- Email: [email protected]
Tribological studies of potential vegetable oil-based lubricants containing environmentally friendly viscosity modifiers
Tribology International 69 (2014) 110–117
Contents lists available at ScienceDirect
Tribology International journal homepage: www.elsevier.com/locate/triboint
Tribological studies of potential vegetable oil-based lubricants containing environmentally friendly viscosity modifiers L.A. Quinchia a,d, M.A. Delgado a,b,n, T. Reddyhoff c, C. Gallegos a,b,d, H.A. Spikes c a
Departamento Ingeniería Química, Facultad de Ciencias Experimentales, Universidad de Huelva, Campus de “El Carmen”, 21071 Huelva, Spain Pro2TecS – Chemical Process and Product Technology Research Center, Universidad de Huelva, 21071 Huelva, Spain c Tribology Section, Department of Mechanical Engineering, Imperial College of Science, Technology and Medicine, London SW7 2AZ, UK d I&D Centre Complex Formulations, Fresenius Kabi Deutschland GmbH, Daimlerstrasse 22, 61352 Bad Homburg, Germany b
art ic l e i nf o
a b s t r a c t
Article history: Received 3 June 2013 Received in revised form 29 August 2013 Accepted 30 August 2013 Available online 7 September 2013
The amphiphilic properties that result from the fatty acid composition of vegetable oils contribute to a better lubricity and effectiveness as anti-wear compounds than mineral or synthetic lubricant oils. Despite these advantages, vegetable oils show only a limited range of viscosities and this constrains their use as suitable biolubricants in many industrial applications. For the reason, ethylene–vinyl acetate copolymer (EVA) and ethyl cellulose (EC) have been added to the vegetable oil-based lubricants studied. To address this issue, the frictional and lubricant film-forming properties of improved vegetable oilbased lubricants (high oleic sunflower (HOSO), soybean (SYO) and castor (CO) oils), blended with 4% (w/ w) of EVA and 1% (w/w) of EC, have been studied. It has been found that castor oil shows the best lubricant properties, when compared to high oleic sunflower and soybean oil, with very good filmforming properties and excellent friction and wear behaviour. This can be attributed to its hydroxyl functional group that increases both the viscosity and polarity of this vegetable oil. Regarding the effect of the viscosity modifiers studied, ethylene–vinyl acetate copolymer exerts a slight effect on lubricant film-forming properties and, thus, helps to reduce friction and wear mainly in the mixed lubrication region. Ethyl cellulose, on the other hand, was much more effective, mainly with castor oil, in improving both mixed and boundary lubrication. & 2013 Elsevier Ltd. All rights reserved.
Keywords: Vegetable oils Boundary lubrication Viscosity modifiers Biolubricants
1. Introduction Lubricants have an important role in world industrial and economic development, mainly by reducing friction and wear in mechanical contacts [1,2]. Thus, about 38 million metric tonnes per year of lubricants have been used globally in last decade, with the majority of these being petroleum-based [3,4]. In the last 25 years, there has been an increasing interest in the use of biodegradable products. This has been driven by environmental problems that have heightened the need to limit pollution from lubricants and hydraulic fluids based on mineral oils. Vegetable oils are potential substitutes for petroleum-based oils; not only they are environmentally friendly, renewable and less toxic, but also they have excellent lubricating properties such as high viscosity index, high lubricity and low volatility [5,6]. For these reasons, vegetable oil-based lubricants are being actively demanded for many green industrial activities [7,8]. n Corresponding author at: Departamento Ingeniería Química, Facultad de Ciencias Experimentales, Universidad de Huelva, Campus de “El Carmen”, 21071 Huelva, Spain. Tel.: þ 34 959219997; fax: þ34 959219983. E-mail address: [email protected] (M.A. Delgado).
0301-679X/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.triboint.2013.08.016
Vegetable oils can act as anti-wear additives and friction modifiers, due to strong interactions with the lubricated surfaces. Their amphiphilic nature gives them a good film/force relationship, due to long fatty acid chains and the presence of polar groups in the vegetable oil structure [9,10]. For this reason, vegetable oilbased lubricants have the peculiarity of being effective as both boundary and hydrodynamic lubricants [2,9,11]. However, to understand fully the tribological properties of vegetable oils, it is important to know the effects of the variability in fatty acid composition on their lubricating properties, film thickness formed, friction and wear [12]. Despite their safety, efficacy and advantages as mentioned above, vegetable oils suffer from several major drawbacks in terms of thermal and oxidative stability, which preclude their use above 120 1C, crystallisation at relatively high temperatures, and the limited range of viscosities available [13–15]. The last of these is critical since a range of lubricant viscosities are required depending on the industrial application. For instance, kinematic viscosity, at 40 1C, ranges typically from 30 mm2/s in the automotive industry, to 120 mm2/s for lubricants used on bearings, while viscosities higher than 240 mm2/s are demanded in lubricants for four-stroke engines and some gear assemblies [5,16]. In previous
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
works, the authors have investigated the use of environmentally friendly additives to improve some of these shortcomings [14,15,17,18]. More specifically, ethylene–vinyl acetate copolymer (EVA) and ethyl cellulose were tested to increase the viscosity range of different vegetable oils and improve their thermal susceptibilities [14,15,18]. However, the introduction of these additives could alter the properties required in lubricant applications; therefore, it is necessary to determine the tribological behaviour of these vegetable oil-based lubricants. The contribution of viscosity modifiers to the viscous flow behaviour of lubricants has been extensively studied and is well understood. However, their influence on the friction and wear is less clear [19,20]. This is because these additives may behave in different and complex ways in lubricated contacts, each of which might contribute to a greater or lesser extent to film formation and, thus, friction and wear [19]. Previous studies with mineral oils have shown that some viscosity modifiers are able to adsorb on to metal surfaces, improving the boundary film formation under low speed and high temperature, producing a significant reduction in friction and wear [19,21,22]. In this sense, the main objective of the research described in this paper was to determine the tribological behaviour of three potential vegetable oil-based lubricants blended with polymers and, in this way, identify links between lubricant composition, friction and wear performance to find the suitable biolubricants features for use in industrial applications.
2. Materials and methods 2.1. Vegetable oils and additives The environmentally friendly base-stocks (vegetable oils) used in this work were soybean oil (SYO), castor oil (CO), and high-oleic sunflower oil with 85% (w/w) oleic acid (HOSO). SYO was supplied by Fresenius Kabi (Germany), CO was received from Guinama (Spain) and HOSO was kindly supplied by the “Instituto de la Grasa”, CSIC (Spain). All of them were commercial grade oils without any further purification. The fatty acid compositions and physical properties of the vegetable oils are shown in Tables 1 and 2, respectively. Ethylene–vinyl acetate copolymer (EVA) with 33% vinyl acetate content (density, at 23 1C, 0.956 g cm 3; molecular weight, 60250 g mol 1; melting temperature, 59 1C), was kindly supplied, in the form of pellets, by Repsol YPF, S.A. (Spain). EVA copolymer is considered inert, nontoxic, and stable material. It is not expected to be biodegradable but is not hazardous, according to the Commission Directive 93/21/EEC [23]. The other viscosity modifier, and biodegradable material, used was ethyl cellulose (EC) (density, at 25 1C, 1.14 g cm 3; molecular weight, 68960 g mol 1; melting temperature, 155 1C). This viscosity modifier is commercially available and was obtained from Sigma Aldrich. 2.2. Preparation of environmentally friendly lubricating formulations EVA and EC were blended with the different vegetable oils used in this study, at a concentration of 4% (w/w) and 1% (w/w), respectively. Blends were prepared by stirring in an open vessel
111
(800 g), at constant velocity (150 rpm) and using an anchor impeller geometry, to disperse the polymer in the oil. EVA and EC blends were heated up to 120 1C and 155 1C, respectively, for approximately 1 h. After that, when the polymers were completely dissolved, the mixtures were returned to room temperature by natural convective cooling. A homogeneous single phase was obtained in all cases, except the blend HOSO/EC that, after some hours, became cloudy. 2.3. Viscosity and density measurements Dynamic viscosities were measured with a rotational controlled-strain rheometer (ARES, TA Instruments, USA), over a temperature range of 25–100 1C. Viscous flow tests were carried out, in a shear rate range of 5–500 s 1, using coaxial cylinders geometry (inner radius 16 mm, outer radius 17 mm, cylinder length 33.35 mm). Kinematic viscosity values, ν, were obtained as the ratio of the dynamic viscosity to the density, at each temperature. The viscosity indexes (VI) were obtained according to ASTM D-2270. A capillary densimeter, model DMA-5000 (Anton Paar, Austria), was used to measure sample densities in a temperature range of 15–100 1C. 2.4. Film thickness measurement The film forming properties of the lubricant samples were determined using ultrathin film interferometry. For this purpose, an EHL optical interferometry rig (PCS Instruments, UK) was used. A high-pressure contact is formed between the flat surface of a glass disk and a reflective steel ball. The glass disk is coated with chromium and silica layers (5 nm and 500 nm, respectively). The principle of this technique is fully described by Johnston et al. [24]. Film thickness measurements were made using a 19.5 mm diameter steel ball in nominally pure rolling contact with a coated glass disk of 100 mm diameter. Both ball and disk were ultrasonically cleaned in toluene, followed by acetone, prior to a test, and a new ball was used for each test. The load applied was 20 N, corresponding to a maximum Hertz contact pressure of 0.54 GPa. Film thickness measurements were carried out over a temperature range of 40–100 1C, and a range of entrainment speeds between 0.005 and 3.0 m/s. The refractive index of the lubricant film must be known to determine the actual film thickness. In the current work, this was measured using an Abbe 60 refractometer from Bellingham and Stanley Ltd. (UK). The measurements were made over a temperature range of 40–100 1C for each vegetable oil and its blends with the viscosity modifier. 2.5. Friction measurements The minitraction machine (MTM) technique allows determination of the “Stribeck curve”, which plays an important role in the identification of the different lubrication regimes: hydrodynamic (HL), elastohydrodynamic (EHL), mixed lubrication (ML) and boundary lubrication (BL) [25]. In this method, a rolling-sliding lubricated contact is formed between a steel ball and the flat
