a new method to warm up lubricating oil to improve the fuel

Deakin Research Online This is the published version of the abstract: Will, Frank and Boretti, Alberto 2011, A new method to warm up lubricating oil to improve the fuel efficiency during cold start, SAE international journal of engines, vol. 4, no. 1, pp. 175-187. Available from Deakin Research Online: http://hdl.handle.net/10536/DRO/DU:30040627 Reproduced with the kind permissions of the copyright owner. Copyright : 2011, SAE

ABSTRACTCold start driving cycles exhibit an increase in friction lossesdue to the low temperatures of metal and media compared tonormal operating engine conditions. These friction losses areresponsible for up to 10% penalty in fuel economy over theofficial drive cycles like the New European Drive Cycle(NEDC), where the temperature of the oil even at the end ofthe 1180 s of the drive cycle is below the fully warmed upvalues of between 100°C and 120°C. At engine oiltemperatures below 100°C the water from the blow bycondensates and dilutes the engine oil in the oil pan whichnegatively affects engine wear. Therefore engine oiltemperatures above 100°C are desirable to minimize enginewear through blow by condensate. The paper presents a newtechnique to warm up the engine oil that significantly reducesthe friction losses and therefore also reduces the fueleconomy penalty during a 22°C cold start NEDC. Chassisdynamometer experiments demonstrated fuel economyimprovements of over 7% as well as significant emissionreductions by rapidly increasing the oil temperature. Oiltemperatures were increased by up to 60°C during certainparts of the NEDC.

It is shown how a very simple sensitivity analysis can be usedto assess the relative size or efficiency of different heattransfer passes and the resulting fuel economy improvementpotential of different heat recovery systems system. Due to itssimplicity the method is very fast to use and therefore alsovery cost effective. The method demonstrated a very goodcorrelation for the fuel consumption within ±1% compared tomeasurements on a vehicle chassis roll.

INTRODUCTIONMaximum efficiencies of up to 35% are possible for gasolineengines but during typical day to day driving in passengercars only between 10% and 20% of the used fuel is convertedinto mechanical energy to move a car [1]. The majority of theenergy is wasted either through the exhaust mass flow intothe ambient air or via direct heat transfer from the engine andradiator into the environment. Engine friction is another largecontributor of efficiency losses. For a fully warmed up enginethe amount of energy wasted through engine friction can beanywhere between 10% at wide open throttle (WOT) up to40% during part load, for example during city driving orcruising at constant speeds of up to 120km/h on the highway[2]. At lower temperatures, for example between 20°C and30°C, the start temperature of the NEDC, the engine frictioncan be even 2.5 times higher [2, 3]. This strong increase infriction at low temperature is caused by logarithmic increaseof oil viscosity with decreasing temperature [4]. The effectsof ambient temperature and vehicle soak temperature wereinvestigated by Ford at a vehicle level where three vehicles ofdifferent size and manufacturer with different engineconcepts, for example direct injection gasoline and port fuelinjection, were soaked and tested at different temperatures.Fuel economy improvements of between 5% and 10% weredemonstrated for a temperature increase of only 10°C asshown in figure 1 [5, 6].

As friction levels are reduced with increasing temperaturesand on the other hand a lot of heat is wasted through theexhaust and the cooling system it seems to be logical to usesome of the wasted heat to reduce friction. This idea has beeninvestigated to some extend by warming up the coolant withexhaust heat through an exhaust gas heat exchanger. Such

A New Method to Warm Up Lubricating Oil toImprove the Fuel Efficiency During Cold Start

2011-01-0318Published

04/12/2011

Frank WillDeakin Univ.

Alberto BorettiSchool of Science & Engineering, Univ. of Ballarat

Copyright © 2011 SAE International

doi:10.4271/2011-01-0318

Gratis copy for Frank WillCopyright 2011 SAE International

E-mailing, copying and internet posting are prohibitedDownloaded Wednesday, March 02, 2011 10:35:09 PM

http://dx.doi.org/10.4271/2011-01-0318

systems are already in production for various vehicles withvery efficient drive trains even though these systemsdelivered only very minor fuel economy improvements up toa maximum of 1%, in some instances they even lead to areduction of fuel economy [7, 8]. So the main reason whysuch exhaust gas/coolant heat exchangers were implementedis that they improve the heater performance which means thatthe passenger compartment is warmed up faster at very lowtemperatures that are typical during winter in regions withcold climates like for example Scandinavia.

ANALYSIS OF HEAT TRANSFERFROM EXHAUST GAS TO ENGINEOILA relatively simple sensitivity analysis helps to explain whyonly marginal fuel economy improvements can be realized bytransferring exhaust heat to the engine coolant. To reduceengine friction the ultimate aim is to transfer as much heatinto the engine oil as soon as possible. The analysis considerstypical efficiencies for the various steps of the heat transferprocess that are involved to warm up the engine lubricationoil with the exhaust gas. The advantage of using the approachwith typical efficiencies is that it does not require a detailedand complicated model for the components involved. Theunderlying heat transfer equations are based on empiricalinvestigations anyway with relatively large uncertainties. Themain aim of this analysis is to demonstrate the order ofmagnitude of exhaust heat transferred into the oil. Theefficiencies for the different heat transfer steps are displayedin figure 2 where two different systems are compared, firstlya standard production configuration that uses a heatexchanger to transfer heat from the exhaust gas to thecoolant, and secondly a new approach that transfers theexhaust heat directly to the engine oil with the same heatexchanger.

