This study examined the tribological performance of three gear oils (Oils A, B and C), in relation to surface and microstruc- tural changes. Oil A contains molybdenum dithiophosphate friction modifier, Oil B contains amine molybdate combined with zinc dialkyl dithiophosphate antiwear additive, while Oil C contains phosphonate and a commercial gear oil package. Following sliding tests of a hardened AISI 52100 steel ball on a spheroidized AISI 52100 steel disc, the worn surfaces were chemically studied using Raman and energy-dispersive X-ray spectroscopy. The tribological performance for each oil was different, likewise the nature of the tribofilm formed. After a 5 min sliding test, the hardness-depth profile of the worn surfaces was measured; also the cross-sectional microstructure was examined using scanning electron microscopy combined with focused ion beam preparation and transmission electron backscattered diffraction (t-EBSD) techniques. With Oil A, there was a relatively small increase in surface hardness (33% greater than that of the unworn surface), whereas with Oils B and C, the average hardness near the surface was 100% greater than that of the unworn surface. The cross-sectional microstructure using Oil A also differed from Oils B and C, which were quite similar. The result shows that with Oil A refinement of the ferrite grains spreads deeper into the material (> 10 µm), whilst with Oils B and C it was largely limited to 2–3 µm below the surface. It is concluded that the lubricant formulations and their associated tribofilms influenced the extent of deformation in the subsurface layers and consequently influenced the wear performance. Keywords Boundary lubrication · Gear oils · Surface chemistry · Subsurface microstructure · Mechanical properties 1 Introduction class of additives is particularly important in boundary lubri- cation conditions whereby surfaces come in direct contact Formulated oils used in mechanical systems such as gears as a result of a high load and low speed combination. In this and bearings are primarily designed to sustain low friction case, the lubricant is too thin to prevent surface asperities and wear of moving parts. Minimizing friction and wear from touching. Without these additives functioning properly improves energy efficiency and extends the lifetime of the in the formulated oils, gears and bearings can fail in cata- systems. To achieve this, friction and wear additives are strophic ways. included in the formulation of industrial lubricants. This During sliding of surfaces, these additives are surface active and contribute to the formation of a protective layer or tribofilm that prevents direct metal contact and lowers * Aduragbemi Adebogun friction between the surfaces on the bases of its low shear firstname.lastname@example.org strength relative to that of the metal . Friction and wear additives can be subdivided into subcategories. These International Centre for Advanced Materials (ICAM, include friction modifiers, antiwear and extreme pressure Manchester Hub), The University of Manchester, Manchester M13 9PL, UK additives. Other classes of additives usually added to formu- late oils include corrosion inhibitors, antifoam, dispersants School of Materials, The University of Manchester, Manchester M13 9PL, UK and detergents. The focus of this study will be on the friction and wear additives. The formation mechanism of these tri- BP Europa SE - Castrol Industrial Monchengladbach, Monchengladbach, Germany bofilms and their properties determines the level of friction and wear protection that they provide initially and over time. BP Technology Centre, Whitchurch Hill, Pangbourne RG8 7QR, UK Vol.:(0123456789) 1 3 65 Page 2 of 13 Tribology Letters (2018) 66:65 Friction modifiers such as organomolybdenum com- is a function of the combined properties of the induced sur- pounds (OMCs) react with metal surfaces to form molybde- face film and the underlying tribomutation layer. Cao et al. num disulphide (MoS ) and other compounds such as sul-  reported in a recent study that tribofilms had an effect phides and oxides [2–5]. Transmission electron microscopy on wear by influencing the plastic flow of the nanograined (TEM) characterisation of tribofilms formed on surfaces structures generated near surface of the contacts. Two dif- worn in the presence of OMCs has shown that they contain ferent oils produced very die ff rent tribol fi ms. Although they nanosheets of MoS [6, 7] which adhere well to the surfaces. produced similar levels of friction coefficient in a 2 h sliding Low friction observed with these friction modie fi rs is attrib - test, there was a large disparity in the level of wear pro- uted to the ease of sliding between single layers of MoS in tection. The difference in the level of wear protection was the tribofilms formed . MoS is Raman active, and the attributed to the ability of the tribofilm to hinder or promote 2 1 first-order Raman modes are E , E , E , A observed grain rotation. A softer film would allow for more degrees 1g 1g 2g 2g of rotation of the nanocrystalline layer, as a result reducing −1 2 around 34, 287, 383 and 409 cm , respectively . The E 2g wear significantly. The concept of surface films influenc- and A modes have peaks with the highest intensity and can ing the deformation behaviour of surfaces dates back to the 1g easily be used to identify the presence of MoS . Commonly 2 1990’s. One of the earliest research was conducted by a Rus- used antiwear and extreme pressure additives in industrial sian scientist Peter Rehbinder . He and his colleagues in applications include sulphur–phosphorus compounds, 1928 observed that the presence of a lubricant can signifi- molybdenum–sulphur compounds and ZDDP (zinc dialky- cantly affect changes that take place in a solid contact. The ldithiophosphate). ZDDP additive is the most widely used presence of oleic acid on the surface increased the ability of antiwear additive. A comprehensive review of the history a mineral crystal to deform in a plastic manner. Similarly, and mechanism of ZDDP has