Table 1 Fatty acid composition of the vegetable oils studied. Vegetable oils
Palmitic (16:0)
Stearic (18:0)
Oleic (18:1)
Linoleic (18:2)
Linolenic (18:3)
Ricinoleic (18:1:OH)
Unsaturated/saturated ratio
HOSO CO SYO
3.84 2.63 11.28
4.42 1.51 2.70
83.66 4.74 24.39
8.08 8.36 56.28
– – 5.34
– 82.8 –
11.10 23.20 6.15
112
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
Table 2 Some physical properties of the vegetable oils and lubricating blends studied. Samples
HOSO HOSO þEVA 4% HOSO þEC 1% CO COþEVA 4% COþEC 1% SYO SYOþ EVA 4%
Viscosity 40 1C
Viscosity 100 1C
VI
Dynamic (mPa s)
Kinematic (mm2/s)
Dynamic (mPa s)
Kinematic (mm2/s)
36.00 151.40 45.68 230.00 403.96 518.20 31.00 137.89
38.50 168.14 50.60 242.50 427.10 546.70 33.60 174.00
8.70 25.49 13.04 19.00 46.27 36.76 7.60 30.10
9.90 29.59 15.10 21.00 50.90 40.30 8.60 27.30
surface of a polished steel disk. The disk is fully immersed in lubricant, and the temperature of the lubricant and contact are controlled to a set value within 7 0.5 1C. In these tests, the slideto-roll ratio is held constant and the friction coefficient is measured while varying the entrainment speed. The test specimens used in the MTM were a polished 19.05 mm ball and a 50.0 mm diameter disk, according to specification AISI 52100. Both had a surface root mean square roughness of 11 nm7 1 nm. The load applied was 30 N, corresponding to a maximum Hertz pressure of 0.93 GPa. Tests were carried out over the entrainment speed range of 0.005–2.5 m/s, at a fixed slide-to-roll ratio (SRR) of 0.5 (50%) and in a temperature range of 40–100 1C. A new ball was used in each test. The test rig and specimens were cleaned in the same fashion as in film thickness measurements. Friction tests were also carried out on a high frequency reciprocating rig (HFRR), supplied by PCS Instruments, London (UK). In this method, a 6.0 mm diameter steel ball is held in a chuck and loaded downwards on the flat face of a 10.0 mm diameter steel disk. The disk is held in a bath, which contains lubricant so that the contact between the ball and the disk is fully immersed. The ball is oscillated backwards and forwards in contact with the disk at a stroke length and fixed frequency [22]. Friction is monitored continuously, while wear is determined from the wear scar on the steel ball at the end of the test, by averaging the scar diameters transverse and along the rubbing direction. The test conditions used in this test are listed in Table 3. New balls and disks, cleaned successively in toluene and acetone in an ultrasonic bath, were used for each test. 2.6. Statistical analysis At least, two replicates of each test were done on fresh samples and data shown have statistically significant values, i.e. they did not exceed a significance level of 0.05 in Student's t-test and had a 95% confidence interval.
3. Results and discussion 3.1. Film thickness measurements In this current work, the ability of three vegetable oils (HOSO, CO and SYO), and their blends with two viscosity modifiers (EVA and EC), to form a lubricating film that reduce the friction between solid surfaces in contact has been evaluated. The effect of the vegetable oil-based lubricant composition, temperature and entrainment speed on the film-forming properties has also been studied. For this purpose, the optical interferometry technique has been used [26] and the refractive index of the different blends can be seen in Table 2.
257 218 314 116 183 118 250 214
Density 15 1C (g/cm3)
0.9152 0.9160 0.9192 0.9630 0.9629 0.9622 0.9256 0.9252
Refractive index 40 1C
100 1C
1.466 1.465 1.465 1.476 1.475 1.474 1.472 1.470
1.450 1.451 1.448 1.460 1.461 1.455 1.456 1.456
Table 3 HFRR test conditions. Test duration Stroke length Stroke frequency Test load Temperature Ball properties Disc properties
60 min 1 mm 50 Hz 4N 40, 100 1C AISI 52100, 800 VPN AISI 52100, 750 VPN
In addition to this, it has worth pointing out that ethylene-vinyl acetate copolymer [15] has been successfully tested as viscosity modifier for several common vegetable oils, yielding potentially environmental friendly lubricants for some applications. In the case of ethyl cellulose, the addition of 1% (w/w) of EC always yields an important increase in vegetable oil viscosity within the temperature range studied (Table 2). The most important viscosity increments correspond to CO/EC blends (i.e. 125%, at 40 1C). On the contrary, the lowest increments correspond to HOSO/EC blends, mainly at low temperature.
3.1.1. Effect of vegetable oil type and temperature on film thickness formation The film-forming process of high oleic sunflower, soybean and castor oils, at 40 and 100 1C, has been studied. As can be seen in Fig. 1a, at 40 1C, the film thickness developed by all vegetable oils studied increases linearly with entrainment speed, on a log–log plot, in agreement with the elastohydrodynamic (EHL) theory [27]. This indicates the absence of any anomalous behaviour down to 10 nm thickness, which is evidenced by a change in the slope of the film thickness–speed curves at low speeds. The flattening out of the curve for CO at very high speeds is due to thermal effects. HOSO and SYO display similar film thickness in the whole entrainment speed range assessed. For instance, at 0.1 m/s, they show film thickness values of 49.3 and 49.8 nm, respectively. However, CO, due to its high viscosity, was able to form thicker films (167.7 nm), at low temperatures, than the other neat oils studied. As can be seen in Fig. 1, over most of the speed range, film thickness displays a decrease with increasing temperature. At 100 1C (Fig. 1b), film thickness dependence on entrainment speed (u), for these three vegetable oils, is different, primarily at low entrainment speed. HOSO and CO form apparent thick boundary films (12 and 7 nm, respectively) at very low entrainment speeds. SYO does not form such a film and, indeed, at low speeds, its film thickness appears to fall below linearity on a log– log plot. In addition to this, at low speeds, film formation is not controlled by bulk viscosity, since CO, which is much more viscous than HOSO, forms a slightly thinner boundary film, while SYO, which is only marginally less viscous that HOSO, forms a much thinner film at low speeds.
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
113
Film thickness (nm)
103
102
101 HOSO CO SYO
100
10-1
10-2
10-1
100
101
10-2
Entrainment speed (m/s)
10-1
100
101
Entrainment speed (m/s)
Fig. 1. Effect of vegetable oil type and entrainment speed on the film thickness at (a) 40 1C and (b) 100 1C.
Film thickness (nm)
103
102
101
10-2
10-1
100
Entrainment speed (m/s)
101
10-2
10-1
100
101
Entrainment speed (m/s)
Fig. 2. Effect of viscosity improvers on the film thickness at (a) 40 1C and (b) 100 1C.
According to Biresaw and Bantchev [12], the film thickness in the thin film region can be influenced by physico-chemical properties and chemical composition (degree of unsaturation, functional groups and chain length) rather than viscosity. In this sense, it can be pointed out that film thickness increases when the degree of unsaturation decreases [28]; thus, SYO has high level of unsaturation and its film thickness is the lowest. However, as can be seen in Fig. 1b, HOSO formed a thicker boundary film (12 nm) than CO (7 nm) although the degree of unsaturation of both oils are similar [28]. This difference in film thickness at low speed is possibly due to both the difference in polarity [12] and the tendency of polar components to separate from the bulk and adsorb onto the rubbing surfaces as film thickness decreases [29]. Based on this statement, these results suggest that the fractionation of these vegetable oils in ultrathin films results in a deviation in the measured film thickness from that predicted from the bulk viscosity based on EHD theory [10,29]. Thus the presence of a relatively low viscosity and highly polar linolenic component in SYO may result in a thinner film at low entrainment speed.
3.1.2. Effect of viscosity improvers on film thickness formation of vegetable oil-based lubricants According to the results obtained in the last section, two of the vegetable oils (HOSO and CO), which displayed the best lubricating properties, were chosen to be blended with the viscosity modifiers studied, aiming to evaluate their tribological properties. Although the main role of these polymers is, generally, to influence the bulk rheological properties of their blends, it has been shown that some polymers are also able to form boundary lubricating films [19,20]. Fig. 2 shows the film thicknesses for the blends of HOSO and CO with 4% (w/w) EVA and 1% (w/w) EC, at two temperatures (40 and 100 1C).
