University of East London Institutional Repository:

Transcription

1 Univesity of East London Institutional Repositoy: This pape is made available online in accodance with publishe policies. Please scoll down to view the document itself. Please efe to the epositoy ecod fo this item and ou policy infomation available fom the epositoy home page fo futhe infomation. Autho(s): Rigol, Sascha; Schenk, Michael; Peyman, Roy; Hosny, Wada M. Title: Challenges in the onboad oil condition monitoing Yea of publication: 2009 Citation: Rigol, S., Schenk, M., Peyman, R., Hosny, W.M. (2009) Challenges in the onboad oil condition monitoing Poceedings of Advances in Computing and Technology, (AC&T) The School of Computing and Technology 4th Annual Confeence, Univesity of East London, pp Link to published vesion:

2 26 CHALLENGES IN THE ON BOARD OIL CONDITION MONITORING Sascha Rigol, Michael Schenk, Roy Peyman, Wada M Hosny Univesity of East London, London/UK; Daimle AG, Stuttgat/Gemany.peyman@uel.ac.uk, w.m.hosny@uel.ac.uk, sascha.igol@daimle.com, michael.schenk@daimle.com Abstact: Engine oil is an essential component in moden engines and must fulfil a boad ange of duties. The engine oil condition is closely elated to the engine pefomance and to the engine eliability. Diffeent effects duing the engine opeation cause the fomation of contaminants eaching into the oil and degading the oil condition. Monitoing the oil condition with an on boad senso could povide infomation about the cuent oil status. The available senso data and thei coelation to oil contamination effects is futhe illustated and examined. The majo challenge to deive a eliable oil condition statement fom these data is also discussed in this pape. 1. Intoduction: The aim of the automobile industy to contibute to the eduction in global waming gas emissions such as cabon dioxide (CO2) has diven the adoption of biofuels and the implementation of complex exhaust gas afteteatment systems. The combustion of conventional fuel causes the fomation of contaminants esulting in engine oil degadation. The application of biofuels, especially biodiesel, causes a significantly wose deteioation of the oil condition due to the diffeent chemical popeties. Cuent exhaust gas afteteatment systems wosen the engine oil contamination although this educes the emissions. The eason fo this is that the afteteatment equies the engine to povide some unbunt fuel in the exhaust gases causing inceased oil contamination due to condensed fuel tavelling down the cylinde walls into the sump. The application of an on boad oil condition senso would be helpful in poviding infomation about the cuent oil condition in ode to maintain a eliable dive opeation. The coelation between the majo oil contamination effects and available senso data is consideed in this pape using data fom the Daimle intenal oil data base. This data is examined data ae fom a 4 cylinde diesel engine. The final discussion in this pape outlines the main challenge of oil condition monitoing. 2. Oil condition measuing pinciples: Oil condition paametes obtained fom cuent senso technology ae in tems of the physical oil condition eithe the kinematic o dynamic viscosity (Mölle, 2001) and in tems of the chemical elated oil condition paamete of elative pemittivity ε. The following oveview of the applied techniques is theefoe only based on the viscosity and the elative pemittivity to detect physical and chemical oil condition vaiations.