been reported by Spikes . in 1969, Donald Buckley  showed that the mechani- ZDDP can decompose on rubbing surfaces to form a rough cal behaviour of calcium fluoride crystal was sensitive to pad-like tribofilm of glassy zinc phosphate/polyphosphate extremely small concentrations of surfactant. Although this [10, 11] material with thickness of 50–200 nm. This tribo- concept has been investigated by some [15, 16, 19] in recent film is effective at reducing wear by acting as a sacrificial times, there is still more room for exploration particularly in layer with a faster rate of formation than the rate of removal. the development of industrial lubricants. During sliding of surfaces under high load, surface asperities The aim of this study was to investigate the relationship cannot be completely separated. Plastic deformation and/or between tribological performances of three industrial gear fracture of the asperities is inevitable. Tribofilms are able to oils (via their tribofilm) and subsurface mechanical and reduce friction and can influence plastic deformation of the microstructural changes. surface asperities depending on their properties. Lower fric- tion at the surface will reduce the level of frictional strain induced in the subsurface and play a significant role in wear 2 Materials and Methods reduction . Frictional strain during sliding is usually localised very near the surface and continues to accumulate 2.1 Lubricant with more cycles of sliding. Under significantly high load and sufficient number of sliding cycles, most ferrous metals The test lubricants were three ISO VG 320 gear oils, which form a refined grain structure a few nanometres in depth all have the same kinematic viscosity of 320 ± 32 mm /s at . This layer can eventually delaminate to form plate-like 40 °C. Oil A is used in wind power gearboxes, Oil B in wear debris . heavy industrial applications (such as mining vehicles) and Friction and wear performance of different industrial oils Oil C is a multifunctional gear oil used in a variety of appli- have been widely studied and reported in the literature. An cations. The complete formulation of the lubricants includes established approach to understanding the mechanisms by the base oil, and the complete additive mix are shown in which industrial oils perform in boundary lubricated slid- Table 1. Oil A is formulated with synthetic polyalphaolefin ing is examining the properties of the tribofilm formed and (PAO) base oil and an additive package that includes molyb- surface topography. There is a limited amount of studies that denum dithiophosphate (MoDTP) friction modifier, mixed have investigated the relationship between tribological per- sulphonates and an antifoam. Oil C is also formulated with formance, the nature of tribofilm formed and the subsurface synthetic PAO base oil with a different additive mix, con - changes. One of those who have studied this relationship is taining phosphonate, alkylated naphthalene, antifoam and Reichelt et al. . They studied the surface film formed a commercial gear oil package. Oil B is formulated with during triboaction and the subsurface layer (or tribomuta- a mineral base oil and additive package that combines a tion layer) that formed beneath the surface film. Their work friction modifier (amine molybdate), an antiwear additive showed that the wear protection provided by formulated oils (ZDDP), extreme pressure sulphurised additive and metal 1 3 Tribology Letters (2018) 66:65 Page 3 of 13 65 Table 1 Industrial formulated oils and their base oil/additive mix Oil A Oil B Oil C Base oil Polyalphaolefin (> 90%) Group 1 base stock mix (> 85%) Polyalphaolefin (> 80%) Alkylated naphthalene (5-15%) Additives Molybdenum dithiophosphate (MoDTP)—(< 5%) Amine molybdate complex (< 5%) Phosphonate (< 1%) ZDDP (< 5%) commercial gear oil Sulphurised—extreme pressure additive package (< 5%) Mixed sulphonates—corrosion inhibitor (< 0.5%) Sulphonate mix—corrosion inhibitor (< 5%) Copper—corrosion inhibitor (< 0.5%) Antifoam—trace Antioxidant (< 1%) Antifoam—trace Methacrylate polymer—pour point depressant (< 5%) sulphonate (corrosion inhibitor). These three oils with dif- formation of a fully insulating film means that the electrical ferent base oil-additive mixes were selected with the inten- contact voltage reaches a maximum voltage of 15 mV. tion of generating different types of tribofilms with different tribological properties. 2.3 Ball and Disc 2.2 Tribotesting The test ball and disc were made of AISI 52100 steel. The Sliding tests were carried out using a high-frequency recip- ball is 6.0 mm in diameter with hardness of 11.7 ± 1.0 GPa rocating rig (HFRR, PCS Instruments, London, England). and surface finish or roughness (Ra) of less than 0.05 µm. The test configuration is a ball-on-disc, where the disc is The test disc is 10 mm in diameter and 3 mm thick, with a held stationary, submerged in the test lubricant. Load is hardness of 3.3 ± 0.2 GPa and surface roughness of about applied on the test ball which oscillates linearly on the disc. 0.02 µm. Before each test, the test ball and disc were cleaned For the entire sliding test, a load of 9.8 N is applied on the with toluene and acetone. After each experiment, the test test ball and it oscillates on the disc with a frequency of samples were cleaned with heptane to remove the residual 50 Hz for a stroke length of 1 mm providing average Hertz- oil. The test discs were further degreased with soapy water ian contact pressure of 0.96 GPa. All tests were carried out and ethanol prior to characterisation in the scanning electron at 80 °C by heating up the lubricant bath and maintaining the microscope. temperature throughout the test. The temperature of 80 °C was selected to match the temperature at which gear oils typical operate in industrial systems. The test duration was 2.4 Wear Measurement varied between 5 min and 2 h. Each test was carried out three times to give average values of friction coefficient. The A KEYENCE VK—X200 3D confocal microscope was used HFRR system measures and stores the friction