Fig. 2a shows the behaviour of HOSO with both viscosity improvers at the two temperatures studied. At 40 1C, the HOSO/ EVA blend shows a linear evolution (log–log plot) of film thickness with entrainment speed; this is similar to the neat oil, indicating no boundary film formation. At high temperature (100 1C), the addition of 4% (w/w) EVA to HOSO avoids the boundary film formation seen with neat HOSO. Indeed, EVA has the effect of reducing film thickness below linearity at low entrainment speed. According to Zhu [30], EHL films can break down due to different reasons: low speed or viscosity, heavy load and roughness. Thus, the lower viscosity of HOSO and the lower solubility of EVA in this vegetable oil, in comparison to CO, may contribute to break down the EHL film at the lowest entrainment speeds. Fig. 2a also shows the behaviour of HOSO blended with 1% EC. For the tests ran at 40 1C, the cellulose has the effect of forming a very thick boundary film ( 70 nm). At 100 1C, the film formed in the EHL contact was so thick and unstable that film thickness measurement was not possible. Fig. 2b shows the behaviour of CO with both viscosity improvers at the two temperatures studied. The addition of 4% (w/w) EVA to CO has the effect of increasing low speed film thickness, which is in agreement with the formation of a viscous boundary film within the temperature range studied. At 40 1C, the lubricant forms a boundary film of around 50 nm at 0.01 m/s, while, at 100 1C, the boundary film thickness formed dropped to 15 nm. By contrast, the addition of 1% (w/w) EC to CO has only a marginal effect on the film thickness, in the whole entrainment speed range, at 40 1C, while, at 100 1C, a similar effect to the addition of 4% (w/w) EVA is noticed. The results above are consistent with the general mechanism of boundary film formation by viscosity modifier polymers, reported by Smeeth et al. [19], which suggest that polymer molecules are adsorbed on the polar solid surface (steel and silica, in this case) to form a film with a higher viscosity than that of the bulk
114
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
0.1 100ºC HOSO SYO CO
Friction coefficient
40ºC HOSO SYO CO
0.01
10-2
10-1
100
101
Entrainment speed (m/s)
Fig. 3. MTM friction coefficient of the neat vegetable oils studied at 40 1C and 100 1C.
solution. At high speed, the lubricating oil/polymer blend is dragged into the contact inlet, forming an EHD film whose thickness varies according to the viscosity at high shear rate. This is seen with all blends studied at high entrainment speed (Fig. 2), taking into account that the addition of 4% (w/w) EVA or 1% (w/w) EC yields an increase in vegetable oil viscosity. However, when the speed is decreased, in thin-film conditions, the contact inlet is filled with the adsorbed polymer layer, much more viscous than the lubricating blend, which leads to a thicker than expected boundary film. This fact is dramatically observed for the HOSO/EC blend at 40 1C. In the other cases, at very low speeds, the film is compressed to a thickness of a few nanometres, depending on the ability of the polymer to support the applied pressure and form a boundary film [20,21,31]. 3.2. Frictional properties by MTM and HFRR 3.2.1. Effect of vegetable oil nature and temperature on frictional properties Fig. 3 shows the Stribeck curves, at 40 and 100 1C, for the different vegetable oils studied. At 40 1C (Fig. 3a), the curve shows a very different lubrication behaviour for CO as compared to SYO and HOSO. SYO and HOSO show classical Stribeck’s curves, with friction coefficients falling from a high value of ca 0.06, at low entrainment speeds (and thus thin EHL films), to a low value of ca 0.023, at high entrainment speeds when the contact is in full EHD lubrication. Thus, the friction curves represent the transition from mixed boundary-EHD to full film EHD friction. However, CO has a much higher viscosity and this means that it forms a full EHD film, even at entrainment speeds as slow as 0.01 m/s. Thus, its friction coefficient curve represents full film EHD friction over its whole speed range, showing a drop-off, at high speeds, due to thermal heating of the film as the sliding speed increases with entrainment speed. It is noteworthy that the EHD friction coefficient of CO (0.035) is much higher than that for the other two vegetable oils. This must be a result of intermolecular bonding between the hydroxyl groups on the ricinoleic molecules, which are also responsible for the higher viscosity of CO. At 100 1C (Fig. 3b), quite similar Stribeck’s curves can be seen for the three vegetable oils studied, excepting the higher EHD friction coefficient for CO. EHD friction measured at high speeds is lower at 100 1C than at 40 1C. This reduction in EHD friction with an increase in temperature is a general phenomenon seen for all fluids. It can be observed that, in the temperature range studied, the Stribeck curve for HOSO shows no reduction in friction at low speeds, even though optical film thickness measurements showed
a marked boundary film. This discrepancy probably results from the remarkable sliding in the MTM contact, as opposed to the rolling optical contact and the different tribosystem used (steel/ steel contact for the MTM versus glass/steel contact in optical interferometry technique). Regarding SYO, it is noteworthy that the friction coefficient is lower than that of HOSO at intermediate speeds and 100 1C since SYO forms a thinner film than HOSO under these conditions. It may be due to some interactions with the rubbing steel surfaces of the most SYO unsaturated components to form protective boundary films under the sliding conditions of the MTM. Fig. 4 shows the friction coefficient of the different vegetable oils studied, obtained using HFRR, at 40 and 100 1C. At 40 1C (Fig. 4a), CO has the highest friction coefficient ( 0.089) and, after a short surface-conditioning phase, constant and stable behaviour over time. In contrast, SYO and HOSO friction coefficients fluctuate with time, a fact that suggests a worse lubrication, but with lower friction and wear than CO. The HFRR forms a contact that operates in boundary lubrication conditions with extensive initial solidsolid contact. This behaviour of CO, where friction falls, initially, from a high value, is indicative of the formation of a protective surface film, promoted by tribochemical processes due to the rubbing action. By contrast, SYO and HOSO appear to form protective films almost immediately, perhaps reflecting the presence of more strongly adsorbing species. The higher steady-state boundary friction of CO is consistent with its high EHD friction coefficient and suggests that the boundary film formed by this oil is dominated by the ricinoleic component. At 100 1C (Fig. 4b), CO continues to show a very stable friction coefficient with a lower value than at 40 1C. The friction coefficients of SYO and HOSO, at 100 1C, are higher and less regular than at 40 1C. The wear scars obtained by HFRR (Fig. 5) are consistent with the friction coefficients, the high wear for CO, at 40 1C, probably occurring primarily in the initial high friction part of the test. The differences observed at high temperature, in both friction and wear, reflect the strength of the boundary film formed by the oils on the surface. Those formed by SYO and HOSO are relatively weak and, thus, less able to withstand rubbing at 100 1C than at 40 1C. It is likely that the protective films formed by vegetable oils having many polyunsaturations are weak and unstable over time, and therefore less effective [32]. Castor oil appears to form a much stronger film, probably assisted by its higher viscosity and its more polar feature, it means, hydrogen bonding from the hydroxyl groups of ricinoleic molecule could be formed on the steel surface. These latest results demonstrate, once again, that the frictional properties of different vegetable oils are strongly affected by the chemical composition thereof. In this sense, the results obtained in the boundary lubrication regime show a beneficial influence of the hydroxyl group of the ricinoleic acid in the lubrication process, achieving more stable lubricating conditions with time. This functional group in CO is the main responsible factor for the intermolecular interactions favouring the high viscosity of castor oil and the relatively high EHD friction coefficient, and gives the polar character to the oil, favouring the formation of a strong boundary film, effective in friction and wear [33].
3.2.2. Effect of viscosity improver type on vegetable oil-based lubricants friction coefficient As previously discussed for film thickness results, the addition of 4% (w/w) EVA to HOSO had the effect of reducing low speed film thickness (Fig. 2a). This seems to demonstrate that no boundary film was formed with 4% (w/w) EVA. The corresponding Stribeck curves (Fig. 6a) show negligible effect on boundary friction, at very low entrainment speed, or on EHD friction, at
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
115
Friction coefficient
0.15
0.10
SYO
CO
HOSO SYO HOSO
0.05
CO
0
1000
2000
3000
4000 0
1000
2000
3000
4000
Rubbing time (s)
Rubbing time (s)
Fig. 4. HFRR friction coefficient for the neat vegetable oils studied at (a) 40 1C and (b) 100 1C.
Fig. 5. Wear scars, obtained by HFRR, for the neat vegetable oils studied.
Friction coefficient
0.1
0.01
10-2
10-1
100
Entrainment speed (m/s)
101
10-2
10-1
100
101
Entrainment speed (m/s)
Fig. 6. Effect of viscosity improvers on the MTM frictional properties for (a) high oleic sunflower oil (HOSO) and (b) castor oil (CO).
high entrainment speed. However, the addition of EVA does appear to produce a reduction in friction at intermediate, especially at 100 1C. This suggests the presence of a on the surfaces [19], which promotes separation of the surfaces and, thus, the transition from mixed to full film EHD lubrication. No such effect was seen by using optical interferometry, suggesting that the formation of this film was promoted by steel–steel rubbing in the MTM test. This is further supported by the HFRR friction values obtained, which decrease from 0.08 down to 0.06 for the EVAcontaining HOSO-based lubricant, at 100 1C (Fig. 7).
On the other hand, the presence of 1% (w/w) EC in a HOSO-based lubricant results in a very different lubrication regime distribution. The Stribeck curve for the HOSO/EC blend supports the existence of a solid-like boundary film, which exhibits significantly reduced friction at 100 1C (Fig. 6a). Thus, the results obtained at 100 1C show friction coefficients below 0.02 over the whole speed range. By contrast, as shown in Fig. 7, EC effect on the boundary friction, from HFRR tests, is to increase friction. This may result from the increased resistance to motion that occurs during ploughing of the thick solid boundary film and the action of the resulting debris.
