Vegetable- Based Oil as a Gear Lubricant
IBmis IKrzan and Jloief Vi.zintin
Summary to be decomposed by the action of bacteria into Universal tractor transmission oil (UTI1Q) i C02, water, mineral compounds and bacterial rnulrifunctional tractor oil formulated for use in bodies. Biodegradability is influenced by numer- transmissions, final drives. differentials, wet brakes, ous factors, of which the main are the molecular and hydraulic systems of farm tractors employing a structure and the chemical properties of organic common oil reservoir. In the present work, the gear compounds and the environmental conditions of protection properties of two formulated vegetable- biodegradability, such as the presence of oxygen, based uno oils. one synthetic ester-based UTIO the possible level of nutrition and the pH (Ref. I). oil,. one synthetic ester gear oil and one mineral- Vegetable oils and synthetic esters are the most based UTIO oil are investigated. common base stocks for biodegradable lubricants. The data, presented. in this paper, have demon- Synthetic oils represent a fairly recent. development strated thai the formulated vegetable-based in the lubrication market. They can be made by UTIO oils nave high lubricity, high vi cosily reacting alcohols with fatty acid. Synthetic oils indexes and provide equivalent Of-in some offer improved performance compared with all aspects-superior gear protection performance other lubricants, but at a price. Both vegetable oils compared with the mineral-ba ed UTTO oil. The and synthetic esters are highly biodegradable and high-oleic sunflower oil formulation derived readily available, but vegetable oils occur naturally, from the genetically modified plant ha shown have a "greener" image, and are, in general, three better results than the rape eed-ba ed (canola- times cheaper. Properly balanced additives can based) oil formulation. compensate for low temperature performance and Introduction oxidative stability of the vegetable oils and favor It is generally recognized that mineral oil lubri- them as the base stock of choice (Refs. 2-3), cants represent a potential danger in many applica- A comparison of the simplified chemical struc- tions because they are not readily biodegradable tures of mineral and vegetable oils shows great and are toxic. The need for biodegradable and non- similarities. The major difference is that vegetable toxic lubricants has been recognized especially in oil is an ester, while mineral oil is a hydrocarbon. the areas where they can come in contact with soil, The presence ofthe polar ester group impacts ev- ground water, and crops. eral propertie , making vegetable oil better than Biodegradability is the ability of a ubstance mineral oil in reducing friction and wear. The ia'ble1-iest Oils. J polar group also makes vegetable oil a better 01- Viscosity Imm2ls1 I· vent for sludge and dirt, which would be other-
Base stock Oiltype v4ll"C vlI)(),c VI Oil code ,I wise deposited Oil the surfaces being lubricated. Rapeseed oil biodegradable I 48.8 10.4 209 R Because of these properties, it maybe pas ihleto uno I High-oleic sunflower oil biodegradable uno 51.4 10.6 2.03 S I reduce the amount of friction modifiers. antiwear Synthetic ester biodegradable gear oil 101 17.8 195 G I agents. and dispersants required to formulate veg- Synthetic ester biodegradable uno 51.3 10.9 211 H I etable oil-based lubricants. Mineral oil UTTO 55.1 9.2 150 M The agricultural equipment is ideally suited to use vegetable oil-based lubricants, because the ilihle 2-Fatly Acid Content In Test Vegetable Base Stocks. equipment operates dose to the environment The Fatty acid content (%) opportunity exists to create a continuous cyc\ein Base stock Palmitic I Stearic Oleic Linoleic Linolenic OtllBr I which the agricultural equipment is lubricated by C 16:0 C 18:0 C 18:1 C 18:2 C 18:3 High·oleic sunflower oil 4.7 3.7 72.6 17.0 I 2.0 the oil from a plant growing in the field being cul- !Rapeseedoil 6.1 2.5 49.1 32.2 6.9 3.2 tivated! by that same equipment (Ref. 2). Universall tractor transmission oj] (urrO) i C X:.Y .... fatty acid chain of length X and containing Y double bonds; e.g. C 18:3 is an 18 multipurpose nil widely used for agricultural, COIl- carbon- chain fatly acid with three double bonds. struction and other off-road vehicles. UTTO oil i 28 JULY/AUGUST 2003- GEAR:TIECHNOLOGiY - www ..gea.rfech.nology.com. wWW.pOWfUlra.nsmission'.com pecially designed for lubricating the transmi - sions, final drives. wet brakes and hydraulic sys- 0.35 terns employing a common oil re ervoir. UTIO E E 0.3 oil has to meet some specific reqniremems to ~ operatein agricultural and construction equip- I: 0.25 .