coefficient to scan the profile of the wear scar on the test disc, followed continuously throughout the test. Additionally, electrical by post-processing of the measurement with KEYENCE VK contact resistance (ECR) is monitored. ECR gives an indi- Analyser. Tilt correction was applied to adjust for the slight cation of whether or not a separating film is formed between tilt of the sample on the microscope stage. Wear volume was the ball and disc. A 15-mV potential is applied between the calculated by the product of the wear scar length and the ball and disc which form a resistance referred to as the con- cross-sectional areas. The average of 10 2D profiles along tact resistance; this potential is also applied to a balance the wear scar was used to determine the cross-sectional area. resistor which is in series with the contact resistance and The Ra of the worn surfaces was measured by applying a forms a potential divider . A series resistance of 1000 cut-off wavelength (λ ) of 0.025 µm to the primary profile Ω is selected in the HFRR control software. The change which eliminate the influence of the longer wavelength in potential between the contacts (referred to as electrical components. contact voltage) is a measure of the ‘contact resistance’ in comparison with the balance resistor. A zero electrical 2.5 Raman Spectroscopy contact voltage indicates a direct metal–metal contact, and no electrical contact resistance between the contact and the A Renishaw 1000 microscope system was used to analyse residual tribochemical products on the wear scars. The 1 3 65 Page 4 of 13 Tribology Letters (2018) 66:65 wavelength of the laser source was 514 nm. All scans were 2.8 SEM–FIB carried out with an Olympus objective lens which gives a laser spot size of about 1 µm. Raman spectra were obtained A SEM equipped with FIB was used to characterise the from the wear scar at 1 s exposure and 20 accumulations to subsurface microstructure of the spheroidized AISI 52100 reduce the signal-to-noise ratio. steel discs before and after sliding test with the three oils. The study was carried out using a FEI Nova NanoLab 600i system. The dual beam system has secondary electron and 2.6 SEM–EDX focused ion beams. Electron images were taken with volt- age of 5 kV and current of 16 nA. Ion imaging was taken A scanning electron microscopy (SEM) FEI Quanta 650 with voltage of 30 kV and beam current of 9.7 pA. Imaging FEG equipped with energy-dispersive X-ray (EDX) spec- of the subsurface microstructure was done using ion chan- troscopy detector was used to characterise the wear surface nelling contrast technique . With this technique, grains topography and map the elements present on the surface. appear either dark or bright based on their orientation. The Surface imaging and EDX mapping were performed with region of interest is set as the centre of the scar for the three accelerating voltages of 5 and 30 kV, respectively. worn surfaces. This is one of the advantages of using the FIB technique, as it allows site-specific analysis. The FIB 2.7 Nanoindentation process (Fig. 1) involves firstly depositing a platinum (Pt) layer on the surface to protect the region of interest; a trench The hardness of the near surface after sliding tests was is generated by ion beam milling to reveal the cross-sectional measured using the MTS Nano Indenter XP. The continual microstructure. stiffness measurement (CSM) mode makes it possible to continuously measure mechanical response with depth as 2.9 Transmission‑EBSD an indenter tip penetrates the surface [21, 22]. The Diamond Berkovich indenter tip was displaced into the material at a T-EBSD in comparison with the conventional EBSD tech- −1 constant strain rate of 0.05 s to a depth of 2 µm. Hardness nique provides improved spatial resolution as a result of a measurements were taken on the as-received surface and the reduced interaction volume of the electron beam in the three worn surfaces. sample . This is particularly useful in resolving nano- sized grains. An FEI Magellan 400 XHR SEM equipped with NordlyNano EBSD detector was used to take EBSD Fig. 1 SEM micrographs showing the process of subsurface microstructural examination using the SEM–FIB technique 1 3 Tribology Letters (2018) 66:65 Page 5 of 13 65 measurements. Using FIB lift-out technique  in an FEI Oil C contains sulphur-and-phosphorus-based antiwear and Nova NanoLab 600i dual beam system, a thin FIB lamella extreme pressure additives. (< 200 nm) was generated from the cross sections (Fig. 9b, The three fully formulated oils were tested in the HFRR c) and mounted in transmission geometry before characteri- tribometer for durations of 5 min and 2 h. Three tests were sation with the SEM. An accelerating voltage of 30 kV was run for 5 min, and another three were run for 2 h continu- used with probe current of 1.6 nA and a step size of 20 nm. ously. The recorded mean value of friction coefficient at The Kikuchi patterns were indexed using Aztec 2.2 software. the end of the 5 min tests was averaged and is reported in EBSD data processing was done using Channel 5 software Fig. 3a. The same goes for the 2 h tests. After the test, suite developed by Oxford Instruments HKL. wear measurements were taken from the wear scars gener- ated on the test discs since there was no evidence of wear on the test balls. At the start of the 2 h sliding test, Oil A 3 Results and Discussion provides a low friction coefficient of about 0.08 and this drops slightly to 0.07 within 5 min (Fig. 2a). Raman spec- 3.1 Friction and Wear Performance troscopy results indicated that MoS was present on the worn surface at this time (Fig. 4) as evident by E and A 1g 2g In this section, friction and wear performance are discussed −1 peaks at 378 and 411 cm , respectively. The EDX ele- in relation to tribofilm formation on the surface. Although mental map shows the surface contains oxygen, sulphur all the additives in the mix play a role in the performance and calcium after 5 min (Fig. 5). One of the additives of the formulated oils, this discussion focuses on the role of