116
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
Friction Coefficient
0.15
0.10 HOSO HOSO+EC1% HOSO+EC1% HOSO+EVA4%
HOSO
0.05 HOSO+EVA4%
Friction Coefficient
0.15 0.00
0.10
CO
CO+EVA4%
CO+EVA4%
0.05
CO
CO+EC1%
CO+EC1%
0.00
0
1000
2000
3000
Rubbing time (s)
0.0 4000 0
1000
2000
3000
4000
Rubbing time (s)
Fig. 7. Effect of viscosity improvers on the HFRR friction behaviour for high oleic sunflower oil (HOSO) and castor oil (CO) at (a) 40 1C and (b) 100 1C.
Fig. 8. Wear scars, obtained by HFRR, for blends of neat vegetable oils and the viscosity improvers studied.
The addition of EVA and EC to CO has a negligible effect on friction at 40 1C. This was expected since, at this temperature, a full EHD film is present, which is unaffected by the addition of low concentrations of polymer. At 100 1C, the addition of EVA was found to increase low speed film thickness (Fig. 2b), which is in agreement with the formation of a viscous boundary film (Fig. 6b). On the contrary, the addition of 1% (w/w) EC has a negligible effect on friction at 100 1C. Regarding the results obtained by using the HFRR technique, the friction properties were evaluated under very severe conditions, as this test operates in almost full boundary lubrication conditions rather than in mixed lubrication, as is the case for MTM tests [34]. Thus, HFRR tests measure the ability of a lubricant to form strong boundary films to withstand high asperity loads at low speeds rather than its ability to form viscous film due to enhanced entrainment [21]. According to the results obtained (Fig. 7), it can be inferred that EVA is much more effective in mixed lubrication conditions (Stribeck's curve), whilst, on the contrary, EC is more effective in extreme boundary lubrication conditions, mainly at 40 1C. The wear results support these observations (Fig. 8). Moreover, these experimental data suggest a more positive effect, on the tribological behaviour, of the addition of ethyl cellulose to CO than to HOSO. Thus, the effect, in the case of high oleic sunflower oil, is
totally different due to poor solubility of the polymer in this vegetable oil.
4. Conclusions Both castor and high oleic sunflower oils form quite thick boundary films (8 to 12 nm), which can be seen at low speeds and 100 1C, when negligible EHD film is generated. By contrast, soybean oil shows no boundary film formation, and indeed a thinner film than expected from EHD theory, yielding also much higher friction and wear. This may be attributed to its higher content in polyunsaturated fatty acids, which results in a possible thermal degradation of the boundary film. Castor oil shows better EHD film-forming properties compared to high oleic sunflower and soybean oils. Thus, castor oil also shows excellent behaviour in boundary friction and, thus, friction and wear characteristics. This enhanced lubricity of castor oil can be attributed to its hydroxyl functional group that increases the viscosity and polarity of the oil. On the other hand, it is important to note that viscosity modifiers can affect the lubricating properties of vegetable oils. In this sense, ethylene-vinyl acetate copolymer exerted a slight effect on oil filmforming properties and, thus, helps to reduce friction and wear
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
mainly in the mixed lubrication region. On the contrary, ethyl cellulose was much more effective, mainly with castor oil, in boundary lubrication as well as mixed lubrication. In addition to this, it has worth pointing auto that the most important viscosity increments correspond to CO/EC blends (i.e. 125%, at 40 1C).
Acknowledgements This work is part of two Research Projects (PSE-320100-2006-1 and PSE-420000-2008-4) sponsored by MEC. One of the authors (L.A. Quinchia) has received a Ph.D. Research Grant from FPU-MEC Programme (Spain). The authors gratefully acknowledge their financial support. References [1] Rakic R. The influence of tribological properties of lubricating oils on the reliability of gears. Industrial Lubrication and Tribology 1999;51:170–9. [2] Maleque M, Masjuki H, Sapuan S. Vegetable-based biodegradable lubricating oil additives. Industrial Lubrication and Tribology 2003;55:137–43. [3] Siniawski MT, Saniei N, Adhikari B, Doezema LA. Influence of fatty acid on the tribological performance of two vegetable-based lubricants. Journal of Synthetic Lubrication 2007;24:101–10. [4] Shashidhara YM, Jayaram SR. Vegetable oils as a potential cutting fluid – an evolution. Tribology International 2010;43:1073–81. [5] Lathi P, Mattiasson M. Green approach for the preparation of biodegradable lubricant base stock from epoxidized vegetable oil. Applied Catalysis B 2007;9:207–12. [6] Jayadas N, Nair K. Coconut oil as base oil for industrial lubricants-evaluation and modification of thermal, oxidative and low temperature properties. Tribology International 2006;39:873–8. [7] Rudnick LR. Synthetics, mineral oils, and bio-based lubricants: chemistry and technology. In: Rudnick LR, Erhan SZ, editors. Natural oils as lubricants. New York: CRC/Taylor & Francis Group; 2006. [8] Lazzeri L, Mazzoncini M, Rossi A, Balducci E, Bartolini G, Giovannelli L, et al. Biolubricants for the textile and tannery industries as an alternative to conventional mineral oils: an application experience in the Tuscany province. Industrial Crops and Products 2006;24:280–91. [9] Adhvaryu A, Erhan SZ. Tribological studies of thermally and chemically modified vegetable oils for use as environmentally friendly lubricants. Wear 2004;257:359–67. [10] Biresaw G. Elastohydrodynamic properties of seed oils. Journal of the American Oil Chemists' Society 2006;83:559–66. [11] Fox NJ, Stachowiak GW. Vegetable oil-based lubricants – a review of oxidation. Tribology International 2007;40:1035–46. [12] Biresaw G, Bantchev G. Effect of chemical structure on film-forming properties of seed oils. Journal of Synthetic Lubrication 2008;25:159–83. [13] Miles P. Synthetics versus vegetable oils: applications, options, and performance. Journal of Synthetic Lubrication 1998;15:43–52.
117
[14] Quinchia LA, Delgado MA, Valencia C, Franco JM, Gallegos C. Viscosity modification of high-oleic sunflower oil with polymeric additives for the design of new biolubricant formulations. Environmental Science and Technology 2009;43: 2060–5. [15] Quinchia LA, Delgado MA, Valencia C, Franco JM, Gallegos C. Viscosity modification of different vegetable oils with EVA copolymer for lubricant applications. Industrial Crops and Products 2010;32:607–12. [16] Mang T, Dresel W. Lubricants and Lubrication. second ed.. Germany: WileyVCH; 2001. [17] Quinchia LA, Delgado MA, Valencia C, Franco JM, Gallegos C. Natural and synthetic antioxidant additives for improving the performance of new biolubricant formulations. Journal of Agricultural and Food Chemistry 2011;59: 12917–24. [18] Quinchia LA, Delgado MA, Franco JM, Spikes HA, Gallegos C. Low-temperature flow behaviour of vegetable oil-based lubricants. Industrial Crops and Products 2012;37:383–8. [19] Smeeth M, Gunsel S, Spikes HA. The formation of viscous surface films by polymer solutions: boundary or elastohydrodynamic lubrication. Tribology Transactions 1996;39:720–5. [20] Müller M, Topolovec-Miklozic K, Dardin A, Spikes HA. The design of boundary film-forming PMA viscosity modifiers. Tribology Transactions 2006;49: 225–32. [21] Müller M, Fan J, Spikes HA. Design of functionalized PAMA viscosity modifiers to reduce friction and wear in lubricating oils. Journal of ASTM International 2007;4:1–10. [22] Fan J, Müller M, Stöhr T, Spikes HA. Reduction of friction by functionalised viscosity index improvers. Tribology Letters 2007;28:287–98. [23] Commission Directive 93/21/EEC of 27 April 1993 adapting to technical progress for the 18th time Council Directive 67/548/ EEC on the approximation of the laws, regulations and administrative provisions relating to the classification, packaging and labelling of dangerous substances, Off J. E.C. L110, 1993. [24] Johnston GJ, Wayte R, Spikes HA. The measurement and study of very thin lubricant films in concentrated contacts. Tribology Transactions 1991;34: 187–94. [25] Moshkovich A, Perfilyev V, Lapsker I, Rapoport L. Stribeck curve under friction of copper samples in the steady friction state. Tribology Letters 2010;37: 645–53. [26] Spikes HA, Cann PM. Techniques for studying the behaviour lubricants in elastohydrodynamic contacts. Acta Tribologyca 1992;1:33–48. [27] Anghel V, Bovington C, Spikes HA. Thick-boundary-film formation by friction modifier additives. Lubrication Science 1999;11:313–35. [28] Graciani E. Los aceites y grasa. Composición y propiedades. first ed. Madrid; A Madrid Vicente : Mundi-Prensa: 2006. [29] Guangteng G, Spikes HA. Fractionation of liquid lubricants at solid surfaces. Wear 1996;200:336–45. [30] Zhu D. On the EHL film breakdown. Tribological Research and Design for Engineering Systems 2003;41:559–71. [31] Fox NJ, Tyrer B, Stachowiak GW. Boundary lubrication performance of free fatty acids in sunflower oil. Tribology Letters 2004;16:275–81. [32] Asadauskas S, Perez JM, Duda JL. Oxidative stability and antiwear properties of high oleic vegetable oils. Lubrication Engineering 1996;52:877–82. [33] Ogunniyi DS. Castor oil: a vital industrial raw material. Bioresource Technology 2006;97:1086–91. [34] Caporiccio G, Cann PM, Spikes HA. Additives for fluorosilicone oils at high temperature. Wear 1996;193:261–8.