~co 'Vi, ment. The oil must provide the correct frictional co 0.2 <>. balance to' prevent wet brake chatter and to allow ..E 0.15 u smooth transmission clutch engagement, S, 0.1 At tile same time, the oil must provide 'enough <= '"E D.!05 clutch capacity for ,efficient power transrni '01'1 and w'" enough brake capacity to stop the tractor ina reason- 0 A S G H M able lime and distance. urro oil mu t also provide Oil ufficient antiwear (AW) and extreme pressure (EP) Figure I-Elemental ,analysisof oils. properties for tile whole transmi ion y tern. espe- cia1]y for the piral bevel ring and pinion gears in the OJ5
, axle , The AWIEP additives must not be so active as t: co 0.12 to cause corro ion in the tractor's hydraulic sy tern. ..".., :E 0.09 where pump containing alloy of copper can be "0 present (Refs. 2 and 4). t: .!- u 0,06 Rape eed and uoflewer oil are currently :5., 00 used in Europe for the fonnulation of the lUI 0.03; biodegradable lubricants. In the present work. the 0' +-...... ,._...... JL...-__ antiwear and extreme pressure propertie of for- R S G H M mulated rapeseed and high-oleicsunflower-based Oil VITO oils, synthetic ester-based UTTO oil, syn- Hartsian Pressure, 2.0 GPa;.Frequency, 50 Hz; Amplitude,. 1,'00'011m; thetic ester gear oi], and mineral-based VITO oil Speed, 0.1 mrs; Iemparature, 50°C; Duration. 120 min. are investigated on FZG test equipment. Tile .FiguTie 2-.FFiction coeffic.ient meafl"'al'14es,, selected formulated vegetable-based uno oil content of oleic acid than rapeseed oil Due to the wa further te led on the helical gear res trig, higher saturation, the high-oleic sunflower oil ha Sample Preparation better oxidation lability thanthe rape eed oil. Oil samples. We have formulated two differ- ElemenJaI anaJysiSo! ,additive. Spectrometry ent vegetable-based UITO oil for the investiga- via ED-XRF (energy disperse X-ray fluore • B'oris Krian tion . The fmt fortnulation is ba ed on tile rape- cence) has been u ed to obtain the elemental is ,an assistant researcher in eed ba e lock, while the second ba e tock is composition of additives for the te t oil (see th» mechanical eliginur;llg ftu:'ltlty at the University oj derived from a genetically modified ul'1lflower Figure n. The elemental composition of additive LjubljaIla, Slovenia. He also plant with a high oleic content. The arne additive is quite similar for the formulated vegetable- IWJrb ill ,lte university's system is u ed for both formulation . The proper- based UTIO oil .R and S and the reference syn- CI'IIrrl'for TFibolagy and lies of these two fully formulated vegetable-based thetic e ter-based UITO oil labeled H. The min- Technical Diagnostics. HiI research involves lubrication, VITO oils were compared with a commercially eral UlTO oil labeled M contains a significantly ",ilh special interest ill lItt'd available mineral-based UITO od,a fast higher level of calcium. and ulfur than any ether oil Q/wlysis. /ermgmphy and biodegradable synthetic-based UTIO oil and a lest oil. The syntheti.c ester-based gear oil labeled biodl'grntiable. vegetable- based lubricants. . ynlhetk ester-based gear oil (see Table 1). G shows a low amount of additive concentration The test' TIO oils have a kinematk vi cosi- compared. with the UITO test oils. Dr: Jozef V1ifintin ly between 9' mm2/s and ~] mm2/ . at ]OO°C. This Preselection Experiments and Test Re uUs i a prole sor ill (he vi cosily offers ufflciem thickne s to promote SRV test results. The coefficient of friction U"iw'rsily oJ Ljubljallo's mechanical engineering fac- good gear protection and is tin suitable for 'the measurements have been performed on an SRV ult» and is head of the Celli" hydraulic sy tern. The synthetice ter G has a vis- high frequency test device. SRV tands for the for Tribolog)' and Technical