3 Physical oil condition measuing pinciples: The methods of viscosity measuement ae based on the detection of oil flow popeties with the help of diffeent physical pinciples. The main pinciple of cuent sensos is the eaction of a cystalline quatz based electomechanical system to changes in the viscosity popeties of the suounding oil. Most of the concepts use cystalline quatz based SAW (suface acoustic wave) o the tuning fok flexual esonato technique due to pactical difficulties in applying sensos in vehicles, (Dobinski et al., 2007). The basic theoy behind this measuing pinciple is the viscosity dependent attenuation of the initial oscillation of the quatz esonato, (Matsiev, 1998). Othe appoaches ae based on the physical pinciple of heat/ themal conduction using a cylindical pipe and heate configuation, (Lin, 2004). The theoy behind this pinciple using themal conduction is vey simila to the diffusion mechanism and can be deived fom Fick s laws. The diffusion coefficient in liquids is invesely popotional to the viscosity, but inceases with tempeatue, (Foth, 2005). Anothe possibility to detemine the viscosity is the measuing of fluid dynamics with the help of a steamed cylindical capacito configuation. This system is theoetically based on the fluid dynamics pinciple accoding to the Hagen Poiseulle law whee the viscosity is by definition one vaiable in the equation, (Foth, 2005) Chemical oil condition measuing pinciple: The on boad measuing of chemical oil popeties is mainly focused on the detection of polaisation effects with a capacitive element. The electodes ae sepaated by the engine oil as a dielectic fluid. The dielectic becomes polaised as an electic field E is applied. This esults in the induction of dipoles fom non pola molecules o in the oientation of dipoles fom pola molecules in the dielectic substance. The oientation of existing o induced dipoles in the opposite diection to the applied altenating electic field esults in a change of the dielectic chaacteistic of the senso aangement. This change can be expessed and quantified by the elative pemittivity ε. This value is usually displayed by the senso and is a complex quantity: ' '' ε = ε jε (2.1) The eal pat of pemittivity ε is a measue of how much enegy fom an extenal electic field due to the polaisation effects is stoed in the dielectic mateial, i.e. the engine oil. The imaginay pat of pemittivity ε is called the loss facto and is a measue of how dissipative a mateial is to an extenally applied electic field, (Agilent, 2006). The dissipation accompanying the polaisation effects will contibute to the dielectic losses. The elative pemittivity ε of fesh engine oil is usually known and is typically in the ange of The senso evaluation of the elative pemittivity of oil theefoe povides a diect measue of the inceased concentation of substances fom any contamination effect contibuting to the oveall polaisation of the dielectic. 3. Contamination effects and thei influence on selected oil condition data: The degadation of pimay oil is coelated to the oil condition data of viscosity and pemittivity to illustate existing inteelations with possible senso data. The selected datasets cove all available dive and test pofiles as each pofile contibutes a

4 28 diffeent loading of contaminants that affect the oil. It should be noted that some geneated data sets ae fom test uns that exceed nomal diving applications. The appoved SAE 5W 30 engine oil fo this engine type is used in these plots. This should povide a ealistic summay of the complete bandwidth of available load pofiles coveed by the plots pesented hee. The tade off of the selected spectum has caused the lage stay aea of the scatteplots. The equation fo the linea egession and the coefficient of detemination R² in each plot should theefoe povide a lead to indicate the tendency of the coelation Effect of fuel dilution on viscosity and pemittivity: A cetain amount of fuel can each the thin film of oil on the cylinde wall pio to combustion fo an intenal combustion engine duing nomal opeation. The amount of fuel eaching the oil depends on the tempeatue, the exhaust afte teatment stategy and the load paametes. Cold tempeatues of the cylinde wall will cause condensation of the fuel and thus esult in highe fuel dilution, (Mölle, 2001). New emission equiements fo diesel engines have esulted in the adoption of afteteatment devices including diesel paticulate filtes (DPFs). Paticulate filtes accumulate paticulate matte (PM) and peiodically the filtes must be egeneated by oxidizing the PM at elevated tempeatues. Late injections of fuel, and paticulaly late in cylinde injections cause unbuned hydocabons fo DPF egeneation beaing a high isk of impingement of liquid fuel on the sufaces of the combustion chambe. Fuel may ultimately be tanspoted down the walls and into the oil pan. Fuel dilution of oil esults fistly in a eduction of oil viscosity and inceases the isk of scuffing on the cylinde walls o beaings. The kinematic viscosity of an SAE 30 oil at 100 C is between mm²/s. The viscosity of the evised fuels as discussed, is significantly lowe and is at tempeatues of 40 C in the ange of mm²/s accoding to cuent standads. The oil tempeatue coelates to the opeational, i.e. load pofile of the engine. Typical boiling anges of the type of fuel eaching into the oil ae: (bio)ethanol 78.5 C (boiling end point), standad petol 40 C 200 C, standad diesel 180 C 380 C and biodiesel 325 C 355 C, (Luthe, 2008). The oil tempeatue in the sump of passenge cas can each 100 C 160 C, (Reinhadt et al., 1992). Fuel dilution is thus less poblematic fo ethanol and petol but is sevee fo diesel and especially fo biodiesel. With highe oil tempeatues then the low boiling factions of the diesel evapoate, but biodiesel emains due to the high boiling tempeatue of the oil. This accumulation effect is the majo poblem with biodiesel as it causes a pemanent fuel dilution and theeby a viscosity loss. It additionally causes deposits especially on the pistons and piston ings. Stuck piston ings will then wosen the effect of fuel dilution and blow by eaching into the oil, (Mölle, 2001). The inceased amount of unsatuated fatty acids fom the biodiesel also affects the oxidation disposition of the oil; efe to section 3.3 in this pape. The scatteplots in figue 3.1 confim the stong coelation between the fuel dilution in the oil samples in [weight %] and the decease of the kinematic viscosity at 100 C in [mm 2 /s]. It is noticeable that the diesel values ae typically in the ange of 0 8 wt. %. Highe values >15 wt. % esult fom special test pogammes such as egeneation test uns, which undeline the wall impingement poblem caused by DPF.