Contents lists available at ScienceDirect
Tribology International journal homepage: www.elsevier.com/locate/triboint
Tribological studies of potential vegetable oil-based lubricants containing environmentally friendly viscosity modifiers L.A. Quinchia a,d, M.A. Delgado a,b,n, T. Reddyhoff c, C. Gallegos a,b,d, H.A. Spikes c a
Departamento Ingeniería Química, Facultad de Ciencias Experimentales, Universidad de Huelva, Campus de “El Carmen”, 21071 Huelva, Spain Pro2TecS – Chemical Process and Product Technology Research Center, Universidad de Huelva, 21071 Huelva, Spain c Tribology Section, Department of Mechanical Engineering, Imperial College of Science, Technology and Medicine, London SW7 2AZ, UK d I&D Centre Complex Formulations, Fresenius Kabi Deutschland GmbH, Daimlerstrasse 22, 61352 Bad Homburg, Germany b
art ic l e i nf o
a b s t r a c t
Article history: Received 3 June 2013 Received in revised form 29 August 2013 Accepted 30 August 2013 Available online 7 September 2013
The amphiphilic properties that result from the fatty acid composition of vegetable oils contribute to a better lubricity and effectiveness as anti-wear compounds than mineral or synthetic lubricant oils. Despite these advantages, vegetable oils show only a limited range of viscosities and this constrains their use as suitable biolubricants in many industrial applications. For the reason, ethylene–vinyl acetate copolymer (EVA) and ethyl cellulose (EC) have been added to the vegetable oil-based lubricants studied. To address this issue, the frictional and lubricant film-forming properties of improved vegetable oilbased lubricants (high oleic sunflower (HOSO), soybean (SYO) and castor (CO) oils), blended with 4% (w/ w) of EVA and 1% (w/w) of EC, have been studied. It has been found that castor oil shows the best lubricant properties, when compared to high oleic sunflower and soybean oil, with very good filmforming properties and excellent friction and wear behaviour. This can be attributed to its hydroxyl functional group that increases both the viscosity and polarity of this vegetable oil. Regarding the effect of the viscosity modifiers studied, ethylene–vinyl acetate copolymer exerts a slight effect on lubricant film-forming properties and, thus, helps to reduce friction and wear mainly in the mixed lubrication region. Ethyl cellulose, on the other hand, was much more effective, mainly with castor oil, in improving both mixed and boundary lubrication. & 2013 Elsevier Ltd. All rights reserved.
Keywords: Vegetable oils Boundary lubrication Viscosity modifiers Biolubricants
1. Introduction Lubricants have an important role in world industrial and economic development, mainly by reducing friction and wear in mechanical contacts [1,2]. Thus, about 38 million metric tonnes per year of lubricants have been used globally in last decade, with the majority of these being petroleum-based [3,4]. In the last 25 years, there has been an increasing interest in the use of biodegradable products. This has been driven by environmental problems that have heightened the need to limit pollution from lubricants and hydraulic fluids based on mineral oils. Vegetable oils are potential substitutes for petroleum-based oils; not only they are environmentally friendly, renewable and less toxic, but also they have excellent lubricating properties such as high viscosity index, high lubricity and low volatility [5,6]. For these reasons, vegetable oil-based lubricants are being actively demanded for many green industrial activities [7,8]. n Corresponding author at: Departamento Ingeniería Química, Facultad de Ciencias Experimentales, Universidad de Huelva, Campus de “El Carmen”, 21071 Huelva, Spain. Tel.: þ 34 959219997; fax: þ34 959219983. E-mail address: [email protected] (M.A. Delgado).
0301-679X/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.triboint.2013.08.016
Vegetable oils can act as anti-wear additives and friction modifiers, due to strong interactions with the lubricated surfaces. Their amphiphilic nature gives them a good film/force relationship, due to long fatty acid chains and the presence of polar groups in the vegetable oil structure [9,10]. For this reason, vegetable oilbased lubricants have the peculiarity of being effective as both boundary and hydrodynamic lubricants [2,9,11]. However, to understand fully the tribological properties of vegetable oils, it is important to know the effects of the variability in fatty acid composition on their lubricating properties, film thickness formed, friction and wear [12]. Despite their safety, efficacy and advantages as mentioned above, vegetable oils suffer from several major drawbacks in terms of thermal and oxidative stability, which preclude their use above 120 1C, crystallisation at relatively high temperatures, and the limited range of viscosities available [13–15]. The last of these is critical since a range of lubricant viscosities are required depending on the industrial application. For instance, kinematic viscosity, at 40 1C, ranges typically from 30 mm2/s in the automotive industry, to 120 mm2/s for lubricants used on bearings, while viscosities higher than 240 mm2/s are demanded in lubricants for four-stroke engines and some gear assemblies [5,16]. In previous
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
works, the authors have investigated the use of environmentally friendly additives to improve some of these shortcomings [14,15,17,18]. More specifically, ethylene–vinyl acetate copolymer (EVA) and ethyl cellulose were tested to increase the viscosity range of different vegetable oils and improve their thermal susceptibilities [14,15,18]. However, the introduction of these additives could alter the properties required in lubricant applications; therefore, it is necessary to determine the tribological behaviour of these vegetable oil-based lubricants. The contribution of viscosity modifiers to the viscous flow behaviour of lubricants has been extensively studied and is well understood. However, their influence on the friction and wear is less clear [19,20]. This is because these additives may behave in different and complex ways in lubricated contacts, each of which might contribute to a greater or lesser extent to film formation and, thus, friction and wear [19]. Previous studies with mineral oils have shown that some viscosity modifiers are able to adsorb on to metal surfaces, improving the boundary film formation under low speed and high temperature, producing a significant reduction in friction and wear [19,21,22]. In this sense, the main objective of the research described in this paper was to determine the tribological behaviour of three potential vegetable oil-based lubricants blended with polymers and, in this way, identify links between lubricant composition, friction and wear performance to find the suitable biolubricants features for use in industrial applications.
2. Materials and methods 2.1. Vegetable oils and additives The environmentally friendly base-stocks (vegetable oils) used in this work were soybean oil (SYO), castor oil (CO), and high-oleic sunflower oil with 85% (w/w) oleic acid (HOSO). SYO was supplied by Fresenius Kabi (Germany), CO was received from Guinama (Spain) and HOSO was kindly supplied by the “Instituto de la Grasa”, CSIC (Spain). All of them were commercial grade oils without any further purification. The fatty acid compositions and physical properties of the vegetable oils are shown in Tables 1 and 2, respectively. Ethylene–vinyl acetate copolymer (EVA) with 33% vinyl acetate content (density, at 23 1C, 0.956 g cm 3; molecular weight, 60250 g mol 1; melting temperature, 59 1C), was kindly supplied, in the form of pellets, by Repsol YPF, S.A. (Spain). EVA copolymer is considered inert, nontoxic, and stable material. It is not expected to be biodegradable but is not hazardous, according to the Commission Directive 93/21/EEC [23]. The other viscosity modifier, and biodegradable material, used was ethyl cellulose (EC) (density, at 25 1C, 1.14 g cm 3; molecular weight, 68960 g mol 1; melting temperature, 155 1C). This viscosity modifier is commercially available and was obtained from Sigma Aldrich. 2.2. Preparation of environmentally friendly lubricating formulations EVA and EC were blended with the different vegetable oils used in this study, at a concentration of 4% (w/w) and 1% (w/w), respectively. Blends were prepared by stirring in an open vessel
111
(800 g), at constant velocity (150 rpm) and using an anchor impeller geometry, to disperse the polymer in the oil. EVA and EC blends were heated up to 120 1C and 155 1C, respectively, for approximately 1 h. After that, when the polymers were completely dissolved, the mixtures were returned to room temperature by natural convective cooling. A homogeneous single phase was obtained in all cases, except the blend HOSO/EC that, after some hours, became cloudy. 2.3. Viscosity and density measurements Dynamic viscosities were measured with a rotational controlled-strain rheometer (ARES, TA Instruments, USA), over a temperature range of 25–100 1C. Viscous flow tests were carried out, in a shear rate range of 5–500 s 1, using coaxial cylinders geometry (inner radius 16 mm, outer radius 17 mm, cylinder length 33.35 mm). Kinematic viscosity values, ν, were obtained as the ratio of the dynamic viscosity to the density, at each temperature. The viscosity indexes (VI) were obtained according to ASTM D-2270. A capillary densimeter, model DMA-5000 (Anton Paar, Austria), was used to measure sample densities in a temperature range of 15–100 1C. 2.4. Film thickness measurement The film forming properties of the lubricant samples were determined using ultrathin film interferometry. For this purpose, an EHL optical interferometry rig (PCS Instruments, UK) was used. A high-pressure contact is formed between the flat surface of a glass disk and a reflective steel ball. The glass disk is coated with chromium and silica layers (5 nm and 500 nm, respectively). The principle of this technique is fully described by Johnston et al. [24]. Film thickness measurements were made using a 19.5 mm diameter steel ball in nominally pure rolling contact with a coated glass disk of 100 mm diameter. Both ball and disk were ultrasonically cleaned in toluene, followed by acetone, prior to a test, and a new ball was used for each test. The load applied was 20 N, corresponding to a maximum Hertz contact pressure of 0.54 GPa. Film thickness measurements were carried out over a temperature range of 40–100 1C, and a range of entrainment speeds between 0.005 and 3.0 m/s. The refractive index of the lubricant film must be known to determine the actual film thickness. In the current work, this was measured using an Abbe 60 refractometer from Bellingham and Stanley Ltd. (UK). The measurements were made over a temperature range of 40–100 1C for each vegetable oil and its blends with the viscosity modifier. 2.5. Friction measurements The minitraction machine (MTM) technique allows determination of the “Stribeck curve”, which plays an important role in the identification of the different lubrication regimes: hydrodynamic (HL), elastohydrodynamic (EHL), mixed lubrication (ML) and boundary lubrication (BL) [25]. In this method, a rolling-sliding lubricated contact is formed between a steel ball and the flat
Table 1 Fatty acid composition of the vegetable oils studied. Vegetable oils
Palmitic (16:0)
Stearic (18:0)
Oleic (18:1)
Linoleic (18:2)
Linolenic (18:3)
Ricinoleic (18:1:OH)
Unsaturated/saturated ratio
HOSO CO SYO
3.84 2.63 11.28
4.42 1.51 2.70
83.66 4.74 24.39
8.08 8.36 56.28
– – 5.34
– 82.8 –
11.10 23.20 6.15
112
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
Table 2 Some physical properties of the vegetable oils and lubricating blends studied. Samples
HOSO HOSO þEVA 4% HOSO þEC 1% CO COþEVA 4% COþEC 1% SYO SYOþ EVA 4%
Viscosity 40 1C
Viscosity 100 1C
VI
Dynamic (mPa s)
Kinematic (mm2/s)
Dynamic (mPa s)
Kinematic (mm2/s)
36.00 151.40 45.68 230.00 403.96 518.20 31.00 137.89
38.50 168.14 50.60 242.50 427.10 546.70 33.60 174.00
8.70 25.49 13.04 19.00 46.27 36.76 7.60 30.10
9.90 29.59 15.10 21.00 50.90 40.30 8.60 27.30
surface of a polished steel disk. The disk is fully immersed in lubricant, and the temperature of the lubricant and contact are controlled to a set value within 7 0.5 1C. In these tests, the slideto-roll ratio is held constant and the friction coefficient is measured while varying the entrainment speed. The test specimens used in the MTM were a polished 19.05 mm ball and a 50.0 mm diameter disk, according to specification AISI 52100. Both had a surface root mean square roughness of 11 nm7 1 nm. The load applied was 30 N, corresponding to a maximum Hertz pressure of 0.93 GPa. Tests were carried out over the entrainment speed range of 0.005–2.5 m/s, at a fixed slide-to-roll ratio (SRR) of 0.5 (50%) and in a temperature range of 40–100 1C. A new ball was used in each test. The test rig and specimens were cleaned in the same fashion as in film thickness measurements. Friction tests were also carried out on a high frequency reciprocating rig (HFRR), supplied by PCS Instruments, London (UK). In this method, a 6.0 mm diameter steel ball is held in a chuck and loaded downwards on the flat face of a 10.0 mm diameter steel disk. The disk is held in a bath, which contains lubricant so that the contact between the ball and the disk is fully immersed. The ball is oscillated backwards and forwards in contact with the disk at a stroke length and fixed frequency [22]. Friction is monitored continuously, while wear is determined from the wear scar on the steel ball at the end of the test, by averaging the scar diameters transverse and along the rubbing direction. The test conditions used in this test are listed in Table 3. New balls and disks, cleaned successively in toluene and acetone in an ultrasonic bath, were used for each test. 2.6. Statistical analysis At least, two replicates of each test were done on fresh samples and data shown have statistically significant values, i.e. they did not exceed a significance level of 0.05 in Student's t-test and had a 95% confidence interval.
3. Results and discussion 3.1. Film thickness measurements In this current work, the ability of three vegetable oils (HOSO, CO and SYO), and their blends with two viscosity modifiers (EVA and EC), to form a lubricating film that reduce the friction between solid surfaces in contact has been evaluated. The effect of the vegetable oil-based lubricant composition, temperature and entrainment speed on the film-forming properties has also been studied. For this purpose, the optical interferometry technique has been used [26] and the refractive index of the different blends can be seen in Table 2.
257 218 314 116 183 118 250 214
Density 15 1C (g/cm3)
0.9152 0.9160 0.9192 0.9630 0.9629 0.9622 0.9256 0.9252
Refractive index 40 1C
100 1C
1.466 1.465 1.465 1.476 1.475 1.474 1.472 1.470
1.450 1.451 1.448 1.460 1.461 1.455 1.456 1.456
Table 3 HFRR test conditions. Test duration Stroke length Stroke frequency Test load Temperature Ball properties Disc properties
60 min 1 mm 50 Hz 4N 40, 100 1C AISI 52100, 800 VPN AISI 52100, 750 VPN
In addition to this, it has worth pointing out that ethylene-vinyl acetate copolymer [15] has been successfully tested as viscosity modifier for several common vegetable oils, yielding potentially environmental friendly lubricants for some applications. In the case of ethyl cellulose, the addition of 1% (w/w) of EC always yields an important increase in vegetable oil viscosity within the temperature range studied (Table 2). The most important viscosity increments correspond to CO/EC blends (i.e. 125%, at 40 1C). On the contrary, the lowest increments correspond to HOSO/EC blends, mainly at low temperature.
3.1.1. Effect of vegetable oil type and temperature on film thickness formation The film-forming process of high oleic sunflower, soybean and castor oils, at 40 and 100 1C, has been studied. As can be seen in Fig. 1a, at 40 1C, the film thickness developed by all vegetable oils studied increases linearly with entrainment speed, on a log–log plot, in agreement with the elastohydrodynamic (EHL) theory [27]. This indicates the absence of any anomalous behaviour down to 10 nm thickness, which is evidenced by a change in the slope of the film thickness–speed curves at low speeds. The flattening out of the curve for CO at very high speeds is due to thermal effects. HOSO and SYO display similar film thickness in the whole entrainment speed range assessed. For instance, at 0.1 m/s, they show film thickness values of 49.3 and 49.8 nm, respectively. However, CO, due to its high viscosity, was able to form thicker films (167.7 nm), at low temperatures, than the other neat oils studied. As can be seen in Fig. 1, over most of the speed range, film thickness displays a decrease with increasing temperature. At 100 1C (Fig. 1b), film thickness dependence on entrainment speed (u), for these three vegetable oils, is different, primarily at low entrainment speed. HOSO and CO form apparent thick boundary films (12 and 7 nm, respectively) at very low entrainment speeds. SYO does not form such a film and, indeed, at low speeds, its film thickness appears to fall below linearity on a log– log plot. In addition to this, at low speeds, film formation is not controlled by bulk viscosity, since CO, which is much more viscous than HOSO, forms a slightly thinner boundary film, while SYO, which is only marginally less viscous that HOSO, forms a much thinner film at low speeds.
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
113
Film thickness (nm)
103
102
101 HOSO CO SYO
100
10-1
10-2
10-1
100
101
10-2
Entrainment speed (m/s)
10-1
100
101
Entrainment speed (m/s)
Fig. 1. Effect of vegetable oil type and entrainment speed on the film thickness at (a) 40 1C and (b) 100 1C.
Film thickness (nm)
103
102
101
10-2
10-1
100
Entrainment speed (m/s)
101
10-2
10-1
100
101
Entrainment speed (m/s)
Fig. 2. Effect of viscosity improvers on the film thickness at (a) 40 1C and (b) 100 1C.
According to Biresaw and Bantchev [12], the film thickness in the thin film region can be influenced by physico-chemical properties and chemical composition (degree of unsaturation, functional groups and chain length) rather than viscosity. In this sense, it can be pointed out that film thickness increases when the degree of unsaturation decreases [28]; thus, SYO has high level of unsaturation and its film thickness is the lowest. However, as can be seen in Fig. 1b, HOSO formed a thicker boundary film (12 nm) than CO (7 nm) although the degree of unsaturation of both oils are similar [28]. This difference in film thickness at low speed is possibly due to both the difference in polarity [12] and the tendency of polar components to separate from the bulk and adsorb onto the rubbing surfaces as film thickness decreases [29]. Based on this statement, these results suggest that the fractionation of these vegetable oils in ultrathin films results in a deviation in the measured film thickness from that predicted from the bulk viscosity based on EHD theory [10,29]. Thus the presence of a relatively low viscosity and highly polar linolenic component in SYO may result in a thinner film at low entrainment speed.
3.1.2. Effect of viscosity improvers on film thickness formation of vegetable oil-based lubricants According to the results obtained in the last section, two of the vegetable oils (HOSO and CO), which displayed the best lubricating properties, were chosen to be blended with the viscosity modifiers studied, aiming to evaluate their tribological properties. Although the main role of these polymers is, generally, to influence the bulk rheological properties of their blends, it has been shown that some polymers are also able to form boundary lubricating films [19,20]. Fig. 2 shows the film thicknesses for the blends of HOSO and CO with 4% (w/w) EVA and 1% (w/w) EC, at two temperatures (40 and 100 1C).
Fig. 2a shows the behaviour of HOSO with both viscosity improvers at the two temperatures studied. At 40 1C, the HOSO/ EVA blend shows a linear evolution (log–log plot) of film thickness with entrainment speed; this is similar to the neat oil, indicating no boundary film formation. At high temperature (100 1C), the addition of 4% (w/w) EVA to HOSO avoids the boundary film formation seen with neat HOSO. Indeed, EVA has the effect of reducing film thickness below linearity at low entrainment speed. According to Zhu [30], EHL films can break down due to different reasons: low speed or viscosity, heavy load and roughness. Thus, the lower viscosity of HOSO and the lower solubility of EVA in this vegetable oil, in comparison to CO, may contribute to break down the EHL film at the lowest entrainment speeds. Fig. 2a also shows the behaviour of HOSO blended with 1% EC. For the tests ran at 40 1C, the cellulose has the effect of forming a very thick boundary film ( 70 nm). At 100 1C, the film formed in the EHL contact was so thick and unstable that film thickness measurement was not possible. Fig. 2b shows the behaviour of CO with both viscosity improvers at the two temperatures studied. The addition of 4% (w/w) EVA to CO has the effect of increasing low speed film thickness, which is in agreement with the formation of a viscous boundary film within the temperature range studied. At 40 1C, the lubricant forms a boundary film of around 50 nm at 0.01 m/s, while, at 100 1C, the boundary film thickness formed dropped to 15 nm. By contrast, the addition of 1% (w/w) EC to CO has only a marginal effect on the film thickness, in the whole entrainment speed range, at 40 1C, while, at 100 1C, a similar effect to the addition of 4% (w/w) EVA is noticed. The results above are consistent with the general mechanism of boundary film formation by viscosity modifier polymers, reported by Smeeth et al. [19], which suggest that polymer molecules are adsorbed on the polar solid surface (steel and silica, in this case) to form a film with a higher viscosity than that of the bulk
114
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
0.1 100ºC HOSO SYO CO
Friction coefficient
40ºC HOSO SYO CO
0.01
10-2
10-1
100
101
Entrainment speed (m/s)
Fig. 3. MTM friction coefficient of the neat vegetable oils studied at 40 1C and 100 1C.
solution. At high speed, the lubricating oil/polymer blend is dragged into the contact inlet, forming an EHD film whose thickness varies according to the viscosity at high shear rate. This is seen with all blends studied at high entrainment speed (Fig. 2), taking into account that the addition of 4% (w/w) EVA or 1% (w/w) EC yields an increase in vegetable oil viscosity. However, when the speed is decreased, in thin-film conditions, the contact inlet is filled with the adsorbed polymer layer, much more viscous than the lubricating blend, which leads to a thicker than expected boundary film. This fact is dramatically observed for the HOSO/EC blend at 40 1C. In the other cases, at very low speeds, the film is compressed to a thickness of a few nanometres, depending on the ability of the polymer to support the applied pressure and form a boundary film [20,21,31]. 3.2. Frictional properties by MTM and HFRR 3.2.1. Effect of vegetable oil nature and temperature on frictional properties Fig. 3 shows the Stribeck curves, at 40 and 100 1C, for the different vegetable oils studied. At 40 1C (Fig. 3a), the curve shows a very different lubrication behaviour for CO as compared to SYO and HOSO. SYO and HOSO show classical Stribeck’s curves, with friction coefficients falling from a high value of ca 0.06, at low entrainment speeds (and thus thin EHL films), to a low value of ca 0.023, at high entrainment speeds when the contact is in full EHD lubrication. Thus, the friction curves represent the transition from mixed boundary-EHD to full film EHD friction. However, CO has a much higher viscosity and this means that it forms a full EHD film, even at entrainment speeds as slow as 0.01 m/s. Thus, its friction coefficient curve represents full film EHD friction over its whole speed range, showing a drop-off, at high speeds, due to thermal heating of the film as the sliding speed increases with entrainment speed. It is noteworthy that the EHD friction coefficient of CO (0.035) is much higher than that for the other two vegetable oils. This must be a result of intermolecular bonding between the hydroxyl groups on the ricinoleic molecules, which are also responsible for the higher viscosity of CO. At 100 1C (Fig. 3b), quite similar Stribeck’s curves can be seen for the three vegetable oils studied, excepting the higher EHD friction coefficient for CO. EHD friction measured at high speeds is lower at 100 1C than at 40 1C. This reduction in EHD friction with an increase in temperature is a general phenomenon seen for all fluids. It can be observed that, in the temperature range studied, the Stribeck curve for HOSO shows no reduction in friction at low speeds, even though optical film thickness measurements showed
a marked boundary film. This discrepancy probably results from the remarkable sliding in the MTM contact, as opposed to the rolling optical contact and the different tribosystem used (steel/ steel contact for the MTM versus glass/steel contact in optical interferometry technique). Regarding SYO, it is noteworthy that the friction coefficient is lower than that of HOSO at intermediate speeds and 100 1C since SYO forms a thinner film than HOSO under these conditions. It may be due to some interactions with the rubbing steel surfaces of the most SYO unsaturated components to form protective boundary films under the sliding conditions of the MTM. Fig. 4 shows the friction coefficient of the different vegetable oils studied, obtained using HFRR, at 40 and 100 1C. At 40 1C (Fig. 4a), CO has the highest friction coefficient ( 0.089) and, after a short surface-conditioning phase, constant and stable behaviour over time. In contrast, SYO and HOSO friction coefficients fluctuate with time, a fact that suggests a worse lubrication, but with lower friction and wear than CO. The HFRR forms a contact that operates in boundary lubrication conditions with extensive initial solidsolid contact. This behaviour of CO, where friction falls, initially, from a high value, is indicative of the formation of a protective surface film, promoted by tribochemical processes due to the rubbing action. By contrast, SYO and HOSO appear to form protective films almost immediately, perhaps reflecting the presence of more strongly adsorbing species. The higher steady-state boundary friction of CO is consistent with its high EHD friction coefficient and suggests that the boundary film formed by this oil is dominated by the ricinoleic component. At 100 1C (Fig. 4b), CO continues to show a very stable friction coefficient with a lower value than at 40 1C. The friction coefficients of SYO and HOSO, at 100 1C, are higher and less regular than at 40 1C. The wear scars obtained by HFRR (Fig. 5) are consistent with the friction coefficients, the high wear for CO, at 40 1C, probably occurring primarily in the initial high friction part of the test. The differences observed at high temperature, in both friction and wear, reflect the strength of the boundary film formed by the oils on the surface. Those formed by SYO and HOSO are relatively weak and, thus, less able to withstand rubbing at 100 1C than at 40 1C. It is likely that the protective films formed by vegetable oils having many polyunsaturations are weak and unstable over time, and therefore less effective [32]. Castor oil appears to form a much stronger film, probably assisted by its higher viscosity and its more polar feature, it means, hydrogen bonding from the hydroxyl groups of ricinoleic molecule could be formed on the steel surface. These latest results demonstrate, once again, that the frictional properties of different vegetable oils are strongly affected by the chemical composition thereof. In this sense, the results obtained in the boundary lubrication regime show a beneficial influence of the hydroxyl group of the ricinoleic acid in the lubrication process, achieving more stable lubricating conditions with time. This functional group in CO is the main responsible factor for the intermolecular interactions favouring the high viscosity of castor oil and the relatively high EHD friction coefficient, and gives the polar character to the oil, favouring the formation of a strong boundary film, effective in friction and wear [33].
3.2.2. Effect of viscosity improver type on vegetable oil-based lubricants friction coefficient As previously discussed for film thickness results, the addition of 4% (w/w) EVA to HOSO had the effect of reducing low speed film thickness (Fig. 2a). This seems to demonstrate that no boundary film was formed with 4% (w/w) EVA. The corresponding Stribeck curves (Fig. 6a) show negligible effect on boundary friction, at very low entrainment speed, or on EHD friction, at
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
115
Friction coefficient
0.15
0.10
SYO
CO
HOSO SYO HOSO
0.05
CO
0
1000
2000
3000
4000 0
1000
2000
3000
4000
Rubbing time (s)
Rubbing time (s)
Fig. 4. HFRR friction coefficient for the neat vegetable oils studied at (a) 40 1C and (b) 100 1C.
Fig. 5. Wear scars, obtained by HFRR, for the neat vegetable oils studied.
Friction coefficient
0.1
0.01
10-2
10-1
100
Entrainment speed (m/s)
101
10-2
10-1
100
101
Entrainment speed (m/s)
Fig. 6. Effect of viscosity improvers on the MTM frictional properties for (a) high oleic sunflower oil (HOSO) and (b) castor oil (CO).
high entrainment speed. However, the addition of EVA does appear to produce a reduction in friction at intermediate, especially at 100 1C. This suggests the presence of a on the surfaces [19], which promotes separation of the surfaces and, thus, the transition from mixed to full film EHD lubrication. No such effect was seen by using optical interferometry, suggesting that the formation of this film was promoted by steel–steel rubbing in the MTM test. This is further supported by the HFRR friction values obtained, which decrease from 0.08 down to 0.06 for the EVAcontaining HOSO-based lubricant, at 100 1C (Fig. 7).
On the other hand, the presence of 1% (w/w) EC in a HOSO-based lubricant results in a very different lubrication regime distribution. The Stribeck curve for the HOSO/EC blend supports the existence of a solid-like boundary film, which exhibits significantly reduced friction at 100 1C (Fig. 6a). Thus, the results obtained at 100 1C show friction coefficients below 0.02 over the whole speed range. By contrast, as shown in Fig. 7, EC effect on the boundary friction, from HFRR tests, is to increase friction. This may result from the increased resistance to motion that occurs during ploughing of the thick solid boundary film and the action of the resulting debris.
116
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
Friction Coefficient
0.15
0.10 HOSO HOSO+EC1% HOSO+EC1% HOSO+EVA4%
HOSO
0.05 HOSO+EVA4%
Friction Coefficient
0.15 0.00
0.10
CO
CO+EVA4%
CO+EVA4%
0.05
CO
CO+EC1%
CO+EC1%
0.00
0
1000
2000
3000
Rubbing time (s)
0.0 4000 0
1000
2000
3000
4000
Rubbing time (s)
Fig. 7. Effect of viscosity improvers on the HFRR friction behaviour for high oleic sunflower oil (HOSO) and castor oil (CO) at (a) 40 1C and (b) 100 1C.
Fig. 8. Wear scars, obtained by HFRR, for blends of neat vegetable oils and the viscosity improvers studied.
The addition of EVA and EC to CO has a negligible effect on friction at 40 1C. This was expected since, at this temperature, a full EHD film is present, which is unaffected by the addition of low concentrations of polymer. At 100 1C, the addition of EVA was found to increase low speed film thickness (Fig. 2b), which is in agreement with the formation of a viscous boundary film (Fig. 6b). On the contrary, the addition of 1% (w/w) EC has a negligible effect on friction at 100 1C. Regarding the results obtained by using the HFRR technique, the friction properties were evaluated under very severe conditions, as this test operates in almost full boundary lubrication conditions rather than in mixed lubrication, as is the case for MTM tests [34]. Thus, HFRR tests measure the ability of a lubricant to form strong boundary films to withstand high asperity loads at low speeds rather than its ability to form viscous film due to enhanced entrainment [21]. According to the results obtained (Fig. 7), it can be inferred that EVA is much more effective in mixed lubrication conditions (Stribeck's curve), whilst, on the contrary, EC is more effective in extreme boundary lubrication conditions, mainly at 40 1C. The wear results support these observations (Fig. 8). Moreover, these experimental data suggest a more positive effect, on the tribological behaviour, of the addition of ethyl cellulose to CO than to HOSO. Thus, the effect, in the case of high oleic sunflower oil, is
totally different due to poor solubility of the polymer in this vegetable oil.
4. Conclusions Both castor and high oleic sunflower oils form quite thick boundary films (8 to 12 nm), which can be seen at low speeds and 100 1C, when negligible EHD film is generated. By contrast, soybean oil shows no boundary film formation, and indeed a thinner film than expected from EHD theory, yielding also much higher friction and wear. This may be attributed to its higher content in polyunsaturated fatty acids, which results in a possible thermal degradation of the boundary film. Castor oil shows better EHD film-forming properties compared to high oleic sunflower and soybean oils. Thus, castor oil also shows excellent behaviour in boundary friction and, thus, friction and wear characteristics. This enhanced lubricity of castor oil can be attributed to its hydroxyl functional group that increases the viscosity and polarity of the oil. On the other hand, it is important to note that viscosity modifiers can affect the lubricating properties of vegetable oils. In this sense, ethylene-vinyl acetate copolymer exerted a slight effect on oil filmforming properties and, thus, helps to reduce friction and wear
L.A. Quinchia et al. / Tribology International 69 (2014) 110–117
mainly in the mixed lubrication region. On the contrary, ethyl cellulose was much more effective, mainly with castor oil, in boundary lubrication as well as mixed lubrication. In addition to this, it has worth pointing auto that the most important viscosity increments correspond to CO/EC blends (i.e. 125%, at 40 1C).
Acknowledgements This work is part of two Research Projects (PSE-320100-2006-1 and PSE-420000-2008-4) sponsored by MEC. One of the authors (L.A. Quinchia) has received a Ph.D. Research Grant from FPU-MEC Programme (Spain). The authors gratefully acknowledge their financial support. References [1] Rakic R. The influence of tribological properties of lubricating oils on the reliability of gears. Industrial Lubrication and Tribology 1999;51:170–9. [2] Maleque M, Masjuki H, Sapuan S. Vegetable-based biodegradable lubricating oil additives. Industrial Lubrication and Tribology 2003;55:137–43. [3] Siniawski MT, Saniei N, Adhikari B, Doezema LA. Influence of fatty acid on the tribological performance of two vegetable-based lubricants. Journal of Synthetic Lubrication 2007;24:101–10. [4] Shashidhara YM, Jayaram SR. Vegetable oils as a potential cutting fluid – an evolution. Tribology International 2010;43:1073–81. [5] Lathi P, Mattiasson M. Green approach for the preparation of biodegradable lubricant base stock from epoxidized vegetable oil. Applied Catalysis B 2007;9:207–12. [6] Jayadas N, Nair K. Coconut oil as base oil for industrial lubricants-evaluation and modification of thermal, oxidative and low temperature properties. Tribology International 2006;39:873–8. [7] Rudnick LR. Synthetics, mineral oils, and bio-based lubricants: chemistry and technology. In: Rudnick LR, Erhan SZ, editors. Natural oils as lubricants. New York: CRC/Taylor & Francis Group; 2006. [8] Lazzeri L, Mazzoncini M, Rossi A, Balducci E, Bartolini G, Giovannelli L, et al. Biolubricants for the textile and tannery industries as an alternative to conventional mineral oils: an application experience in the Tuscany province. Industrial Crops and Products 2006;24:280–91. [9] Adhvaryu A, Erhan SZ. Tribological studies of thermally and chemically modified vegetable oils for use as environmentally friendly lubricants. Wear 2004;257:359–67. [10] Biresaw G. Elastohydrodynamic properties of seed oils. Journal of the American Oil Chemists' Society 2006;83:559–66. [11] Fox NJ, Stachowiak GW. Vegetable oil-based lubricants – a review of oxidation. Tribology International 2007;40:1035–46. [12] Biresaw G, Bantchev G. Effect of chemical structure on film-forming properties of seed oils. Journal of Synthetic Lubrication 2008;25:159–83. [13] Miles P. Synthetics versus vegetable oils: applications, options, and performance. Journal of Synthetic Lubrication 1998;15:43–52.
117
[14] Quinchia LA, Delgado MA, Valencia C, Franco JM, Gallegos C. Viscosity modification of high-oleic sunflower oil with polymeric additives for the design of new biolubricant formulations. Environmental Science and Technology 2009;43: 2060–5. [15] Quinchia LA, Delgado MA, Valencia C, Franco JM, Gallegos C. Viscosity modification of different vegetable oils with EVA copolymer for lubricant applications. Industrial Crops and Products 2010;32:607–12. [16] Mang T, Dresel W. Lubricants and Lubrication. second ed.. Germany: WileyVCH; 2001. [17] Quinchia LA, Delgado MA, Valencia C, Franco JM, Gallegos C. Natural and synthetic antioxidant additives for improving the performance of new biolubricant formulations. Journal of Agricultural and Food Chemistry 2011;59: 12917–24. [18] Quinchia LA, Delgado MA, Franco JM, Spikes HA, Gallegos C. Low-temperature flow behaviour of vegetable oil-based lubricants. Industrial Crops and Products 2012;37:383–8. [19] Smeeth M, Gunsel S, Spikes HA. The formation of viscous surface films by polymer solutions: boundary or elastohydrodynamic lubrication. Tribology Transactions 1996;39:720–5. [20] Müller M, Topolovec-Miklozic K, Dardin A, Spikes HA. The design of boundary film-forming PMA viscosity modifiers. Tribology Transactions 2006;49: 225–32. [21] Müller M, Fan J, Spikes HA. Design of functionalized PAMA viscosity modifiers to reduce friction and wear in lubricating oils. Journal of ASTM International 2007;4:1–10. [22] Fan J, Müller M, Stöhr T, Spikes HA. Reduction of friction by functionalised viscosity index improvers. Tribology Letters 2007;28:287–98. [23] Commission Directive 93/21/EEC of 27 April 1993 adapting to technical progress for the 18th time Council Directive 67/548/ EEC on the approximation of the laws, regulations and administrative provisions relating to the classification, packaging and labelling of dangerous substances, Off J. E.C. L110, 1993. [24] Johnston GJ, Wayte R, Spikes HA. The measurement and study of very thin lubricant films in concentrated contacts. Tribology Transactions 1991;34: 187–94. [25] Moshkovich A, Perfilyev V, Lapsker I, Rapoport L. Stribeck curve under friction of copper samples in the steady friction state. Tribology Letters 2010;37: 645–53. [26] Spikes HA, Cann PM. Techniques for studying the behaviour lubricants in elastohydrodynamic contacts. Acta Tribologyca 1992;1:33–48. [27] Anghel V, Bovington C, Spikes HA. Thick-boundary-film formation by friction modifier additives. Lubrication Science 1999;11:313–35. [28] Graciani E. Los aceites y grasa. Composición y propiedades. first ed. Madrid; A Madrid Vicente : Mundi-Prensa: 2006. [29] Guangteng G, Spikes HA. Fractionation of liquid lubricants at solid surfaces. Wear 1996;200:336–45. [30] Zhu D. On the EHL film breakdown. Tribological Research and Design for Engineering Systems 2003;41:559–71. [31] Fox NJ, Tyrer B, Stachowiak GW. Boundary lubrication performance of free fatty acids in sunflower oil. Tribology Letters 2004;16:275–81. [32] Asadauskas S, Perez JM, Duda JL. Oxidative stability and antiwear properties of high oleic vegetable oils. Lubrication Engineering 1996;52:877–82. [33] Ogunniyi DS. Castor oil: a vital industrial raw material. Bioresource Technology 2006;97:1086–91. [34] Caporiccio G, Cann PM, Spikes HA. Additives for fluorosilicone oils at high temperature. Wear 1996;193:261–8.