included the formulation of Oil A is calcium sulphonate. the friction and wear additives. Both Oils A and B contain The elemental print of calcium on the wear scar indicates organmolybdenum friction modifiers, MoDTP and amine that the calcium sulphonate is surface active and contrib- molybdate, respectively. Oil B contains ZDDP antiwear utes to the tribofilm formed. The formation of MoS additive in addition to amine molybdate friction modifier. Fig. 2 Plots of friction coefficient and ECR film coverage for a Oil A, b Oil B and c Oil C Fig. 3 Average a friction coefficient and b wear volume for the three oils after 5 and 120 min of HFRR sliding test 1 3 65 Page 6 of 13 Tribology Letters (2018) 66:65 Fig. 4 Raman spectra obtained from surfaces lubricated by Oils A and B after 5 and 120 min (2 h). The red dot at the centre of the crosshair in the optical images shows the location each spectrum was taken from Fig. 5 SEM images and EDX maps of wear scars lubricated by Oils A, B and C after 5 min and 2 h between sliding metal surfaces is known to be the cause of surface generated after 2 h of sliding (Fig. 5) shows an significant reduction in friction [2 , 4, 5, 26]. From 5 min increase in the intensity of oxygen, sulphur and calcium. to the end of the 2 h test, the friction coefficient continues As the friction coefficient drops, the ECR plot (Fig. 2a) to drop steadily to reach 0.06. MoS remains present on shows the build-up of an unsteady insulating film. At the the worn surface after 2 h of sliding as evidenced by the end of the 2 h test, wear has increased substantially with Raman spectra in Fig. 4. The elemental map of the wear Oil A (Fig. 3b). This suggests that the insulating film was 1 3 Tribology Letters (2018) 66:65 Page 7 of 13 65 either insufficient to protect against wear or contributed to The average friction coeci ffi ent and wear volume for each the increase. of the oils at 5 and 120 min are presented in Fig. 3. Oil A When testing Oil B, the friction coec ffi ient was 0.1 for the provides the lowest friction after both 5 and 120 min of slid- first 10 min of the sliding test and this was then followed by ing. This can probably be attributed to the formation of low a sharp drop to 0.065 (Fig. 2b). This drop in friction coef-shear MoS tribofilm at the contact, which prevents direct ficient has been attributed to the formation of a low friction metal to metal contact and reduces friction. However, in film [27, 28]. The ECR result for Oil B shows a sharp rise in the 2 h test, wear increased considerably. The result shows electrical contact resistance at the start of sliding suggest- that Oil A provides relatively good friction performance, but ing a film forms quickly at the contact. Raman spectroscopy average antiwear performance, particularly over longer peri- was used to scan the surface for evidence of MoS forma- ods. Oils B and C provide higher average friction coefficients tion on the surface after 5 min and 2 h. The results pre- of 0.087 and 0.089, respectively, after 5 min of sliding test sented in Fig. 4 indicate that MoS was formed after 5 min compared to Oil A. After 2 h of sliding test, Oil B provides and remains present at the end of the 2 h test. The EDX the lowest wear and forms an insulating film that covers elemental map of the surface shows that oxygen is present about 75% of the contact surfaces (Fig. 2). Oil B appears to on the surface after 5 min; and after 2 h of sliding, sulphur, form a complex tribofilm consisting of MoS and a phos- phosphorus and zinc are present in addition to oxygen. The phate constituent. This complex tribofilm provides relative presence of zinc and phosphorus after 2 h of sliding coupled low friction and excellent wear performance. The elemental with the relatively low wear obtained with Oil B suggests map of the surface worn by Oil C for 2 h suggests that a S–P- the formation of zinc phosphate film which is known to play based film might have formed on the surface although this a significant role in the low wear performance when ZDDP may not have reflected on the ECR plot due to its conductive is present [10, 11]. nature. The roughness of the friction coefficient plot and the Similar to Oil A, the friction coefficient of Test Oil C is substantial increase in wear after 2 h suggests that any film initially at about 0.08, but the friction coefficient profile for formed was probably too unstable to adequately protect the Oil C is rougher than that of Oils A and B, as evident by surface, as evident by the relatively high wear. the short order spikes and dips. This suggests that the tribo- The surface topography of the worn surfaces after 2 h is film, if present, is very unstable, and this is also implied by presented in Fig. 6. The worn surfaces from the three oils the lack of electrical contact signal throughout the 2 h test (Fig. 6a–c) show evidence of plastic deformation caused by (Fig. 2c). The EDX elemental map (Fig. 5) shows that oxy- ploughing of a hardened AISI 52100 steel balls on the softer gen and a smaller proportion of sulphur are present on the AISI 52100 steel discs. Figure 7 shows the cross section surface after 5 min. After 2 h, the surface contains oxygen, profile of the wear scars generated with the three oils which sulphur and phosphorus. corresponds to the wear result (Fig. 3b) after 2 h sliding. For Fig. 6 SEM images of wear surfaces generated from 2 h sliding test with Oil A (a, a1), Oil B (b, b1) and Oil C (c, c1) 1 3 65 Page 8 of 13 Tribology Letters (2018) 66:65 3 frictional heat and will raise the temperature of the surfaces significantly even in well lubricated surfaces . The rest of the energy is partitioned into other processes including: formation and shearing of tribofilm, transformation of the microstructure (plastic deformation) and the process of wear . How frictional energy is dissipated during sliding has Oil A been the subject of previous studies [30–32]. These studies -1 highlight the importance of considering just how frictional Oil B -2 energy is dissipated to different process taking place in a Oil C