5 29 The SAE 30 lowe limit is 9.3 mm²/s and becomes violated fo values >5% fuel. RME and gasoline data shows the same chaacteistic of viscosity as diesel. Anothe liquid contamination of the oil besides fuel is wate and is usually caused by condensation at low oil tempeatues. Wate initially educes the viscosity of the oil but it will not be examined futhe in this pape since small amounts of condensed wate evapoate quickly with highe oil tempeatues. Coolant wate can also get into the oil in the case of a sevee engine poblem. popeties. The most significant dipole moment due to the polaisation effect with any fuel occus with ethanol. The esulting elative pemittivity is about 16.2, (Holdefe, 1999). An ethanol dilution in the oil fom a theoetical consideation will be detectable, but this will only be possible at low oil tempeatues due to the high volatility of ethanol. The effect of wate in the oil is also detectable as the dipole chaacteistic of the pola wate molecules causes a high dipole moment and thus a high elative pemittivity of 80.3, (Holdefe, 1999). Data plots with wate contamination of the oil wee not available and ae not consideed in this wok. Figue 3.1.: Effect of oil dilution by Diesel fuel on kinematic viscosity Figue 3.2 indicates that fuel dilution has no significant effect on the elative pemittivity ε e. The eason is the similaity of the pemittivity numbe of diesel with ε e of 2.1, compaed to fesh engine oil which is in the ange of The diesel molecules ae non pola and thei contibution to an electic dipole moment in the pesence of an altenating field is elatively low. The elative pemittivity of petol is 2.0 and the elative dielectic constant of biodiesel (B100) is about as the este components cause a slightly highe dipole moment. The slight positive slope in figue 3.2 can be ascibed to scatteing effects. Howeve, in house expeiments with a 10% B100 fuel dilution in fesh oil showed no identifiable change in the dielectic Figue 3.2.: Effect of oil dilution by Diesel fuel on elative pemittivity 3.2. Effect of soot on viscosity and pemittivity: Soot is entained in the oil usually due to blow by of exhaust gas. Soot in the oil causes abasion and acceleated wea in the beaing, valves and camshaft. Elevated levels of soot in the oil cause sludge and the fomation of deposits. Deposits on the piston ing goves ae especially hamful as they limit the feedom of movement of the piston ings and ultimately cause the piston ings to get stuck. When the feedom of movement of the piston ings becomes limited, they will not fully seal the gap between the piston

6 30 and the cylinde wall. This causes moe blow by and moe soot loading of the oil. Figue 3.3 indicates a stong viscosity incease with highe soot loading of the oil. The amount of soot in the evised oil data fo the diesel engine was up to 8% wt. It is noticeable that even an amount of 2% soot may cause the viscosity to incease >12.5 mm²/s and thus to violate the SAE 30 uppe limit at 100 C. Soot fomation with biodiesel is less than fo standad diesel due to a highe cetane numbe and 11% oxygen (Schmiedel, 2004). Figue 3.3.: Soot accumulation effect on kinematic viscosity The elative pemittivity of soot is 18.8 and diffes consideably fom the pemittivity of fesh oil, (Holdefe, 1999). The high polaisability of the non pola soot paticles causes a elatively high dipole moment in the pesence of an extenal altenating field. Figue 3.4. shows a significant coelation between the amount of soot in the engine oil and the incease in the elative pemittivity Effect of oxidation/nitation on viscosity and pemittivity: The stating point fo the oxidation pocess is un bunt hydocabons fomed in the leane combustion flame egions whee the combustion tempeatue is too low to ensue a complete oxidation of the fuel. Figue 3.4.: Soot accumulation effect on elative pemittivity The unbunt hydocabons in the oil combined with oxygen is the stating point of the fomation of adicals, (Mölle, 2001). The catalytic influence of metal in the oil, e.g. fom abasion and high tempeatues acceleate the fomation of adicals and substantially influences thus the oxidation pocess. The end poduct of the oxidation chain is the fomation of oil insoluble polymes and cabon acids. The polymes cause a sevee incease in the viscosity and the cabon acids affect the engine beaings causing acceleated engine wea. The oxidation pocess is much moe pogessive with unbunt biodiesel as the unsatuated fatty acids ae pone to the eaction with oxygen. The additive agents of the oil ae depleted much faste and the oil lifetime is significantly educed especially with biodiesel fom less oxidation stable feedstock like SME (Soy Methyl Este), (Schmiedel, 2004). The main diffeence with nitation is the stating point which is with nitogen, i.e. NOx, compounds eaching into the oil due to blow by gases. The chain eaction is simila to the oxidation pocess and the end poduct is sludge, causing sevee deposits and nitic acids depleting the additive agents of the oil, esulting in coosion. Figue 3.5 indicates a distinct incease of the viscosity caused by oxidation due to the fomation of oil insoluble poducts. The IR (infa ed)