co it)' two ISO grades higher than other test oils German words "Schwingung" (oscillation), Diagnostics. His I1IQjor research' interests art' I'tg- and i mtable a a gear oil onJ:y. "Reibung" (friction) and "versehleiss" (wear), etable-based lubricants, ....!'Or The main difference between the vegetable The device produce linear 0 cillatiog motion of resistanc« 01advanced mate base rocks for R and S fonnulatioa lies in fa~ty a ban on a flat pecimen under boundary lubri- rial, and diagnostics alld prediction a/wear failllPl's ill acids contem (see Table 2). The high-oleic un- eating conditions. A thin layer of lubricant is mechanical sy Il'rns. For his flower base lock is, derived from a genetically spread over 'the flat specimen before eachtest. On doaorat», V;tilllin studi d altered plant and possesses a sigl1ificantly higher the SRV test rig. just the friction coefficient, at p01l'l'r losses in gears. WWIW.IHI'wfulrsflsm;ss;o'fI,,/:om. www.geIT.tednology ..com • GEAR TiECHNOLOG,V • JULY/AUGUST 2003 2'9 The load-carrying capacity of lubricants was t.4 investigated by using the standard FliG A/8.3/90 E 1.2 test procedure. The test oil is subjected to a load.. !. ,~ increasing by stages, until the scuffing failure cri- .2l m E, O.B terion bas been reached. Twenty mini meters of ~'" tooth scuffing indicate a test failure. The failure ''- 0.6 0 load stage is reported as a result (Ref. 6-7). '\0'" '- 0.4 Investigations of the pitting resistance were '"II> ~ 0.2 performed on the FliG gear test rig in the standard
a Oleic Stearic Palmitic Linoleic 'linolenic pitting test CI8.3/90. After a two-hour run-in at fatty Acids load stage 6 (135..3 Nm), the test is run at load Hartzian Pressure, 3.17 .GPa; FrequencY,s 50Hz; A,mplitude,.1 ,000 11m; Speed, 0. I m/s; Temperature, SO'C; Duration, 120 min. stage 9 (302 Nm) until the failure criterion is Figure 3-Antiwea,. propertief o/fatty acids. recorded. The number of pinion lead cycles when the critical damage of the tooth flanks occur is reported as a result (Ref 8). UTIU oils are intended to Jubricatetransmis- sions and gearboxes of the tractors ..In such sys- tems, high temperatures, high loads and low speeds are very common conditions. Tile primary mode of failure observed witb the spiral bevel gearing is scuffing, while the planetary units encounter normal abrasive wear. There are a num- ber of methods to evaluate scuffing, but the pri- mary concern of this investigation is to simulate (l) test ,pinion 141brace coupling m torque maasu ring coupling HOI shaft 1 !2) lest gear 151locking pin (8)temperature sensor 1111shaft 2 normal.rubbing wear of the planetary gears. In the (3) slave gearbox (61load lever and weights (91tast gearbox 021 electric mater slow-speed, high-lead wear test, the C-type gears Figure 4-Schematic section of.tlle FZG K,eartest rig. were used to reduce the sliding velocity and con- sliding motion was measured, The test rig config- sequently the probability of scuffing. The test pro- uration and the test specimens are described in cedure is divided into two stages. The test gears DIN 51 834 T2 (Ref. 5). are weighed before and after each stage and the FIgure 2 shews the results of friction coeffi- weight loss associated with wear is rec-orded as a cient measurement of the test oils. The mineral- result which indicates the lubricant antiwear per- based UTTO oil M shows the highest value of formance (Ref. 9-10). average friction coefficient. The biodegradable The main FZG test conditions are summarized oils exhibit less friction, especiallythe test oil G. in Table 3. Biodegradable UTIO oils with similar elemental, The results of the FZG tests are summarized in compositions of additives-R, S, and H-obtain Table 4. The best results on the FZG test rig were almost the same value for friction coefficient. obtained with the synthetic ester-based oil G, five different fatty acids contained in the veg- which has a viscosity that's two ISO grades high- etable oils were tested OIl the SRV test device to er than other oils. demonstrate their antiwear properties (see Fig. 3). The results of scuffing load capacity for UITO The oleic fatty acid proved to have the best anti- oils show better scuffing resistance for the formu- wear