tribosystem. Evaluating the energy balance in the tribosys- -3 0 200 400 600 800 tem during the sliding test is outside the scope of this study. Posion (μm) Consequently, analysis of the results in this study has been simplified by assuming that the percentage of frictional Fig. 7 Cross section profiles of the wear scars generated from 2 h energy dissipated to tribofilm formation and shearing, plastic sliding tests with Oils A, B and C deformation and wear is equal for the three oils. Oil A pro- vided the lowest average friction coefficient of 0.073 during Oil A, the higher magnification micrograph (Fig. 6a1) shows the 5 min sliding test, while Oils B and C provided similar the softer ferrite matrix organised in a wavy pattern around levels of friction at 0.087 and 0.089, respectively. Based the hard cementite particles, which has a darker contrast. on the assumption made, the higher friction coefficient of The surface worn by Oil B (Fig. 6b1) appears smooth and Oils B and C would equate to higher stored energy in the homogenous, which was confirmed by the surface rough- subsurface layer and wear. ness measurement. For Oil B, the Ra is 0.12 µm, which Nanoindentation technique was used to measure the makes it the smoothest of the three surfaces. On the other hardness-depth profile for the unworn substrate and those hand, the surface lubricated by Oil C appears to be rough worn using Oils A, B and C to a maximum depth of 2 µm. with a Ra value of 0.21 µm; it is the roughest of the three The hardness-depth profile for the unworn surface in Fig. 8a surfaces. Also, the surface appears to be densely populated shows that the average hardness is approximately 3 GPa with cementite particles. The cementite particles appear to and was relatively homogenous up to the measured depth of be sticking-out, which could be the reason for the high sur- 2 µm. The hardness-depth profiles for Oils B and C in Fig. 8c face roughness value measured. This combined with the high and d, respectively, are similar and different to that of Oil density of cementite particles possibly suggests that most of A. With Oils B and C, there is significant rise in hardness the soft ferrite matrix surrounding the hard cementite parti- towards the top surface from 3.5–4 GPa at 2 µm below the cles has been worn away. surface to 7.5–10 GPa at 200 nm. However, with Oil A there is a relatively low increase in the hardness from 3GPa in the 3.2 Subsurface Changes (Mechanical unworn substrate to average of 4 GPa. Another distinction and Microstructural) with Oil A is that the increase in hardness is uniform with depth up to 2 µm. The significant hardening of the subsur - Section 3.1 focuses on how the tribological performance face layer with Oils B and C in comparison with Oil A is of the gear oils relates to the nature and extent of tribofilm perhaps the first indication that there might be a correlation formed and also the surface topography. However, the tribo- between the higher level of friction coefficient of Oils B and logical performance of the oils is also a function of changes C to the higher level of energy stored in the subsurface layer. taking place beneath the surface. The aim of this section is Further examination of subsurface transformation was to investigate the effect of friction and wear performance on carried out using FIB technique combined with ion chan- subsurface changes (mechanical and microstructural) and nelling contrast imaging (ICCI) to reveal subsurface micro- any correlation that might exist. Subsurface characterisation structure. This was preferred to the conventional approach was conducted on the worn surfaces generated after 5 min of of sectioning, grinding and polishing, because of the rela- sliding in the HFRR tribometer. Since the test ball is approx- tively small size of the wear scar (approximately 1.5 mm by imately 3 times harder than the test disc (substrate), plastic 0.5 mm). Additionally, this technique makes it possible to deformation and wear are expected to occur entirely on the analyse roughly the same location (centre of the wear scar) substrate. No wear was observed on the test ball. Therefore, for each of the three different samples (Fig. 1). Figure 9a investigation of subsurface changes was conducted on the shows the subsurface microstructure of the substrate in the spheroidized AISI 52100 steel substrate. unworn state which consists of cementite (FeC) particles Frictional energy generated during sliding of surfaces is uniformly distributed in large ferrite grains with average dissipated in many ways. Most of the energy is dissipated as grain size of 15 µm. As surfaces plastically deforms during 1 3 Wear depth (µm) Tribology Letters (2018) 66:65 Page 9 of 13 65 Fig. 8 Variation of hardness with depth on a the unworn surface, and the worn surface after sliding test of 5 min with b Oil A, c Oil B and d Oil sliding, dislocations or defects are generated and accumulate microstructure below the surface worn with Oil A (Fig. 9b) in the ferrite grains. As sliding progresses, these dislocations reveals refined ferrite grains of varying sizes surrounded with strain energy associated with them begin to form dislo- by cementite particles to a large depth greater than 10 µm cation tangles (DT) or dislocation dense walls (DDW) . below the surface. To better characterise the grain struc- Intersecting DDW subdivides the original ferrite grains into tures observed in Fig. 9b and c, the transmission electron dislocation cells or refined blocks, and with the accumula- backscatter diffraction (t-EBSD) technique described in tion of more dislocations, these cells evolve into sub-grains Sect. 2.9 was employed. The process involved extracting with small mis-orientations between them. With progres- out a thin slice of about 100–150 nm from the cross section sive refinement, of the ferrite grains, the subsurface layer shown in Fig. 9b and c. The thin slice was then examined is substantially work-hardened and explains the increase in in transmission mode. This analysis was conducted for the hardness in Fig. 9. The presence