-54. [2] J-L LIGIER, B NOEL: Friction Reduction and Reliability for Engines Bearings. Lubricants 2015; 3: 569-96. [3] PJ BLAU, KM COOLEY, DG BANSAL, I SMID, TJ EDEN, M NESHASTEHRIZ, ET AL: Spectrum loading effects on the running-in of lubricated bronze and surface-treated titanium against alloy steel. Wear 2013; 302: 1064-72. [4] PJ BLAU: On the nature of running-in. Tribol Int 2005; 38: 1007-12. [5] Tribotechnique - Testing of friction and wear - SLPGtester for testing of (precision engineering) plain bearings, ÖNORM M8124: 2009 2009. [6] F DITRÓI, S TAKÁCS, F TÁRKÁNYI, E CORNIANI, RW SMITH, M JECH, ET AL: Sub-micron wear measurement using activities under the free handling limit. J Radioanal Nucl Chem 2012; 292: 1147-52. Aus der Praxis für die Praxis 49 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 DOI 10.30419/ TuS-2019-0005 Figure 6: Comparison of running-in duration (left) and amount of running-in wear and starting coefficient of friction for friction and wear curves (right) T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 49 [11] T WOPELKA, U CIHAK-BAYR, C LENAUER, F DI- TRÓI, S TAKÁCS, J SEQUARD-BASE, ET AL. Wear of different material pairings for the cylinder liner - piston ring contact. Ind Lubr Tribol 2018; 70: 687-99. [12] BRUCE RW: Handbook of Lubrication and Tribology, Volume II: Theory and Design, Second Edition - CRC Press Book. Taylor & Francis Group; 2012. [13] R BOSMAN, J HOL, DJJ SCHIPPER. Running-in of metallic surfaces in the boundary lubrication regime. Wear 2011; 271: 1134-46. Aus der Praxis für die Praxis 50 Tribologie + Schmierungstechnik · 66. Jahrgang · 1/ 2019 [7] C LENAUER, C TOMASTIK, T WOPELKA, M JECH: Piston ring wear and cylinder liner tribofilm in tribotests with lubricants artificially altered with ethanol combustion products. Tribol Int 2015; 82: 415-22. [8] FT BARWELL: Wear of metals. Wear 1958; 1: 317-32. [9] R KUMAR, B PRAKASH, A SETHURAMIAH: A systematic methodology to characterise the running-in and steady-state wear processes. Wear 2002; 252: 445-53. [10] M HANIEF, MF WANI. Modeling and prediction of surface roughness for running-in wear using Gauss-Newton algorithm and ANN. Appl Surf Sci 2015. DOI 10.30419/ TuS-2019-0005 T+S_1_2019.qxp_T+S_2018 29.01.19 09: 11 Seite 50