EXHAUST GAS/COOLANT HEATEXCHANGERAs mentioned before the main purpose of such systems is towarm up the cabin faster at very cold temperatures. Anexhaust gas heat exchanger is installed in the exhaust behindthe catalyst. To avoid overheating of the coolant the heatexchanger can be bypassed on the exhaust side, controlled bya flap valve. On the coolant side the heat exchanger is in the“bypass” circuit that warms up the passenger compartmentwhen required. Such an arrangement is displayed in figure 17of [7]. The first step in this heat transfer process is thetransfer of heat from the exhaust gas to the engine coolant.Here a typical heat exchanger efficiency ηHE of 80% isassumed. This is an arbitrary value that depends on thespecific heat exchanger design and installation as well as onthe inlet temperatures of the two fluids.

In a second step the remaining heat is transferred from thecoolant into the metal structure of the engine that is also incontact with the engine oil, mainly the cylinder head and thecylinder block. So the cylinder head/and -block could beconsidered as a typical heat exchanger to transfer heat fromthe coolant to the engine metal structure. However, this heattransfer path normally acts in the opposite direction: thecombustion heat is transferred into the cylinder head and -block from where it is transferred into the coolant to avoidlocal overheating. With an increase in coolant temperaturethe heat loss from the combustion chamber to the coolant isactually reduced. The effects on the combustion process areneglected in this example for further simplification. Becausethe heat transfer function of cylinder head and a cylinderblock is only a side function, they are not designed liketypical heat exchangers. For example the wall thickness ofthe metal is a multiple of a typical heat exchanger and thesurface area is relatively small due to the absence of fins thatare difficult to incorporate in the manufacturing process of a

Figure 1. Fuel consumption as a function of test temperatures [6]



cylinder block and cylinder head. For dynamic warm upconsiderations the wall thicknesses of the cylinder head and -block with its specific heat capacities would be alsoimportant which in fact further reduces the warm up of theoil, but again these effects are not discussed here. Theefficiency number should be smaller compared to the exhaustgas heat exchanger. Even though in most areas of the cylinderhead and -block the heat is transferred from the metal to thecoolant we can introduce an efficiency factor ηCM of 50%that represents the percentage of heat from the coolant that isremaining in the metal structure as a result of the reduction ofheat loss from the metal to the coolant due to higher coolanttemperatures. Again this is an arbitrary assumption based ona best engineering guess.

The final step is the transfer of heat from the cylinder blockand -head metal structure to the lubrication oil. The additionalheat from the coolant that is remaining in the metal structureas a result of the reduction of heat loss from the metal to thecoolant is transferred in two directions: Firstly the targetdirection is to transfer this excessive heat into the engine oiland secondly the majority of the heat is lost into the ambientair with its much lower temperature. The heat rate transferredinto the ambient air can be described simply by theconvective heat transfer equation

(1)where hA is the convective heat transfer coefficient to the air,AA is the surface area of the engine in contact with theambient air, TM the average metal temperature of cylinderhead and -block and TA the ambient air temperature. In thesame way the heat rate transferred from the engine metal tothe lubrication oil can be described as

(2)

Here hO is the convective heat transfer coefficient to the oil,AO is the surface area of the engine in contact with thelubrication oil and TO the average oil temperature. Thecombined heat transfer rate from the metal to the air and oil is

(3)

So the resulting efficiency of the heat transfer rate into the oilrelated to the combined heat transfer rate is

(4)

By dividing equation (2) through equation (1) the heat rate tothe oil QO is described in relation to the heat rate to the airQA

(5)

With the introduction of three oil to air ratio factors QO canbe described as a function of QA. The ratio factors are

(6)

The combined oil to air ratio factor is

(7)

So the resulting efficiency of the heat transfer rate into the oilrelated to the combined heat transfer rate is

(8)

Some of these ratio factors were determined throughmeasurements of the vehicle that was tested as describedlater. The test object was a Ford Falcon Turbo In-line 6cylinder engine with a displacement of 4.0 liter. At the startof a NEDC test all temperatures are the same at 22°C. At theend of the test the maximum metal temperature of 96°C wasrecorded and the maximum oil temperature was 82°C, asshown in figure 6 for the baseline configuration (heatexchanger off). The resulting average temperatures are TM =59°C and TO = 52°C, so the resulting ‘temperature ratio’ ηTis only 0.19. For the surface ratio a value of ηA=0.1 wascalculated based on the outside dimensions of the engine(0.7m long, 0.2m wide and 0.6m high) and oil gallery (totallength 3.3m, average diameter D=13mm), for more accuratecalculations a CAD model of the block and head would berequired.

The heat transfer processes during an NEDC cold start test isquite dynamic due to the constantly changing temperatures,material properties, vehicle speeds and engine speeds.Describing that process would require the solution of variouscomplex differential equations which is not the purpose ofthis study. To simplify the process average values over thedrive cycle were assumed for vehicle speed and materialproperties. This is a common practice even for complexcombustion simulations with modern simulation tools. Mostof these combustion simulation tools use empirical equationsfor the heat transfer in the combustion chamber in the form ofheat transfer times the temperature difference, for examplethe equations from Woschni [9], Bargende [10], Hohenberg[11], or Annand [12]. None of these equations includesolutions for the fundamental differential equations even



though an heat transfer equation based on the solutions ofthese differential equations has been developed byKleinschmidt [13, 14]. So the heat transfer coefficient for theengine oil was calculated using the following equations

(9)

(10)

Material properties and dimensions that were used for thatcalculation are: effective length of oil gallery L=2m, enginewidth d=0.2m, oil flow rate 301/min at an average enginespeed of 1500RPM, and the oil properties at 50°C: kinematicviscosity ν=161mm2/s, conductivity k=0.142W/(m °C),density ρ= 870kg/m3, specific heat capacity cp=2.0055 kJ/(kg°C) [15]. For the convection of the oil in the galleries thatlead to a heat transfer coefficient hO= 36.2W/(m2 °C).

For the convection of air around the engine a heat transfercoefficient is calculated through [4]

(11)

With an average drive cycle speed of ν=9.3m/s that results inan heat transfer coefficient hA=34.1 W/(m2 °C) which leadsto a heat transfer coefficient ratio ηh of 1.06.

Therefore the combined ratio factor ηC is only 0.0202 leadingto a resulting efficiency for the convective heat transfer fromthe engine structure to the lubrication oil ηR of only 0.0198.

Now the total heat transfer from the exhaust gas to the engineoil is a multiplication of the individual efficiencies of all threeheat transfer steps

(12)

So the total efficiency is 0.8% and therefore veryinsignificant.

That means that only 0.8% of the available exhaust heat istransferred into the engine oil to reduce friction. If theaverage engine friction during a NEDC test is considered as25% of the fuel energy as stated before in [2] and [3] thisresults in a fuel economy improvement of 0.2%. That doesnot include negative effects of increased power requirementsof the water pump to overcome the additional resistance ofthe additional heat exchanger and the relevant hoses and the

additional thermal masses of the heat exchanger itself andparticularly the additional cooling fluid in the heat exchangerwhich will compensate at least some of this theoretical fueleconomy improvement potential.

So the results of this very simple sensitivity analysis are verywell in line with results from actual measurements [7, 8] eventhough such a small amount is very difficult to verify in astandard fuel economy test cycle like the NEDC with typicaltest to test variability in the range of ±1% under verycontrolled conditions.

To get an idea how potential variations to the single valueassumptions would affect the results of this analysis, a best-and worst case scenario are added. This also reflects theproblem that most convective heat transfer calculations arebased on empirical equations and it is not unusual to seevariations of up to ±25% depending on the type of equationthat is used [15]. Therefore in the best case scenario all threeefficiencies are increased by 25% and in the worst casescenario reduced by 25%. The result is that the totalefficiency will be multiplied by 0.42 in the worst casescenario or with 1.95 for best case. For the worst case thatleads to a new heat transfer efficiency of only 0.3% and a fuelconsumption reduction of 0.08%. In the best case scenario theheat transfer efficiency increases to 1.5% leading to a fuelconsumption reduction of 0.4 %. Both of these results arevery small and therefore still in line with the measurements ofsuch a system [7, 8].

DIRECT HEAT TRANSFER FROMEXHAUST GAS TO ENGINE OILThe above analysis demonstrated that using coolant to warmup the engine oil involves a series of heat transfer processesthat are not very efficient, in particular the heat transfer fromthe coolant to the engine oil by using the engine as a heatexchanger. The reasons are significant heat losses from themetal structure of the cylinder block and head to the ambientair. Even though one primary function of the cylinder headand block is to transfer heat, the aim is to transfer the heat tothe coolant and the ambient air to avoid local overheating ofvarious components. But the engine is not really designed totransfer heat quickly to the engine oil. So if the major lossfactors in equation (12) could be eliminated, particularly ηR,that would dramatically improve the efficiency of such anexhaust gas recovery system. Therefore a new approach is towarm up the engine oil directly with the exhaust gas. Astandard production exhaust gas heat exchanger that isdesigned to warm up the coolant was installed in thelubrication system instead of the coolant system. An adaptorbetween engine block and oil filter was used to pick up the oilfrom behind the oil pump. From the exhaust gas heatexchanger the oil was returned back into the oil filter andfrom there back into the engine oil gallery. A system diagram



is shown in figure 3. For such a system the same heat transferefficiency analysis was conducted as for the standard systemwith the exhaust gas/coolant heat exchanger. From theprevious three efficiency factors only one remains, the heatexchanger efficiency ηHE of 80%. This value was notchanged because even though the oil has a lower conductivitythan coolant - which could result in a lower efficiency - thespecific heat capacity is also lower which partially offsets thenegative effect of the reduced conductivity, but the mainreason is that the mass flow of an engine oil pump is lowerthan the mass flow of an engine water pump if both aredriven mechanically through the engine. This result in largertemperature differences of the oil through the heat exchangerwhich in fact improves the heat exchanger efficiency definedas

(13)

TO2 is the oil temperature after the heat exchanger, TO1describes the temperature of the oil entering the heatexchanger and exhaust temperature into the heat exchanger isTE1. Now unfortunately once the oil is warmed up a little, itflows back into the engine oil pan. Here a significant portionof the energy is also lost into the ambient air in a similar wayas previously discussed for the heat loss from the cylinderhead/-block into the ambient air. Similar as the cylinder blockand -head the second main function of the oil pan is totransfer heat from the oil into the ambient air under certainconditions like wide open throttle. However, the oil pan canbe considered as a slightly more efficient heat exchangercompared to the cylinder block and -head because the ‘walls’are much thinner for example by using sheet metal as amaterial or in case aluminum is used with its high thermalconductivity often ribs are used on the outside of the oil panto improve the stiffness for reduced noise emissions but alsoto increase the heat transfer. So for the purpose of thisanalysis an efficiency ηOA of 60% is assumed which isslightly higher than the 50% for the heat transfer from coolantto the metal structure. As we are interested in the heat fromthe oil/exhaust gas heat exchanger that is actually retained inthe oil this efficiency ηOP is

(14)

So the total efficiency for the oil/exhaust gas heat exchangeris 0.32.

(15)

With 32% the efficiency of an oil/exhaust gas heat exchangeris over 27 times more efficient compared to heating up the

coolant with exhaust gas. Assuming the same frictionpercentage of 25% over the NEDC as in the previous chapterthis would result in a fuel economy improvement of 8%without consideration of additional pumping losses of the oilpump and the additional thermal masses of the additional heatexchanger and the additional oil required.

Figure 2. Heat transfer efficiencies for coolant/exhaustgas heat exchanger and oil/exhaust gas heat exchanger

Even though this analysis just focuses on static conditions,the transient effects of using engine oil instead of coolant alsohave a positive effect due to the lower heat capacity of the oilcompared to the coolant so the additional fluid that isrequired by both such systems has a less negative effect withusing oil.

A similar best case and worst case scenario was considered asfor the exhaust gas/coolant heat exchanger. Here only twoefficiency factors need to be reduced or increased by 25%. Ina worst case scenario the total heat exchange efficiencywould reduce to 18% compared to a solid improvement of50% for the best case range. For the fuel consumptionreductions that would mean a worst case of 4.5% which isstill quite remarkable and a maximum reduction of 12.5% inbest case. If different engines will be analyzed obviouslythese numbers would also depend on a couple of other factorslike the material properties (aluminum versus cast iron), anddetailed designs and dimensions of the engine and the heatexchanger.

TEST RESULTS FOR OIL/EXHAUSTGAS HEAT EXCHANGERThe system configuration with the exhaust gas/oil heatexchanger as shown in figure 3 was tested in a vehicle overthe NEDC test cycle in two conditions firstly without the heatexchanger activated and secondly with an active heatexchanger. The heat exchanger with a by-pass valve that wasmanually operated was installed in the exhaust behind thecatalyst. Due to an under-floor catalyst configuration thatresulted in relatively long oil hoses that were connected to the



engine via an adapter plate that was mounted between oilfilter and engine block. Such adaptors are typically used forinstallations on engine dynamometers where the oiltemperature is controlled or if an additional oil cooler needsto be installed in a vehicle.

To reduce the influence of test to test variability like driverinfluence, soak temperature and others [6] five tests were runin both conditions and the results were averaged to increasethe confidence level of this study. The tests were performedon vehicle chassis rolls of the Advanced Centre forAutomotive Research and Testing, as described in [6, 16].The most interesting results were the improvements in fueleconomy as depicted in figure 4. Over the combined NEDCtest the fuel economy was improved by over 7%.

For the urban part of the drive cycle the fuel economy waseven improved by 8% and also in the extra urban part the fueleconomy was increased by 7%. An equivalent reduction ofCarbon Dioxide (CO2) emissions of 21g/km was measured.For the regulated exhaust emissions also significantreductions were verified (figure 5). The Carbon Monoxide(CO) emissions were influenced at the most with a reductionof 27% at the tailpipe. Also the Nitrogen Oxides (NOx)emissions were reduced largely by 19%. Only the Hydro-Carbon emissions (HC) remained more or less constant witha minor reduction of only 2%.

The analysis of the oil temperatures revealed some veryinteresting phenomenon. Figure 6 shows several oil

temperatures: for the configuration with the active heatexchanger all traces are displayed with dotted lines and thesolid lines show the configuration without heat exchanger.The oil temperatures out of the engine into the heat exchangerare drawn in blue, the red lines are the oil temperatures afterthe heat exchanger flowing into the engine and the black linesare the recordings from the standard oil temperature sensor inthe main oil gallery.

Figure 4. Fuel economy improvements with oil/exhaustgas heat exchanger (vehicle measurements averaged

over 5 tests in each configuration)

The most interesting finding with the heat exchanger active isthat the oil temperature rises sharply within the first 100seconds to almost 80°C. This is followed by a temperaturedrop and then it rises again to 92°C in line with the increasein vehicle speed. This is followed by another unexpectedtemperature drop after 180 seconds. Then the temperature

Figure 3. Layout of the tested system with oil/exhaust gas heat exchanger



difference compared to the standard configuration stabilizesat 25°C and it increases to a difference of over 35°C duringthe high speed part of the extra urban cycle at the end of thetest. A similar behavior with some delay and dampening isobserved for the temperature in the main oil gallery. The fastincrease happens only after 100 seconds and there is nodecrease in temperature evident afterwards. Only thedifference between the standard conditions decreases a little.A growing difference over time between the on- and offcondition can be seen for the oil that is pumped into the heatexchanger out of the oil pan.

DISCUSSION OF SYSTEM BENEFITSAND THE UNDERLYING CAUSESOIL TEMPERATURESThe fast temperature increase out of the heat exchanger (HE)is caused through the operation of the pressure relieve vale:this valve opens during the cold start due to high oil pressurecaused by the low oil viscosity and most of the oil flowsthrough the relieve valve by-pass instead through the oilgalleries. Therefore the heat exchanger oil flow is quite smalland causes a rapid temperature increase. When the oil warmsup the pressure drops and at some stage the pressure relievevalve closes, with the consequence that the flow through theheat exchanger immediately increases which causes that thetemperature drops. The temperature starts to increase againduring the middle of the test with a quite constant offsetcompared to the configuration without heat exchangerindicating a stable condition. Finally at the highest speed of120km/h this oil inlet temperature starts to increase again andmuch faster than without HE. This is caused by the higherexhaust gas flow rates.