of the cementite particles in cross sections of only Oils A and B since the cross-sectional ferrite facilitates refinement of the large ferrite grains . microstructures of Oils B and C are quite similar. The EBSD The cross-sectional microstructures of surfaces lubricated inverse pole figure maps (Fig. 10a, c) allow clear observation by Oils B and C (Fig. 9c, d) are similar and different to of the grain size, grain orientation which is colour-coded that of Oil A (Fig. 9b). With Oils B and C, the subsurface and the grain size distribution with depth. One obvious microstructure reveals a top layer (< 3 µm thick) of refined distinction between the microstructures of Oils A and B in ferrite grains above large ferrite grains with the cementite Fig. 10a and c, respectively, is the difference in the depth of particles uniformly distributed. However, the subsurface the refined grains. Grain refinement is localised very near the 1 3 65 Page 10 of 13 Tribology Letters (2018) 66:65 Fig. 9 Microstructures beneath a the unworn surface, and the worn surface after sliding test of 5 min with b Oil A, c Oil B and d Oil C surface with Oils B and C and corresponds to the hardness profile and subsurface microstructures. Strain localisation profiles in Fig. 8c and d, respectively. However, with Oil A can be induced by a stress gradient developing in the sub- grain refinement spreads deeper into the material. Also with surface layer during mechanical processes . Oil A, some of the grains within the top 3 µm below the Going by the earlier assumption made that the percent- surface appear elongated and below 3 µm the average size age of frictional energy partitioned into tribofilm forma- of the refined ferrite grain appears larger than those in the tion and shearing, plastic deformation and wear process top layer for Oil B. The subsurface microstructures in Figs. 9 are equal during sliding with the three oils, higher wear and 10 correspond to the hardness profiles in Fig. 8, and yet (Fig. 3b) would be expected with Oils B and C which both again, there is similarity between subsurface changes with have higher average friction coefficient (Fig. 3a) in compari- Oils B and C. This is perhaps a second indication that there son with Oil A after 5 min of sliding. Whilst that was the might be a correlation between higher friction coefficients case in this study, it is not always so. It is clear from previ- obtained with Oils B and C to the localisation of strain near ous studies [16, 19] that although different oils can produce the surface, whereas with Oil A providing lower friction a similar steady-state friction coefficient during boundary coefficient the strain accumulated to a larger depth below lubricated sliding this does not always correlate to similar the surface. Friction coefficient influences the subsurface levels of wear. Nevertheless, subsurface microstructure gen- stresses during sliding. The zone of maximums stress moves erated during boundary lubrication sliding can be related to from the subsurface closer to the surface as friction coef- the process of wear. Nanocrystalline layers typically formed ficient increases [35 , 36]. This might explain why with Oils below the surfaces under boundary lubrication play a crucial B and C with higher friction coefficient strain is localised role in generating wear particles  and the extent of wear nearer to the surface as evidenced by the hardness-depth [39, 40]. Buscher et al.  in their study of nanocrystalline 1 3 Tribology Letters (2018) 66:65 Page 11 of 13 65 Fig. 10 Cross-sectional EBSD images of FIB lamella extracted from contrast map for Oils A and B, respectively. Black lines represent the centre of the worn surfaces: a, c inverse pole figure map for Oils high-angle GB’s (> 10°), and white lines represent low-angle GB’s A and B, respectively, b, d grain boundary superimposed on band (2° ≤ θ ≤ 10°) layer and wear particle generation found that the size of wear test influences how much friction is generated in the tri- debris during tribo-process corresponds to the nanocrystal- bosystem, how friction evolves and is dissipated into other line grain size. Therefore, they suggest that the wear parti- simultaneous processes such as microstructural transfor- cles are generated from nanocrystal torn off from the subsur - mation and wear. Although it was assumed that the fric- face layer. Higher wear with Oils B and C might be linked to tional energy generated during sliding is partitioned into significant localisation of strain very near the surface form- other simultaneous processes in the same manner for the ing fine ferrite grains near the surface in comparison with different oils; by simply changing the lubricant used, the Oil A where the strain accumulation spreads deeper into the dynamics of how friction is dissipated into heat and other material. Perhaps as the refined grains gets smaller near the simultaneous process will likely change. There are several surface the boundary between the fine ferrite grains and the characteristics of the tribofilm formed that can influence hard cementite particles weakens forming voids and eventual the dynamics. Some of these characteristics have been delamination of wear debris. investigated by other researchers in the past including: The only variable in the sliding tests conducted for this mechanical properties (shear strength, hardness, etc.) of study was the lubricants and their corresponding additives the tribofilm [ 41–45], how quickly forms and stabilizes mix. This will influence the nature and extent of tribofilm [46–48] and its durability [49–52]. formed during sliding as discussed in Sect. 3.1. The tri- bofilms formed from the different oils during the sliding 1 3 65 Page 12 of 13 Tribology Letters (2018) 66:65 mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- 4 Conclusions tion, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the This experimental paper has investigated the importance Creative Commons license, and indicate if changes were made. of combining surface characterisation (surface chemistry and