properties at the selected parameters, lated. high-oleic sunflower-based oil S, which Antiwear properties and good oxidation stability passed the 11th load stage while the other veg- make the oleic fatty acid the most desired fatty etable-based formulation R and reference UITO acid in the vegetable lubricant oil formulation. oils passed the 10th load stage. Generally, UTTO FZG test results. The FZG gear test rig is oils exhibit a scuffing load stage between 9 and commonly used to evaluate scuffing load capaci- 11; therefore all tested oils meet the e require- ty, pitting resistance and slow-speed, high-load ments (Ref. 11). wear resistance. Experiments are based on a fail- The results of pitting investigations show supe- ure of a standard gear set, lubricated with thetest nor pitting resistance of the vegetable- and the oil under specific test conditions, using the test synthetic esters-based oils, compared with the rig illustrated in Figure 4 (Refs. 6-7). mineral-based UTTO eu The high-oleie sun- 3D JULY/AUGUST 2003 • GEAR TECH:NOUI'GY • www.geaf.re.c.hnol.ogy.com.www;fJowertransmission.com flower oil S showed very good pitting performance Table, 3:--fZG Test Conditions. among the formulated UTIO oil . Parameters, u~ Scuffing Pitting ~~_Wear The results of slow-speed, high-load wear inves- Stage I Stage II tigations indicate no significant difference in wear I Test gears type A c
1111.08 d stage 9 HI rates among the TIO oils. All te I oils show low " .. 302 37.2.6 372.6 wear rate in a low-speed. high-load fZO test Circumferential speed m/s 8.3 8.3 0.35 0.20 On the basis of the .fZG test results, we select- Pinion rotational speed rpm 2,170 2,170 '93 53 ed the formulated high-oleic sunflower-based for- Bunning time hour 1/.,-one stage until failure 20 30 mutation S for further investigations, Sump temperature 'C 90. at start 90 121 Uelicaw Gear 'Test ..... incrementally increased toad. Helical gear test rig. A helical gear lest rig was Table 4--HG Tes~Results. used in order to demonstrate visco ity lability, anti- Oil Sc uffing Pitting Slow speed wear wear properties, oxidatienresi ranee and seaJ compat- A/B.3/90 C/B.3!90 C/0.35-0.25/120 ibility of the formulated high-oleic sunflower-based . _ (FZG load staOfl'1 ~ ~ycles 01 pinianl Iwe.ight loss in mgl R 10 13.96 11(16 UT1U oil labeled S in contmlledlaborarory condi- s 111 26.75 1'()6 11 tions. Through periodic sampling of 'the lubricant and 19/22 G :>12 30.00" 1,(16 used oil analysis, the condition of thetestoil and parts 1'0 l'5.1fi61OS ofithe gearbox were detennined. ------M HI 7.70 lOS The helical gear test rig is schematically shown * pining, test. was stopped, but critical failure did not occur. on Figure 5.. The AC drive motor with frequency 1) Stage IIStage III. regulation runs a test gear-unit which is lubricated with the formulated high-oleic sumflower-based tJTI10 oil S. For load simulation, as a brake. the DC generator and electric heaters are 1.1 ed, The Dm CK60 pinion and DIN CK45 gear, ea e hard- ened to 60-62 HRC and uot undercut, were used electrometer test gearbox. generater electric brak,e during the test. The e helical gears had face widths of 30 mm, normal modules of 2.5 mm, and the .FigllJ';eS-ScI,emalic d4JgrallJ oj the helical gear lest rig. drive pinion had 31 teeth meshing in a 1:L5 rari o, I 5 The test rig was run continuously [2 hours per day ;;; '50 L_ - - ...... IF'" at a constant load of approximately 60' Nm of ~ 1 .- - - .- .- 4 .§. - torque. The oil temperature was maintained in the 40 IKinematic viscosity .1 u 'CO range of' 78-82°C. A pair of helical gears was :I: 3 ,0-- rotated until the lubricant deteriorated. i 30 / 2 .~.., OJ Oxidation of lubricants is normally measured e .s u ~ 2 z .20 -. I: TAN by total acid number (TAN) and viscosity increase. 'S t- - r .s u .z::> • -- New oils have an initial. TAN, therefore the -l- E 10 ~ '" 1111" increase over the initial value measures oxidation. c: II :;;;:'" i i 0 If the TAN exceeds 2.0 mg KOHIg over the origi- 0 - nal value, the oil shouldbe changed. (TIle unit "mg a 150 300 450 !600 750 900 Operating hours KOHIg" is the quantity in milligrams of potassium hydroxide (KOB) needed to neutralize the acid Figure fi...-..Trelld values for viscosity and TAN, constituents in one gram of lubrication oil) A by a table value until the sudden rise at 650 oper- strong indicator of oil degradation is also its ating hours, which indicates the oil deterioration. increase in viscosity. Normally a 20% increa e in ,Gearbox condition. Oi1in the gearbox couJd be viscosityis a warning that the oil. is reaching the a very u eful condition monitoring media. If we end of its useful life, can separate the debris from the oil, we can iden- Oil cOlJditiolJ. The top line on 'the graph in tify and. track an abnormal wear condition without Figure 6 represents kinematic viscosity oflhe tearing down the equipment. Wear particles con- high-oleic sunflower-based UITO oil S, measured tained in the lubricating oil carry detailed and at 40°C. ArLer initial shear-down, the viscosity is important information about the condition of the table until. 600 working, hours. when a slight oil-wetted cemponemsin the gearbox. If noe ces- increase is observed. The bottom TAN line shows ive wear is observed, then this indicates that the three distinct ection: initial increa e i .followed effective lubrication in the gearbox i maintained Iwww.p.oweruaRsm;ss;on.com .. w.ww.gear.ff ...;hnol.ogr.com. GEARi TECHNOLOGY. JULY/AUGUST 2003 3'1 2000 count. In normal operating conditions, a baseline ... of normal wear may be established and the aver- 1600 -- 1.- DL O-DS r--i age of WPC values can be calculated. I The average of Wear Particle Concentration , ! til 1200 • • value (WPC",ea,,) 0 I I J • WPC",ean l/n2;WPC* (3) D = 800 r. • • • •• li- I oj 0 • where WPC* means that only normal wear data 0 t.,,(') • () ~ are summed (outliers are excluded) . 400 v -v-v u' 9 The WPC value should not exceed the value I ( 'f! of an alarm limit-the critical wear particle con- 0 I 0 150 300 450 600 750 900 centration that is based on the WPC"'.1l1Ivalue and I Operating hours standard deviation of the normal samples (out- Figur,e 7-Qltantitative Jerrography reading». liers are excluded). The critical Wear Particle Concentration (WPCc,J 10,000 100 WPCcr = WPCmean + zo (4) WPCcr where (J is a population's standard deviation. '. WPC '" PLP J= 80 The most informative method for DR results .... representation is a plot of WPC and PLP in the jft•..- 60 .... ~ CL. same graph, because an increase in both WPC • • ...... J • •• CL. 40 and PLP is the best indication of an abnormal I wear condition. Figure 8 shows (he calculated .., - - 20 values for WPC and PLP, which are plotted over I-ot Running-in wear .....---- Normal wear ------. I time. The WPC value shows an initial sharp rise I lOa I o through a running-in process, during which the o 150 300 450 600 750 900 quantity of wear particles quickly increases and Operating hours then settles to a lower value after 350. operating Figl,re 8-Tr:elld values Jor WPC and PLP. hours, when a normal wear period is beginning. during the operation. The WPC and PLP values remain relatively con- The method used for the quantitative evalua- stant in the normal running operation period, tion of the wear particle concentration was direct because the gearbox wear reaches a state of equi- reading (DR) ferrography. DR ferrography mag- librium in which the particle loss rate equals the netically separates wear particles from lubricants particle production rate. and optically measures the relative concentration An alarm limit for severe wear WPCcr can be of ferrous particles present in the oil sample. The calculated for the last six samples, because at instrument is able to detect particles in the length least three data in the normal wear period are nec- range of 1-300 microns. The output of the DR fer- essary to determine WPCmean and a population's rography consists of two digital readings, a DL standard deviation cr before the WPCcr for the (density large) for large particles (> 5 um) and a first point can be calculated. All WPC values are DS (density small) for small particles (1-2 um) beyond the alarm limit. therefore the gearbox (Ref. 12). operates in the normal wear mode. The test was Figure 7 shows the measured values for DL ended when the test oil started to deteriorate and and OS for high-oleic sunflower-based oil S as a before the gear failures occurred. function