7 31 absoption goes up to 25 cm 1. The viscosity value of SAE 30 oils exceeds the allowed uppe viscosity limit of 12.5 mm²/s fo oxidation values of about 5 cm 1. Figue 3.5.: Oxidation effect on kinematic viscosity The cabon acids which ae fomed fom the oxidation eaction chain and the nitic acids fom the nitation pocess ae pola molecules and have a distinct dipole chaacteistic poviding a high dipole moment when influenced by an electic field. The elative pemittivity of these acids is 57.9 fo the cabon acid and 33.6 fo nitic acid, (Holdefe, 1999). Despite the boad stay aea in figue 3.6 thee is a stong coelation between the incease in the IR absoption and the elative pemittivity. Figue 3.6.: Oxidation effect on elative pemittivity 4. Discussion: Some senso manufactues and potential uses have published eseach papes whee some discete contaminations and the measued viscosity o pemittivity change in the senso output value was successfully demonstated, usually in laboatoy expeiments, (Schmitigal, 2005). The obsevations suggest the conclusion that diffeent contaminations ae all well detectable, eithe with the viscosity o with the pemittivity value fom the senso. This appoach appeas to wok fo discete laboatoy expeiments with known levels of contamination of the oil sample. Howeve, the cucial issue with the tansfe of these coelations fom laboatoy expeiments to vehicle applications is the inclusion of pevious knowledge about the atificially doped oil samples in the laboatoy. As soon as the oil is opeating in a eal engine unde eal cicum stances it must be consideed as a black box whee the opeato is unawae of all of the pocesses in the oil and the oil condition is only fully known when a complete oil change occus. Each of the contamination effects can attach value to the oveall oil which is measued by the senso. The viscosity and elative pemittivity data do not povide sufficient knowledge in ode to be able to distinguish which contamination influence actually contibutes to the oveall value. Extenal aspects such as the initial oil quality, the fuel type and quality, the opeational pofile and tempeatue also have an influence on the pogess of most of the contamination effects. The following illustation, figue 4.1 undelines the poblem and will claify the challenge with an on boad oil condition monitoing system. The senso data values fo the available oil condition data only pesents a combination of numeous and possible contamination

8 32 mechanisms. The contaminations listed above in tems of the chemical mixtue pocesses can be divided into homogenisation (mixing of soluble liquids), emulsification (mixing of limited soluble poducts) and suspension (mixing of solids in liquids). Figue 4.1.: Masking of oil degadation pocesses in the sump Each blend component included in the engine oil is consideed as an individual dilution and can be chaacteised by the individual kinematic viscosity υ in [mm²/s]. (Evapoation, volatilisation losses and tempeatue effects ae not consideed hee). The VBN (Viscosity Blend Numbe) fo each substance is, (Maples, 2000): VBN = ln [ ln ( υ )] (4.1) The VBN of the final blend can be deived fom the Lomonossow Lavoisie law and is defined as: VBN = [ χ VBN ] + [ χ VBN ] [ χ VBN ] Blend A A B B n n (4.2) whee χ A to χ n ae the mass faction of each blend component, i.e. the oil and the contaminants. The oveall kinematic viscosity is deived by solving equation 4.1 fo υ. VBN Blend e ν = e (4.3) The final viscosity value υ fom equation 4.3 is the measued value by the on boad senso. Any engine oil contamination examined above, except fuel dilution with conventional fuel, inceases the numbe of pola and non pola molecules in the oil. The pesence of an altenating field causes the polaisation of these molecules, theefoe foming a dipole moment μ in [C m], (Atkins, 1986). The esulting aveage dipole moment of the pola molecules can be expessed as: 2 * µ = µ E / 3 k T (4.4) aveage with E* in [V m 1 ] is the local electic field aound the molecule and k is the Boltzmannconstant in [J/K] and T in [K] as tempeatue, (Foth, 2005). The esulting induced dipole moment of the non pola molecules can be expessed as * µ induced = α E (4.5) whee α is the polaisability of the molecule in [J 1 C 2 m 1 ]. The final polaisation P in elation to the density ρ, the numbe of molecules N A and the molecula mass M m gives * P = ( ρ N / )( 2 A M m α + µ / 3 k T ) E (4.6) whee * P ( ε + 2 ) E = (4.7) 3 ε 0 ( ε 1 ) which gives Debye s equation fo the oveall mola polaisability in elation to the pemittivity 2 ε 1 M m µ P = 3 ε = N + 0 A α ε + 2 ρ 3 k T (4.8) The measued elative pemittivity ε e using the senso based on Debye s equation will povide a quantitative sum of all pesent polaisation effects in the oil. These consideations fo the viscosity and the pemittivity with espect to an on boad application in a vehicle, poves the

9 33 assumption that a subsequent sepaation of the contamination effects is not possible with the cuently available senso data. Futhe eseach activities in the cuent pogamme ae based on an evaluation of the time dependent total viscosity and elative pemittivity pogess as indicated in figue 4.2 fo a schematic engine test un. algoithm can neve claim to be an oil laboatoy on wheels but it will help to pevent the dive fom loosing mobility caused by oil degadation. 6. Refeences: Agilent Technologies Inc, Basics of Measuing the Dielectic Popeties of Mateials, Application Note, Atkins, PW., Physical Chemisty 3 d Edition, Oxfod Univesity Pess, Oxfod, 1986 Figue 4.2: Oil condition pogess egading the engine untime 5. Conclusion: Cuent on boad oil condition senso technology offes the viscosity and pemittivity to measue the physical and chemical oil condition. Any contamination eaching into the oil, changes at least one available senso paamete. Individual contamination effects cannot be veified and measued in the sump as a multitude of unknown effects ae possible. Futhe eseach wok will be focused on the analysis of available senso data to monito changes fom fesh oil values. The intention of the pesent eseach is to know pecisely whethe something is happen with the engine oil instead of tying to indicate any individual effect as this is not possible using cuent senso technology. An on boad oil condition senso combined with a smat Dobinski H., Buhdof A., Lüdtke O., Knippe U., Multipaamete Oil Condition Senso Based on the Tuning Fok Pinciple, SAE Technical Pape , SAE Intenational, 2007 Wold Congess Detoit Michigan, Apil 2007 Holdefe, M. Relative Dielektizitätskonstante ε von flüssigen und festen Medien (Relative Dielectic Constant ε of Liquid and Solid Substances), Endess+Hausse Notes, 1999 Lin, Y, Rodiguez, LG., Delphi Technologies Inc., US Patent # 6,725,707 B1, 2004 Foth HJ, Thieme Römpp, Geog Thieme Velag KG, Stuttgat Gemany, 2005 Luthe, R., Altenative Kaftstoffe aus Sicht de Motoenschmieung (Altenative Fuels fom Engine Lubication View), MTZ Jahgang 69,MTZ, Mach 2008, pp Maples, RE., Petoleum Refiney Pocess Economics 2 nd. Edition, Pennwell Books, Tulsa USA, 2000

10 34 Matsiev, LF., Bennett, JW., McFaland, E.W., Application of Low Fequency Mechanical Resonatos to Liquid Popety Measuements,IEEE Ultasonics Symposium, IEEE, Sendai Japan 1998, pp Mölle, UJ. Nassa, J., Schmiestoffe im Betieb (Lubicants in Opeation), Spinge Velag, Belin, 2001 Reinhadt, GP., et al., TAE Band 361 Schmieung von Vebennungskaftmaschinen (Lubication of Combustion Engines), Expet Velag, Esslingen Gemany, 1992 Schmiedel, HP., Biofuels Popeties and Expeiences with thei Use Follow Up, DGMK Reseach Repot 611 1, DGMK, Hambug Gemany, 2004 Schmitigal, J., Moye, S., Evaluation of Sensos fo On Boad Diesel Oil Condition Monitoing of U.S. Amy Gound Equipment, SAE Technical Pape , SAE Intenational, 2005 Wold Congess Detoit Michigan, Apil 2005