topography) with subsurface examination (mechani- cal and microstructural changes) to better understand the References mechanisms by which industrial oils perform tribologi- cally in boundary lubricated sliding. Hence, the following 1. Spalvins, T.: A review of recent advances in solid film lubrication. J. Vac. Sci. Technol. A 5(2), 212–219 (1987) conclusions are drawn: 2. Guerret-Piécourt, C., Grossiord, C., Le Mogne, T., Martin, J., Palermo, T.: Tribochemical interactions between molybdenum • The combined results of ECR film formation generated dithiophosphate and succinimide additives. Surf. Interface Anal. by the HFRR tribometer during the sliding test with ex 30(1), 646–650 (2000) 3. Yamamoto, Y., Gondo, S.: Friction and wear characteristics of situ chemical characterisation (Raman and EDX spec- molybdenum dithiocarbamate and molybdenum dithiophosphate. troscopy) of the worn surface improve understanding Tribol. Trans. 32(2), 251–257 (1989) of the tribofilm formation process and its influence on 4. Yamamoto, Y., Gondo, S.: Environmental effects on the composi- tribological performance. tion of surface films produced by an organo-molybdenum com- pound©. Tribol. Trans. 37(1), 182–188 (1994) • The gear oil formulation determines the nature of tri- 5. Unnikrishnan, R., Jain, M.C., Harinarayan, A.K., Mehta, A.K.: bofilm formed on the surfaces during sliding; conse- Additive–additive interaction: an XPS study of the effect of ZDDP quently, it influences the type of transformation that on the AW/EP characteristics of molybdenum based additives. takes place in the metal subsurface layer mechanically Wear 252(3–4), 240–249 (2002). https ://doi.org/10.1016/S0043 -1648(01)00865 -1 and microstructurally. 6. Grossiord, C., Varlot, K., Martin, J.M., Le Mogne, T., Esnouf, After 5 min of sliding, Oil A provided lower friction C., Inoue, K.: MoS2 single sheet lubrication by molybdenum and wear compared to Oils B and C. When compared dithiocarbamate. Tribol. Int. 31(12), 737–743 (1998). https://doi. to the original surface, Oil A produced a 33% increase org/10.1016/S0301 -679X(98)00094 -2 7. Evans, R.D., Doll, G.L., Hager, C., Howe, J.Y.: Influence of steel in the surface hardness after sliding. Also, the depth of type on the propensity for tribochemical wear in boundary lubrica- deformation evident by refined grains near the surface tion with a wind turbine gear oil. Tribol. Lett. 38(1), 25–32 (2010) is > 10 µm. Conversely, Oils B and C produced over 8. Wieting, T.: Long-wavelength lattice vibrations of MoS2 and 100% increase in surface hardness and localised defor- GaSe. Solid State Commun. 12(9), 931–935 (1973) 9. Spikes, H.: The history and mechanisms of ZDDP. Tribol. Lett. mation to a depth of 2–3 µm below the surface. 17(3), 469–489 (2004) Tribosystems are complex due to the inter-relationship 10. Jahanmir, S.: Wear reduction and surface layer formation by a between simultaneously occurring processes such as ZDDP additive. ASME, Trans. J. Tribol. 109, 577–586 (1987) generation of frictional heat, tribofilm formation and 11. Martin, J.M., Grossiord, C., Le Mogne, T., Bec, S., Tonck, A.: The two-layer structure of Zndtp tribofilms: part I: AES, XPS and shearing, plastic deformation and wear. This makes XANES analyses. Tribol. Int. 34(8), 523–530 (2001) it difficult to make direct correlations for exam- 12. Jahanmir, S.: The relationship of tangential stress to wear particle ple between frictional performance and subsurface formation mechanisms. Wear 103(3), 233–252 (1985) changes. 13. Rigney, D., Glaeser, W.: The significance of near surface micro- structure in the wear process. Wear 46(1), 241–250 (1978) • The nature of subsurface changes in tribological pro- 14. Rice, S., Nowotny, H., Wayne, S.: A survey of the development cesses influences the extent of wear therefore to better of subsurface zones in the wear of materials. In: Key Engineering understand how industrial lubricants and the tribofilms Materials, pp. 77–100. Trans Tech Publ (1989) they form influence wear it is important to investigate 15. Reichelt, M., Gunst, U., Wolf, T., Mayer, J., Arlinghaus, H.F., Gold, P.W.: Nanoindentation, TEM and ToF-SIMS studies of their influence on subsurface changes. the tribological layer system of cylindrical roller thrust bearings lubricated with different oil additive formulations. Wear 268(11), 1205–1213 (2010) Acknowledgements The authors would like to gratefully acknowl- 16. Cao, S., Sarasin, F., Cantoni, M., Mischler, S.: Effect of surface edge the financial support of the Engineering Physical Sciences films on tribologically induced metallurgical transformations of Research Council, UK (EPSRC), through Centre for Doctoral Train- steel in oil lubricated contacts. Wear 368, 75–83 (2016) ing in Advanced Metallic Systems and the BP International Centre for 17. Rehbinder, P.: About influence of changing surface on cleav - Advanced Materials (BP-ICAM). Additionally, the BP-Castrol advisory age hardness and other crystal properties. In: Proceedings of 6th team (Hocine Faci and Oliver Sfreddo), David Lewis, Aleks Tedstone, Physical Congress, Moscow (1928) Bart Winiarski, Etienne Bousser are sincerely thanked for useful dis- 18. Buckley, D.H.: Ee ff ct of surface active media on friction, deforma - cussions. Appreciation also goes to Andrew Forrest for his assistance tion, and fracture of calcium fluoride (1969) with nanoindentation experiments. 19. Qu, J., Bansal, D.G., Yu, B., Howe, J.Y., Luo, H., Dai, S., Li, H., Blau, P.J., Bunting, B.G., Mordukhovich, G.: Antiwear perfor- Open Access This article is distributed under the terms of the Crea- mance and mechanism of an oil-miscible ionic liquid as a lubri- tive Commons Attribution 4.0 International License (http://creat iveco cant additive. ACS Appl. Mater. Interfaces. 4(2), 997–1002 (2012) 1 3 Tribology Letters (2018) 66:65 Page 13 of 13 65 20. Kajdas, C., Majzner, M.: Boundary lubrication of low-sulphur localization in metals: high resolution strain cross sectioning via diesel fuel in the presence of fatty acids. Lubr. Sci. 14(1), 83–108 synchrotron X-ray diffraction. J. Eng. Mater. Technol. 