of operating time. These readings can be To avoid leakage problems. seals used in a processed in several ways to allow easier identifi- gearbox should be compatible with the test oil, cation of an abnormal wear mode. Two such ways Fluoroelastomer (Viton®)was used as a seal mate- are briefly described (Ref. 12): rial for the test, The seals were inspected after the Total Wear Particle Concentration (WPC) test and no change in the geometry was found. WPC=DL+DS (1) Discussion Percentage of Large Particles (PLP) Two formulated vegetable-based UTIO oils, PLP= [(DL-DS)IWPC]*100 (2) two synthetic esters, and a mineral UTIO oil Although the magnitude of the WPC is impor- were investigated with respect to their gear pro- tant, the change from historical values is the iadi- tection properties. The study has shown. that the cator of machine wear condition. Quantitative fer- gear protection properties of the formulated veg- rography is a trending tool and not a particle etable-based oils are comparable with the corre- 3.2 JULY/AUGUST 2003' GEAR lU;IINOL'OGY • www.gslIrtechnology.com. www.powertr.ansmi.ss;on.com spending mineral-based oil, However. the FZG Summary investigations show significantly better results The vegetable-based lJITO oil formulations have for pitting re i lance for the vegetable-ba ed oils the advantages of having a green ouree of oil and and synLl1eti.c e ters. lower cost than a biodegradable ynthetic UTfO oils. The vegetable-based oils and synthetic esters Agricultural equipment. is ideally uited to lise veg- exhibit very good lubricity in boundarylubrica- etable-based oil because the tractor is lubricated by oi I tion conditions. because the synthetic esters con- derived from a plant growing in 'the field that has been Lain organic straight chain compounds with polar cultivated by tile arne equipment. end groups=fauy acid (see Table 2). The polar The vegetable-based oil formulations exhibit nature of the biodegradable oils gives them a [ow changes of viscosity with temperature, The greater affinity for metal surfaces than nonpolar vi co it)' index improvers can be used to enhance mineral oil . The need for anti wear additives is the viscosity index of mineral-based oil • but it is reduced. Therefore, with lower concentration of advantageous to have a high viscosity index "buill. the additive , both the vegetable- arid synthetic- into" the base oil molecule itself. ba ed oil how lower fri.cLion coefficients than The te ts how that the pitting resistance of the mineral test oil (see Figs. 1-2). vegetable- and synthetic ester-based UTIO oil The FZG Lest results indicate that oil viscosity formulations is significantly better compared with has 3. strong influence on cuffing performance the mineral UITO oil. and pitting resistance (Ref. 13). The best results The investigations on the helical gear te t rig are obtained with the ynthetic-ba ed gear oil, have shown that the test high-oleic uaflower- which has a vi cosity two ISO grades higher than based UITO oil formulation derived from the the UTTO te t oils. At the arne time. this syn- genetically modified plant. provides ufflcient thetic ester has the lowe t concentration of addi- gearbox lubricationfor 600 operation hours at. a tive . The gear oils are generally of higher vi - maintained oil temperature in the range of cosity than hydraulic oils. but UITO oil is a mul- 78-82°C·O tipurpose lubricant that has 10 meet boththe gear R Ierences protection and hydraulic system requirements. I, Kabuja, 1\.. and J.I... Bozet, "Comparative Analysis of the Lubricating Power Between (l Pure MUlCraI Oil and Biodegmdable Vegetable- and synthetic-based oil have Oils oflhe Same Mean tSOGrnde." whrirum, and uwrirnrwll. 1995. pp.25-3O. excellent viscosity propertie . Their viscosity 2. Gapin ki, RE,. LE. Joseph, and B,D, Layzell. "A Vegetable Oil indexes (VI) exceed 1.95. while the V] for miner- Bused Tractor Lubricant." lntemational Off-Highway <.I: Power plsnu Congress & Exposit/OIl. Milwaukee, WI, U,S.A, September 12-14. aI UITO cil equals ISO (see Table I),. The high- 1994. This Ipaper was pr,esented 3. Ooyan R.L, R.E. Melley. P.A. Wis.,ncr. w.e. (lng, "Biodo;gmdable er VI allow- for the formation of the thicker LubricaJ'llS," WmcrHi