FUEL ECONOMYA fuel economy improvement of 7% is quite significant andin the same range as other much more complex and expensivetechnology so this system and has a great potential to be

introduced into series production. The test results are well inline with the predictions of 8% fuel economy improvement asdiscussed before. The difference of 1% compared to thepredictions is very small. Potential reasons for this smalldifference is that many assumptions and simplifications weremade for the predictions for example the increased pumpinglosses for the oil pump were not calculated and the increasedthermal masses for the heat exchanger with the additional oilwas also not included in the study. But for such a simplesensitivity analysis the results correlate exceptionally wellwith the measurements which demonstrates the viability forsuch an approach as a quick and reliable method to decideabout preferred design configurations. The demonstrated fueleconomy improvement represents between 50% and 70% ofthe cold/hot factor which is between 10% and 15% for mostengines [6, 17, 18]. The cold hot factor is the differencebetween the fuel consumption of a cold test and a hot test thatis followed directly after the cold test, divided through the hottest fuel consumption. Interestingly not only for the urbanpart 1 a large improvement of 8% was demonstrated. Thisimprovement was expected because the temperatures duringthe urban part are lower that during the extra urban part andthe oil viscosity increases logarithmic with reduced oiltemperature. Similar simulations were conducted by Farrantet al. [19] that predicted fuel consumption changes forconstant temperatures compared to the baseline engine. It wasestimated that for a constant engine temperature of 94°C thefuel consumption over the ECE15 cycle could be reduced by20% with completely warm oil. But for the extra urban drivecycle (EUDC) only 2% reductions in fuel consumption werepredicted. For the combined NEDC the maximumimprovement potential was 12%. So the combined results arewell in line with the measurements of this study consideringthe fact that the tests were started with cold temperatures andnot with fully warm oil.

However, the fuel economy improvements for the second partare also very high so they are worth to be discussed in moredetail. The high fuel economy improvement measured for the

Figure 5. Emission reductions with oil/exhaust gas heat exchanger (vehicle measurements averaged over 5 tests in eachconfiguration)



EUDC can be explained by several reasons: firstly during thesecond part the engine load is higher resulting in a higherexhaust flow rate so much more heat is available in theexhaust to warm up the oil. That can be seen in thetemperature difference between the on- and off-position. Thedifference is much bigger in the second part, particularly forthe cold temperature that is coming out of the engine, here thedifference is around 4 times bigger. Secondly the enginespeed is also much higher compared to the urban part. Thatleads to higher friction values, both in absolute terms and as apercentage of the effective power [2].

Thirdly the cold hot factor includes some measures to warmup the catalyst faster, like increased idle speed or cold startspark retard. These effects are only active during the first partbut they are not affected by the oil temperature. And lastlythe friction can actually increase to some extent if the oiltemperatures are higher during the urban cycle. At very lowengine speeds the increase in temperature and viscosity cancause the lubrication to move from the most efficient hydro-dynamic state into mixed friction which can cause an increaseof friction.

Additional fuel economy improvements are possible duringreal world customer driving conditions. Over a whole yearthe average ambient temperature in Europe is around 11°Cwhich is 11°C lower compared to the conducted NEDC tests.According to [6] this lower temperature will increase fuelconsumption in the standard configuration without HE bybetween 5.5% and 11%. Higher fuel consumption at lowertemperatures offers an even bigger potential for fuel economyimprovement through a faster warm up of the oil. The coldhot factor could be reduced by over 50% with the HE, so asimilar percentage improvement is expected for colder realworld conditions. Therefore the real world fuel consumptionfor 11°C ambient temperature with HE would be additionallyimproved by between 2.8% to 5.5% compared to the 7%improvements at normal cold start temperatures of 22°C.

EMISSIONSThe reduced exhaust mass flow (7%) firstly causes a similarreduction of all emissions. Additionally the large COemissions reductions are explained by reduced wallquenching caused by higher oil temperatures. The oil

Figure 6. Oil temperatures with and without new exhaust gas/oil heat exchanger



temperatures are important to define the wall temperatureswithin the combustion chamber on the cylinder liner. Becausethe engine load is reduced due to the friction reduction, themaximum combustion temperatures are also reduced andtherefore less NOx emissions are created during the peakcombustion temperatures. The HE reduces the exhausttemperature after the catalyst which results in an increase ofwater condensation with the exhaust. NOx emissions candissolve in water, therefore the condensation of water is notallowed within an emissions sampling system duringemission certification testing. However, if the more watercondensates already within the exhaust system, this can leadto a further NOx emission reduction.

For the HC emissions also some reductions were expecteddue to the same reasons as for the CO reductions, reducedmass flow and reductions in flame quenching. However, upto 30% of the tailpipe HC emissions come from the oilemissions [20]. With higher oil temperatures more oilevaporates in the combustion chamber which offsets most ofthe benefits of a lower mass flow and the reduced quenching.

ENGINE WEARAround 10kg of water flows through the crankcase of atypical gasoline engine during an oil change interval of15,000km. This amount can be computed with the followingassumptions: an average blow by of 101/min [21], an averagevehicle speed of 50km/h and a 9.2% water content in theexhaust [22]. The crankcase ventilation system re-circulatesmost of that combustion water back into the intake manifold.But especially at low oil- and engine temperatures below100°C some of that blow-by water condensates in thecrankcase and runs into the oil sump where it is mixed withthe oil and “promotes the formation of acids corrosion and oilaging” [23].

Wear rates for valve train chains were measured in [24] atdifferent test temperatures. Wear rates of up to 150μg/hr weremeasured at cold engine oil temperatures between 20°C and50°C. This was 30 times higher compared to tests that wereconducted at higher oil temperatures between 110 and 140°C.So if operating conditions with oil temperatures below thewater condensation temperature of 100°C can be avoided - orat least be reduced - that will also lead to a reduction ofengine wear. This problem only gets worse if the engine isoperated at temperatures below 0°C: the water condensate inthe oil can turn into a solid state which further increases wearbut also can cause serious engine damage by blocking of oilpassages or crankcase ventilation [1] or by even burstingsome of those passages through the expansion when the waterturns into ice.

This problem particularly affects any very efficient modernpowertrains for example hybrids, because they stop duringidle so it take much longer for the oil to warm up. Therefore

many hybrids have a reduced oil change interval, for examplethe 2007 Toyota Camry Hybrid has a reduced oil changeinterval of only 8000km. The handbook even includeswarnings about potential oil thickening in case the shorter oilchange intervals are not followed [25].

A question that is often asked in relation to systems thatwarm up the oil is how overheating of the oil could beprevented which would cause a breakdown of the oil [23].These long term effects will depend on the specific systemconfiguration, for example heat exchanger design and valveoperating strategies, similar as in turbo chargers where oiland hot exhaust gases flow through. One has to rememberthat the hottest and therefore the most critical local areas forthe oil are the surfaces of the combustion chamber, inparticular the oil in the honing groves. Here the surface of theoil film is in direct contact with the combustion gas attemperatures of over 2000K [1]. This is where the oil getsdamaged, it breaks down, it condensates, and it even burnsand the maximum temperatures in the combustion chamberare even much higher than in the exhaust gas. Typicalmaximum oil temperature limits of between 120 and 150°Care specified for the oil sump. However these temperaturesare only reference temperatures as a result of the heat transferfrom the hot gases in the combustion chamber to the oil in thesump. These limit temperatures are set based on extensiveexperience where relationships or transfer functions betweenoil sump temperatures and oil break down in the combustionchamber have been established, which can vary betweendifferent engine configurations and designs.

A NEW SYSTEM CONFIGURATION -OVER7™The analysis of these test results initiated the idea for a novelconfiguration of the lubrication system to further improvefuel economy and to reduce costs. The system increases theheat transfer from the engine metal to the lubrication oil, itseparates the thermal masses of the active oil in the enginethat generates most of the friction from the relative passiveoil in the oil pan that is responsible for a normal slow warmup and it reduces the hydraulic power required by the oilpump.

The novel system has an oil return bypass from the cylinderhead oil galleries that connects directly to the oil pump (or oilpick up tube) so that a certain portion of the oil does not needto flow through the oil sump [26]. Such a by-pass reduces oreven eliminates the dissipating flow through the pressurerelieve valve during cold start. By increasing the oil flow ratethrough the engine and particularly the cylinder head this by-pass also increases the heat transfer from the cylinder head tothe oil, so the overall heat transfer process will be much moreefficient. A valve located in the oil by-pass controls the by-pass flow rate - and therefore also the engine oil pressure.Therefore another positive benefit of that by-pass is that the



hydraulic power of the oil pump is reduced, similar as in avariable oil pump, but without the costly need to redesign theengine to replace the oil pump. By partially separating thethermal masses between the friction active oil in the engineand the passive oil in the oil pan the oil that runs through theoil galleries and bearings warms up much faster than the oilin the oil pan which reduces friction, even without the need ofan exhaust gas heat exchanger. Alternatively the advantagesof an exhaust gas heat exchanger can me multiplied throughinstallation in the cylinder head lubrication by-pass. The newsystem is called Oil Viscosity Energy Recovery SystemOVER7™ highlighting the 7% fuel economy improvementpotential.

Analysis of the theoretical benefits and experimentalverification will be the subject of further studies. It isexpected that one of the main benefits of this new systemconfiguration will be a significant reduction of costs as theheat exchanger size can be reduced significantly and standardproduction proved EGR components can be used.Alternatively if the focus is on the lowest cost per fueleconomy improvement instead of maximizing fuel economy,the system can be designed even without an exhaust gas heatexchanger in its most simple configuration. This even opensthe opportunity for after-market conversions.

CONCLUSIONSWarming up lubrication oil with exhaust heat is a veryattractive method to reduce fuel consumption. For therelatively small cost of a heat exchanger an impressive 7%fuel economy improvement was demonstrated over anaverage of 5 NEDC tests on vehicle chassis rolldynamometers. The fuel economy improvements were verysimilar for the urban part and the extra urban part which isvery important for the real world fuel economy potential e.g.for vehicles that are either driven over longer distances or thatare operated at lower average temperatures compared to the22°C standard test temperature. Oil temperatures wereincreased by up to 60°C during certain parts of the NEDCtest.

Exhaust emissions were also reduced particularly COemissions by 27% and NOx by 19% without changes to theafter treatment system or adjustments to the enginecalibration or emission control strategy. Optimization ofcalibration parameters and control strategies may offerfurther potential that is worth exploring. Utilizing the lowertemperature of the exhaust gas after the heat exchanger forexhaust gas re-circulation offers some further emissionreduction potential.

Reduced engine wear is another positive effect of suchsystems. Ingress of water into the lubrication oil will bereduced and water can be vaporized even during city drivingat low engine speeds and loads. That offers the potential of

extended oil change intervals and to use cheaper oil, twofurther opportunities to reduce operating costs.

This project showed that even the simplest sensitivityconsideration can be a very important tool for the evaluationof different concepts. Especially for new concepts thatinclude transient heat transfer such a qualitative approach incombination with accurate experiments and measurementscan be much faster and cost efficient in helping to find thedesired improvements instead of time consuming detailedsimulations. These simulation models are very useful forparameter studies within a well know area, but often thesedetailed simulation models are only valid for a particularsystem configuration because they also rely on empiricalmeasurements. The method of using heat transfer efficienciesto analyze different heat transfer passes showed a goodcorrelation of within ±1% compared to measurements on avehicle chassis roll.

The study resulted in a new invention named OVER7™ thatinvolved an oil bypass to increase the heat transfer betweencylinder head and lubrication oil and to reduce hydrauliclosses in the lubrication system. This system is expected tooffer further fuel economy and costs benefits due to reductionin size and complexity.

It should be noted that these conclusions are based on thestudy of only one single vehicle. It will be interesting toinvestigate various different system configurations also forother vehicles with different powertrain configurations, forexample a vehicle with smaller engine and manualtransmission, a Diesel engine or a Hybrid powertrain. Theauthors are keen to collaborate with other organizations toconduct further relevant research and development to enablethe implementation of such cost efficient fuel economyimprovement technologies into mass production. Other issuesthat are worth to investigate further are noise behavior, cabinwarm up performance, or on-line wear analysis. Of particularinterest could be an analysis how conventional thermomanagement features like electrical thermostats or splitcooling systems would perform in combination with theOVER7™ system. It is expected that in many instances suchconventional thermo management features would onlyprovide very marginal benefits in combination with theOVER7™ system. This offers some further costs reductionpotential.

SUMMARYThe project successfully verified the significant potential ofusing exhaust heat to reduce engine friction. 7% fuelconsumption reduction was measured in emission tests onvehicle chassis rolls over an average of 5 tests. The testsconfirmed the results of a simple sensitivity analysis thatdemonstrated the advantages of heating up oil directlythrough exhaust gas. Results of the sensitivity analysis



correlated very well with the test results of a physicalprototype. Other advantages of such a system configurationare significant emission reductions and reduced engine wearrates. The analysis of the temperature recordings resulted inthe idea of an even more effective system to reduce fuelconsumption, emissions and engine wear called OVER7™.The system is well suited for any powertrain configurationincluding Diesels, Hybrids or alternative fuels.

REFERENCES1. van Basshuysen, R. and Schaefer, F., “InternalCombustion Engine Handbook,” SAE International,Warrendale, PA, ISBN 978-0-7680-1139-5, 2004, doi:10.4271/R-345.

2. Maassen, F. J., Dohmen, J., Pischinger, S., Schwaderlapp,M., “Reibleistungsreduktion Konstruktive Maßnahmen zurVerbrauchseinsparung”, MTZ 7-8/2005

3. Pischinger, S., Kochanowski, H.-A., Steffens, C., Sonnen,S., Atzler, M., “Akustische Auslegung von Wälzlagern imKurbeltrieb”, MTZ 03/2009

4. n.n. Bosch Automotive Handbook, 7th edition, Wiley,2007

5. Tobergte, M., “Influence of Soak and Start Temperatureon Fuel Economy and Emissions, A Chassis RollInvestigation”, Ford, 1999

6. Will, F., “The importance of advanced test processes toreduce emissions and fuel consumption”, ICSAT conference,2008

7. Diehl, P., Haubner, F., Klopstein, S., and Koch, F.,“Exhaust Heat Recovery System for Modern Cars,” SAETechnical Paper 2001-01-1020, 2001, doi:10.4271/2001-01-1020.

8. Nakagawa, T; Tsubouchi, M; Suzuki, M. “Exhaust HeatRecirculation System for Actual Fuel Economy”, Journal ofthe Society of Automotive Engineers of Japan, Vol. 61, No. 7p. 49, 2007

9. Woschni, G.: “Die Berechnung der Wandwärmeverlusteund der thermischen Belastung der Bauteile vonDieselmotoren”, MTZ 31, (1970), p491-499

10. Bargende, M.: “Ein Gleichungsansatz zur Berechnungder instationären Wandwärmeverluste im Hochdruckteil vonOttomotoren”. Dissertation, TH Darmstadt (1990)

11. Hohenberg, G.: “Experimented Erfassung derWandwärme von Kolbenmotoren”. Habilitation, TU Graz(1980)

12. Annand, W.,: “Instantanious heat transfer rates to thecylinder head surface of a small compression-ignitionengine”, Proc Instn Mech Engrs 185 (72) (1971) 976-987

13. Kleinschmidt, W., Hebel, D.: “InstationaereWaermeuebertragung in Verbrennungsmotoren - Theory,

Simulation and comparison with Test Results”, final report ofresearch project K1600/1-2 German Research Association,1995

14. Kleinschmidt, W., Hebel, D.: “InstationaereWaermeuebertragung in Verbrennungsmotoren - Theory,Simulation and comparison with Test Results”, VDI ProgressReports, No 383 Society of German Engineers (VDI), 1999

15. Holman, J. P., “Heat Transfer” 10th edition, McGrawHill, New York, 2010

16. Boretti, A.A., Will, F., Watson, H.C., Brear, M.J., Dingli,R., Voice, G.: “Comparison of static and dynamic enginemodels on the transient performance of a passenger vehiclepowertrain”, FISITA congress Munich, F2008-12-287

17. Holzer, H.: “Waermemanagement im Antriebsstrang vonVerbrennungsmotoren”, PhD Thesis, Technical University ofVienna, cited in MTZ5/2002

18. Kunze, K., Wolff, S., Lade, I., and Tonhauser, J., “ASystematic Analysis of CO2-Reduction by an Optimised HeatSupply during Vehicle Warm-Up,” SAE Technical Paper2006-01-1450, 2006, doi:10.4271/2006-01-1450.

19. Farrant, P.E., Robertson, A., Hartland, J., and Joyce, S.,“The Application of Thermal Modelling to an Engine andTransmission to Improve Fuel Consumption Following aCold Start,” SAE Technical Paper 2005-01-2038, 2005, doi:10.4271/2005-01-2038.

20. Gohl, M., “Schnelle Ölemissionsmessung bei Otto- undDieselmotoren”, FVV Vorhaben Nr. 758,Massenspektrometrische Bestimmung der Ölemission imAbgas von Otto- und Dieselmotoren, FVV, 2002

21. Ebner, H. W., Jaschek, A. O., “Die Blow-by-MessungAnforderungen und Messprinzipien”, MTZ MotortechnischeZeitschnft 59 (1998) 2, p. 90 - 95

22. Pischinger, S., “Verbrennungsmotoren Band IIVorlesungsumdruck”, RWTH Aachen, 1995

23. van Basshuysen, R. and Schaefer, F., “Modern EngineTechnology from A to Z,” SAE International, Warrendale,PA, ISBN 978-0-7680-1705-2, 2007, doi:10.4271/R-373.

24. Schwarze, H., Brouwer, L., Knoll, G., Schlerege, F.,Müller-Frank, U. Kopnarski, M., Emrich, S., “Ölalterung undVerschleiß im Ottomotor: MTZ, 2008-10

25. n.n., “Owners manual Toyota Camry Hybrid”, ToyotaMotor Corporation, 2006

26. Will, F., “Process and device for lubrication of rotatingand oscillating components”, Patent application no. DE102009013943.5 filed 19/03/2009

CONTACT INFORMATIONFrank Will has over 18 years product developmentexperience with Ford Motor Company in Europe and



http://dx.doi.org/10.4271/R-345

http://www.sae.org/technical/papers/2001-01-1020

http://dx.doi.org/10.4271/2001-01-1020


http://dx.doi.org/10.4271/2006-01-1450


http://dx.doi.org/10.4271/2005-01-2038

http://dx.doi.org/10.4271/R-373

Australia where he was managing the Ford part of theAdvanced Centre of Automotive Research and Testing(ACART). Since 2009 he is senior lecturer at DeakinUniversity in Geelong, Australia. For questions regarding thisstudy or further details about the new OVER7™ system, or incase of interest in a collaborative research project toinvestigate the new system with a different engineconfiguration, for example a Diesel or Hybrid configurationhe can be contacted at: [email protected]

ACKNOWLEDGMENTSThe authors like to thank Dong Yang, Aristos Karavias andIno8 Pty Ltd for their support related to this exciting project.

NOMENCLATUREACART

Advanced Centre of Automotive Research and Testing

AAsurface area of the engine in contact with the ambientair

AOsurface area of the engine in contact with thelubrication oil

COCarbon Monoxide

CO2Carbon Dioxide

cpspecific heat capacity

Ddiameter

dwidth

EUDCextra urban drive cycle

HEheat exchanger

hAconvective heat transfer coefficient to the air

hOconvective heat transfer coefficient to the oil

kconductivity

Leffective length of oil gallery

NEDCNew European Drive Cycle

NOxNitrogen Oxides

NuNusselt number

OVER7™Oil Viscosity Energy Recovery System

PrPrandtl Number

ReReynolds Number

TAambient air temperature

TElexhaust gas temperature into heat exchanger

TMaverage metal temperature of cylinder head and -block

TO1Oil temperature into heat exchanger

TO2Oil temperature out of heat exchanger



mailto:[email protected]

QAheat rate transferred into the ambient

QCcombined heat transfer rate

QOheat rate transferred into the oil

vvelocity

WOTWide Open Throttle

ηAsurface ratio

ηCcombined oil to air ratio factor

ηCMefficiency factor coolant to metal

ηhheat transfer coefficient ratio

ηHEheat exchanger efficiency

ηOPoil pan heat exchange efficiency

ηRheat transfer rate efficiency

ηTtemperature ratio

νkinematic viscosity

ρdensity

The Engineering Meetings Board has approved this paper for publication. It hassuccessfully completed SAE's peer review process under the supervision of the sessionorganizer. This process requires a minimum of three (3) reviews by industry experts.

All rights reserved. No part of this publication may be reproduced, stored in aretrieval system, or transmitted, in any form or by any means, electronic, mechanical,photocopying, recording, or otherwise, without the prior written permission of SAE.

ISSN 0148-7191

Positions and opinions advanced in this paper are those of the author(s) and notnecessarily those of SAE. The author is solely responsible for the content of the paper.

SAE Customer Service:Tel: 877-606-7323 (inside USA and Canada)Tel: 724-776-4970 (outside USA)Fax: 724-776-0790Email: [email protected] Web Address: http://www.sae.orgPrinted in USA



a new method to warm up lubricating oil to improve the fuel

Documents