130(2), (2001) 021005–021010 (2008). https ://doi.org/10.1115/1.28409 62 21. Oliver, W.C., Pharr, G.M.: An improved technique for determining 38. Büscher, R., Täger, G., Dudzinski, W., Gleising, B., Wimmer, hardness and elastic modulus using load and displacement sensing M., Fischer, A.: Subsurface microstructure of metal-on-metal hip indentation experiments. J. Mater. Res. 7(6), 1564–1583 (1992) joints and its relationship to wear particle generation. J. Biomed. 22. Li, X., Bhushan, B.: A review of nanoindentation continuous stiff- Mater. Res. B Appl. Biomater. 72(1), 206–214 (2005) ness measurement technique and its applications. Mater. Charact. 39. Kolubaev, A., Tarasov, S., Sizova, O., Kolubaev, E.: Scale-depend- 48(1), 11–36 (2002) ent subsurface deformation of metallic materials in sliding. Tribol. 23. Suzuki, T., Endo, N., Shibata, M., Kamasaki, S., Ichinokawa, T.: Int. 43(4), 695–699 (2010) Contrast differences between scanning ion and scanning electron 40. Panin, V., Kolubaev, A., Tarasov, S., Popov, V.: Subsurface layer microscope images. J. Vac. Sci. Technol. A 22(1), 49–52 (2004) formation during sliding friction. Wear 249(10–11), 860–867 24. Keller, R.R., Geiss, R.H.: Transmission EBSD from 10 nm (2001) domains in a scanning electron microscope. J. Microsc. 245(3), 41. Ye, J., Kano, M., Yasuda, Y.: Evaluation of local mechanical prop- 245–251 (2012) erties in depth in MoDTC/ZDDP and ZDDP tribochemical reacted 25. Mayer, J., Giannuzzi, L.A., Kamino, T., Michael, J.: TEM sample films using nanoindentation. Tribol. Lett. 13(1), 41–47 (2002) preparation and FIB-induced damage. MRS Bull. 32(5), 400–407 42. Ye, J., Kano, M., Yasuda, Y.: Evaluation of nanoscale friction (2007) depth distribution in ZDDP and MoDTC tribochemical reacted 26. Sarin, R., Tuli, D., Sureshbabu, A., Misra, A., Rai, M., Bhatnagar, films using a nanoscratch method. Tribol. Lett. 16(1–2), 107–112 A.: Molybdenum dialkylphosphorodithioates: synthesis and per- (2004) formance evaluation as multifunctional additives for lubricants. 43. Bec, S., Tonck, A., Georges, J.-M., Roper, G.W.: Synergistic Tribol. Int. 27(6), 379–386 (1994) effects of MoDTC and ZDTP on frictional behaviour of tribofilms 27. Khaemba, D.N., Neville, A., Morina, A.: A methodology for at the nanometer scale. Tribol. Lett. 17(4), 797–809 (2004) Raman characterisation of MoDTC tribofilms and its application 44. Hashimoto, C., Okubo, H., Tadokoro, C., Sasaki, S.: Correlation in investigating the influence of surface chemistry on friction per - between nano-mechanical and macro-tribological properties of tri- formance of MoDTC lubricants. Tribol. Lett. 59(3), 38 (2015) bofilms derived from organic phosphoric additives. Tribol. Online 28. Graham, J., Spikes, H., Korcek, S.: The friction reducing proper- 11(5), 632–638 (2016) ties of molybdenum dialkyldithiocarbamate additives: part I—fac- 45. Landauer, A.K., Barnhill, W.C., Qu, J.: Correlating mechanical tors influencing friction reduction. Tribol. Trans. 44(4), 626–636 properties and anti-wear performance of tribofilms formed by (2001) ionic liquids, ZDDP and their combinations. Wear 354, 78–82 29. Bowden, F., Ridler, K.: Physical properties of surfaces-III—The (2016) surface temperature of sliding metals-The temperature of lubri- 46. Tung, S.C., Gao, H.: A study of break-in film development with cated surfaces. In: Proc. R. Soc. Lond. A 1936, vol. 883, pp. different piston ring coatings and correlation with electrical 640–656. The Royal Society contact resistance measurements. Tribol. Trans. 46(3), 326–331 30. Fouvry, S., Liskiewicz, T., Kapsa, P., Hannel, S., Sauger, E.: An (2003) energy description of wear mechanisms and its applications to 47. Fujita, H., Glovnea, R., Spikes, H.: Study of zinc dialkydithi- oscillating sliding contacts. Wear 255(1–6), 287–298 (2003) ophosphate antiwear film formation and removal processes, part 31. Klamecki, B.E.: Energy dissipation in sliding. Wear 77(1), 115– I: experimental. Tribol. Trans. 48(4), 558–566 (2005) 128 (1982) 48. Ji, H., Nicholls, M.A., Norton, P.R., Kasrai, M., Capehart, T.W., 32. Huq, M., Celis, J.-P.: Expressing wear rate in sliding contacts Perry, T.A., Cheng, Y.-T.: Zinc-dialkyl-dithiophosphate antiwear based on dissipated energy. Wear 252(5–6), 375–383 (2002) films: dependence on contact pressure and sliding speed. Wear 33. Tao, N.R., Wang, Z.B., Tong, W.P., Sui, M.L., Lu, J., Lu, K.: 258(5–6), 789–799 (2005) An investigation of surface nanocrystallization mechanism 49. Morina, A., Neville, A.: Understanding the composition and low in Fe induced by surface mechanical attrition treatment. Acta friction tribofilm formation/removal in boundary lubrication. Tri- Mater. 50(18), 4603–4616 (2002). https://doi.or g/10.1016/S1359 bol. Int. 40(10–12), 1696–1704 (2007) -6454(02)00310 -5 50. Morina, A., Neville, A.: Tribofilms: aspects of formation, stability 34. Zhou, L., Liu, G., Ma, X., Lu, K.: Strain-induced refinement in a and removal. J. Phys. D Appl. Phys. 40(18), 5476 (2007) steel with spheroidal cementite subjected to surface mechanical 51. Komvopoulos, K., Chiaro, V., Pakter, B., Yamaguchi, E., Ryason, attrition treatment. Acta Mater. 56(1), 78–87 (2008) P.: Antiwear tribofilm formation on steel surfaces lubricated with 35. Buckley, D.H.: Influence of surface active agents on friction, gear oil containing borate, phosphorus, and sulfur additives. Tri- deformation, and fracture of lithium fluoride (1964) bol. Trans. 45(4), 568–575 (2002) 36. Tangena, A.G.: Tribology of thin film systems. (1987) 52. Minfray, C., Martin, J.-M., Lubrecht, T., Belin, M., Le Mogne, 37. Croft, M., Jisrawi, N., Zhong, Z., Horvath, K., Holtz, R.L., Shep- T.: A novel experimental analysis of the rheology of ZDDP tri- ard, M., Lakshmipathy, M., Sadananda, K., Skaritka, J., Shukla, bofilms. In: Tribology Series, vol. 41. pp. 807–817. Elsevier, V., Sadangi, R.K., Tsakalakos, T.: Stress gradient induced strain Amsterdam (2003) 1 3
Tribology Letters – Springer Journals
Published: Apr 10, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera