Bai 2. Qua trinh -ánh l-a --ng c- có acohol.pdf

8/11/2019 Bai 2. Qua trinh -nh l-a --ng c- c acohol.pdf

1/12

A study of mechanical variable valve operation with gasolinealcohol fuels

in a spark ignition engine

Alasdair Cairns a,, Hua Zhao a, Alan Todd b, Pavlos Aleiferis c

a School of Engineering & Design, Brunel University, Uxbridge UB83PH, UKb MAHLE Powertrain Ltd, St. James Mill Rd., Northampton, NN55TZ, UKc Department of Mechanical Engineering, University College London, Torrington Place, London, WC1E7JE, UK

h i g h l i g h t s

" High ethanol content fuels enabled further small fuel and NOxemissions savings.

" The primary mechanism was faster burning and improved residual gas tolerance." Butanol had a negligible effect on residual gas tolerance, regardless of volume." For all fuels, variable valve timing offered the greatest NOxreduction potential.

a r t i c l e i n f o

Article history:

Received 19 July 2012Received in revised form 7 October 2012

Accepted 18 October 2012

Available online 7 November 2012

Keywords:

Spark ignition

Variable valve actuation

Ethanol

Butanol

a b s t r a c t

This work involved study of the effects of gasolineethanol and gasolinebutanol blends on the combus-tion, fuel economy and engine-out emissions of a single cylinder research engine equipped with a

mechanical variable valvetrain on the inlet and variable valve timing on the exhaust. Gasoline or iso-

octane were splash blended with varying amounts of ethanol or 1-butanol and studied under a rangeof part-load engine conditions. During warm idle operation, high ethanol content fuels allowed signifi-

cant improvement in tolerance to internally recycled burned gases, primarily associated with increasedburning velocities of such blends when near to stoichiometric fuelling levels. In turn this allowed highervalve lifts to be used, with reduced throttling locally at the inlet valves, further small fuel savings and

reductions in engine-out emissions of NOx. Conversely, the use of 1-butanol had a negligible effect onresidual gas tolerance, regardless of blend volume. At moderate speeds and loads, where throttling losses

were less, it was apparent that the valvetrain could still be used to attain additional thermal efficiencyimprovements including reduced compression losses and further expansion work for all fuels. However,

a trade-off with increased pumping losses during the exhaust stroke was apparent, with the throttlingmoved from the inlet to the exhaust valves at themost retarded valve timings studied. For all fuel blends,it was extremely interesting to note that variable valve timing alone offered the greatest NOx reduction

potential at moderate loads, insinuating the ability to operate variable valve timing with and without

early intake valve closing may offer one viable path to meeting future engine emissions targets.2012 Elsevier Ltd. All rights reserved.

1. Introduction

The ability to vary the inlet and exhaust valve events of theSpark Ignition (SI) engine is well-known to facilitate improved

compromise between performance, fuel economy and emissions.Variable Valve Timing (VVT) is one such established techniquefor improving the fuel economy of the gasoline engine. There areseveral mechanisms by which VVT influences fuel consumption

including:

Increasing or delaying valve overlap, which increases trappedresiduals and reduces engine pumping losses (where thetrapped residual gases occupy part of the cylinder volume andhence allow less vacuum to be used in the inlet system at

part-load). Late Inlet Valve Closure (IVC), which further decreases pumping

losses by allowing further increased intake pressures (given thepiston pushes some of the air back into the inlet system).

Late Exhaust Valve Opening (EVO), which can increase expan-sion work.

As a result, fuel economy can be improved by up to 5%

over the European drive cycle, for example [13]. The choice of

0016-2361/$ - see front matter 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.fuel.2012.10.041

Corresponding author. Tel.: +44 (0)1895265175.

E-mail address:[email protected](A. Cairns).

Fuel 106 (2013) 802813

Contents lists available at SciVerse ScienceDirect

Fuel

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / f u e l
http://dx.doi.org/10.1016/j.fuel.2012.10.041mailto:[email protected]://dx.doi.org/10.1016/j.fuel.2012.10.041http://www.sciencedirect.com/science/journal/00162361http://www.elsevier.com/locate/fuelhttp://www.elsevier.com/locate/fuelhttp://www.sciencedirect.com/science/journal/00162361http://dx.doi.org/10.1016/j.fuel.2012.10.041mailto:[email protected]://dx.doi.org/10.1016/j.fuel.2012.10.041


2/12

optimum VVT strategy is highly dependent on exhaust manifolddesign, engine compression ratio, cam phasing limits due to

valve-to-piston clash, part-load residual dilution tolerance andthe importance of Wide Open Throttle (WOT) performance relativeto part-load fuel consumption and emissions. A so-calleddual-independent VVT strategy (where the timing of the inletand exhaust valves may be varied independently) offers high over-

lap potential and reasonable compromise between maximisingWOT torque and minimising part-load fuel consumption and emis-sions[4,5].

In relatively recent years, there has also been growing interest

in fully variable valvetrain systems for additional improvements,such as:

Further reduction in throttling losses via load control directly at

the inlet valve(s), hence the traditional intake throttle can effec-tively be disregarded.

Increased thermal efficiency through greater effective expan-sion ratio.

Numerous fully variable valvetrain strategies exist dependingon application, but the currently reported work is most concerned

with those SI engine strategies claiming to enable improved part-load fuel efficiency. During his notable study in this vein, Tuttlecompared the effects of Late Inlet Valve Closing (LIVC) and Early In-let Valve Closing (EIVC) in a gasoline single cylinder research en-gine. In early experiments [6], IVC was delayed by between 60and 96 crank. Fuel consumption was observed to reduce by upto 6.5% and was accompanied by lower engine-out emissions ofNOx(24%) but similar hydrocarbon levels. Tuttle concluded that96 was the maximum delay that could be tolerated due to loss

of effective compression ratio, which would significantly limitthe attainable speed-load map (assuming no external compressionwas available). In later work [7] it was concluded that the EIVCstrategy was favourable at part-load, allowing de-throttled opera-

tion over a wider speed-load window, albeit reliant on 200crank

range of inlet valve closing for greatest CO2reduction.For such EIVC operation it is necessary to employ a much short-

er inlet cam duration, allowing the exact required mass of air to en-

ter the cylinder before closing the valves at the appropriate phaseduring the intake stroke. Thereafter the engine effectively acts in amanner akin to an air spring, expanding the fresh charge belowatmospheric pressure prior to the compression stroke. One major

limitation of such operation is reduction of the in-cylinder turbu-lence intensity, associated with increased time under closed valveconditions. This was previously well demonstrated by Cleary andSilvas [8] under part-load cruising conditions (1300 rpm/3.3 bar

net IMEP), where EIVC resulted in prolonged combustion duration,reduced in-cylinder gas temperatures, reduced engine-out emis-sions of NOx(up to 25%) and increased values of unburned hydro-

carbons (also25%). The deterioration in burn rate was reduced byswitching to a LIVC strategy but, under the reduced valve durationconditions tested, the pumping losses were worse than the con-ventionally throttled case.

Recent benefits claimed from part-load EIVC operation arehighly dependent on the valvetrain system employed. At one endof the spectrum, various electro-magnetic and electro-hydraulicsystems have been proposed. Such camless systems have been

reported to allow the greatest potential for reduction in breathinglosses [9,10] and/or can be used to realize advanced modes of oper-ation such as controlled auto-ignition[11,12]. However, in generalthese systems often still have significant issues to overcome

including packaging, noise, limited engine speed and cost. There-fore, the majority of fully variable valvetrain systems entering

production have been mechanically based, often producing sinu-soidal valve lift profiles of reducing valve lift in proportion to valve

duration and providing fuel economy benefits of around 10% overthe European drive cycle [13,14]. Elsewhere, Sellnau and co-workers [15] have also demonstrated how simpler two-stagemechanical valve actuation can allow viable compromise on a

cost-benefit basis, achieving 5.5% improvement in fuel economyand 46% reduction in NOx over the US warmed-up Phase 3Environmental Protection Agency drive tests (cycles 1923).

The combination of EIVC with homogeneous direct fuel injec-

tion has also begun to warrant interest, with potential for furtherpart-load fuel savings via increased compression ratio. For exam-ple, workers on the Hotfire collaborative project previouslyexamined such effects in both optical and thermodynamic single

cylinder SI engine assemblies[16,17]. During this study, greatestfuel consumption benefits were achieved when just one of thetwo inlet valves was actuated. However, the swirl dissipated

quickly once the valve closed and the fuel economy benefits re-corded varied substantially depending on which of the two inletvalves was activated. This was associated with the asymmetric lay-out of the injector and spark plug within the combustion chamber.

Following such fundamental studies, the Fiat Multiair system hasemerged and is now being transferred to advanced SI applications[18].This system decouples the closing of the intake valves from

the conventional mechanical cam via a variable hydraulically col-lapsed connecting piston and arguably provides a good compro-mise in terms of cost between prior proposed fully variableelectro-hydraulic and conventional mechanical systems.

On another note, in recent years there has also been significant

global interest in alcohol fuels for SI engines, especially loweralcohol fuels of reduced carbon count. The idea of using alcoholas automotive fuel is not new [1921]. However, attention hasnow intensified as such fuels may present one viable renewable

solution, with potential to be used in a near CO2-neutral mannerthrough efficient conversion of biomass. First generation biofuelshave largely been based on ethanol, where national fuel qualitystandards typically allow between 5% and 10% volume inclusion

within the existing gasoline pool. The main exception is Brazil,

where fuels produced from sugar cane are widely available in gas-ohol to neat ethanol forms[22]. Elsewhere where less favourablebiomass production limits exist, another exception is availability of

gasoline containing up to 85% ethanol (E85) available as a nicheproduct for flex-fuel vehicles[2325]. In summary, production ofbiofuel from feedstock is clearly limited on a global basis. Nextgeneration processes are therefore currently being investigated

that may allow such fuels to be efficiently mass-produced fromalternative sources such as cellulose, algae or even recovered waste[26].In the meantime efforts are also underway to maximise theimpact of existing biofuel stock[27,28].

Compared to gasoline, the lower alcohols exhibit high latentheat of vaporisation and anti-knock rating, which makes themattractive for use in future downsized highly boosted SI engines

that endure significantly higher peak in-cylinder pressures andtemperatures [29]. Such downsizing may help offset the low en-ergy density of ethanol [30], while problems with cold start dueto reduced volatility can also be reduced via, for example, ad-

vanced DI operating strategies[31,32]. However, gasolineethanolblends are known to exhibit azeotropic behaviour, with profoundeffects on the vaporisation and thermodynamic properties of theblend. Kar and co-workers [33] previously performed cycle re-

solved in-cylinder temperature measurements and reported thatblends with less ethanol content (030%) tended to evaporatemore readily while higher concentrations (>50%) with reduced va-pour pressure did not and hence exhibited reduced evaporative

power.Elsewhere, higher alcohols such as propanol, butanol and

pentanol have also been considered for automotive use [3436].From a thermodynamic stance the higher alcohols generally exhi-

A. Cairns et al. / Fuel 106 (2013) 802813 803


3/12

bit favourable calorific value (and hence better volumetric fuelconsumption), better water tolerance, volatility control and lower

Reid vapour pressure. However, the increased molecular carbonchain reduces the available hydrogen bonding effect and benefitsin latent heat (charge cooling) and hence knock resistance are typ-ically reduced. Differences in source-to-wheel energy consumptionmust also be considered when producing such fuels en masse from

similar sources to bioethanol. For these reasons it can be arguedthat higher alcohols are better suited to widespread inclusion inlow-to-moderate blending levels. In this vein, some of the currentauthors previously examined butanol as a low volume blending

agent (16% volume, equivalent to E10 on an oxygen mass basis)and concluded minimal engine calibration changes were requiredduring typical part-load homogeneous Direct Injection (DI) opera-tion[37]. Later on, the study was extended to unthrottled engine

operation [38] where again for lower quantities of butanol and eth-anol it was clear that minimal calibration adjustments were re-quired (at part-load). However, improved EGR tolerance wasobserved with higher percentages of ethanol (up to E85) under

typical cruising conditions. This was suspected beforehand as pre-vious studies by Brusstar and co-workers[39]generally concludedthat high ethanol content improved part-load EGR tolerance. How-

ever, Brusstars experiments were undertaken under fixed highvalve duration and lift part-load conditions.

In summary, the focus of the current work was to better quan-tify the effects of alcohol and EIVC operation on EGR tolerance un-der the lowest speed-load conditions typically encountered while

also quantifying the changes in optimum valvetrain settings atmoderate speeds and loads where the effects of varying EGR toler-ance were less dominant.

2. Experimental setup

2.1. Experimental engine

The experiments were performed in a naturally aspirated singlecylinder four-valve per cylinder spark ignition research engine.Some general details of this unit are presented in Table 1.

The four-stroke engine assembly was based on a modifiedindustrial Lister-Petter diesel engine sub-assembly, re-fitted witha prototype water-cooled barrel and bespoke con-rod and flat-topped piston assemblies as previously reported in detail [38].

The use of this low-cost bottom end resulted in a peak enginespeed limit of 2200 rpm. The ports and combustion chambergeometry were based on those of a known production engine(Model Year 2004 Audi 2.0L FSI), but with the exception of smallerexhaust valves to allow the option of central DI to be studied in

other work. It is important to note that no form of combustionchamber masking (sometimes alternatively referred to as port

masking) was employed in this work although such masks havebeen adopted in some production solutions to compensate for

the reduced in-cylinder turbulence during EIVC operation[13]. Itis equally important to note that no intake system tumble flapwas employed in this work. Otherwise, although the head was de-

signed with a DI fuel system, a port fuel injector was fitted andused alone in the currently reported study. This injector was alsoa production part, with the four fuel sprays emanating from theinjector designed to straddle the valve stem and allow fuel injector

targeting toward the back of a warm, closed valve.The cylinder head included a prototype mechanical variable val-

vetrain assembly, fitted to both the inlet and exhaust. The systemused was an evolution of the MAHLE Variable Lift and Duration

(VLD) mechanism, previously introduced in detail[40,41]. In brief,the system is based on a lost motion shaft-in-shaft cam operat-ing principle. An example of an inlet VLD mechanism is illustratedin Fig. 1. For each camshaft the two opening control cams are

pressed onto the outer shaft. These cams are equivalent in valveopening profile and must open the two inlet valves in a synchro-nous fashion, via the lever assembly. The opening cam contours

were designed so that, if no closing control cam were available,the inlet valves would remain open at maximum lift for a pro-longed period before eventually closing in a safe manner. However,such operation is hypothetical as the closing control cam is avail-

able and pinned on to the inner shaft. The phase of this closing con-trol cam, relative to the opening control cam, can be advanced so asto close the valve earlier and reduce the valve lift and duration in

proportion. The profiles of the cams were designed so as to ensureacceptable dynamic forces were produced regardless of camshaftphase.

In summary, the opening and closing control cams act in tan-dem to produce a mean cam (and hence valve) lift curve. By

advancing the phasing of the inner shaft relative to the outer shaft,the closing of the valve is advanced, hence allowing reduced liftand duration to be achieved. The phasing of the closing controlcam was controlled using a prototype wide range (140 crank)

hydraulic cam phaser, denoted inFig. 2as the VLD Phaser. In or-der to then achieve fully variable valvetrain operation, a secondhydraulic VVT Phaser was fixed to the other end of the entirecamshaft assembly, attached to the outer tube and providing 40crank timing range. Identical ranges were available on the exhaust.

2.2. Intake valve lift versus duration measurements

The main objective of the current work was to study the com-bined effects of differing inlet valve operating strategies on com-

Table 1

Basic engine characteristics.

No of cylinders 1

Bore (mm) 82.5

Stroke (mm) 88.9Geometric compression ratio 9.8:1

Variable valve timing Fully variable (inlet and exhaust)

Fuel injection Port fuel injection (4 bar)

Spark plug NGK single electrodeFig. 1. Key valvetrain components (example shown for inlet VLD operation).

804 A. Cairns et al./ Fuel 106 (2013) 802813


4/12

bustion, performance and emissions with different ethanol and 1-butanol blends. An illustration of the valve strategies examined isshown inFig. 3. The experiments involved study of EIVC at varied

degrees of valve overlap. The exhaust VLD cam phaser remainedinactive and set for maximum valve lift throughout the experi-ments. In summary, the inlet valvetrain was variable in duration/

lift and timing whereas the exhaust valve actuation was only var-iable in timing as illustrated (with the exhaust timing range 11776bTDC Maximum Opening Position).

The valve lift profiles were designed to maintain acceptable dy-

namic loads at engine speeds up to 6500 rpm (as might be requiredin production). Prior to running the test engine, the valve lift versusduration was measured on a purpose made rig. One set of valveswere spot-faced and fitted to the cylinder head sub-assembly. Thissub-assembly was itself then fitted to a valve motion measurement

rig. In brief, this involved coupling the cylinder head cam drive to acylinder head test rig assembly, details of which are provided inFig. 4. The head was clamped in-line with the motor using a special

mounting plate. A laser differential vibrometer system was thenused to obtain direct valve velocity measurements via the laserDoppler interferometry technique, as illustrated in Fig. 5a. Thevalve lift was estimated via the fringe counting method, with ameasurement resolution of 5m. During the tests, the cylinder headoil circuit was connected to an oil conditioning rig that controlled

oil temperature and pressure to engine-like motoring conditions(90 C 2 C, 3 bar gauge). In conclusion, such measurements can-not directly account for variation in lift versus crank angle due tofiring engine conditions (e.g. torsional vibrations due to combus-

tion and/or increased gas temperature effects) but were still con-sidered to be acceptable for approximating the valve lift from theknown positions of the camshafts.

2.3. Experimental fuels

Set out in Table 2 are details of key fuel blend components

examined during this work, analysed by a third party supplierusing current ASTM standards (including D2699 and D2700 forRON and MON respectively). The exception is the latent heat, takenfrom the literature[35]. A commercial UK pump grade unleaded

gasoline (95 RON ULG) was obtained as a baseline fuel. Samplesof iso-octane (i100), ethanol (E100) and 1-butanol (Bu100) werealso acquired from a UK chemical supplier, where water contentwas guaranteed to fall below 100 ppm. The iso-octane fuel was

tested alone and also used to prepare splash blends of 25% volumeof ethanol (E25i75) and 1-butanol (Bu25i75). A splash blend of 25%volume ethanol and 75% volume gasoline (E25g75) was also pre-pared for comparison. Finally a commercial summer blend gradeE85 (85.35% volume ethanol) was also acquired and an equivalent

two-component splash blend of 85% volume ethanol/15% volume

iso-octane prepared to help compare the effects of practical higherethanol content fuels.

2.4. Test apparatus

During the tests, the cylinder head was fitted with a Kistler

6041A piezo-electric water-cooled in-cylinder pressure transducer,the face of which was mounted flush with the combustion cham-ber walls. Pressure data acquisition was performed using a SMETECCombi system. Corresponding thermodynamic parameters wereevaluated as the average of values compiled over 300 engine cy-

cles. Pressure data analysis was performed using the commercialpackage AVL Concerto Version 3.8, which included a simple sin-gle-zone heat release analysis based upon the first law of thermo-

dynamics, where the net heat release (Q) was computed as afunction of crank angle (h), in-cylinder volume (V) and pressure(p):

dQ

dh

n

n 1pdV

dh

1

n 1Vdp

dh 1

This simple representation neglects variation in polytropic coef-

ficient (n) throughout the cycle and cannot be relied upon for accu-rate post-processing of irreversibilities such as the in-cylinder wallheat transfer and piston blowby. Nonetheless, the model was con-sidered adequate for the qualitative comparisons intended. Other-

wise, the engine-out emissions were sampled to industrystandards using a Horiba MEXA 9100 analyser (with all hydrocar-bon measurements presented herein on a ppm methane basis).Fuel flow measurements were performed using a coriolis fuel flow

meter assembly, calibrated in situ to provide a maximum readingerror of


5/12

must be reduced to such low levels that a degree of throttling oc-curs at the inlet valves themselves. This in turn hinders the fueleconomy benefit achieved. This effect can be best seen when com-

paring the pressurevolume diagrams of a low valve overlap andhigh valve overlap case, as illustrated inFig. 6. This data was ob-

tained at engine speed and load conditions typical of highwaycruising e.g. 2000 rpm/2.7 bar net Indicated Mean Effective Pres-

sure (IMEPn), Maximum Brake Torque (MBT) spark timing andnear-stoichiometric fuelling (k= 0.99). The first case (marked EIVC)was produced with fixed low valve overlap whereas the second

case (denoted EIVC + VVT) was produced with the maximum pos-sible amount of valve overlap where the combustion stability limit

was reached. The limit used throughout this work was governed tobe where the standard deviation in gross IMEP (rIMEPg) reached

Fig. 4. Details of the cylinder head rig used to obtain dynamic valve lift measurements under motored valvetrain conditions.

ValveLift(mm)

0

1

2

3

4

5

6

7

8

9

Duration (crank)

60 80 100 120 140 160 180 200 220 240

(b)(a)

Fig. 5. (a) Example cylinder head fitted to the valve motion rig with the laser beam directed at one inlet valve and (b) inlet valve lift versus duration measurements obtained

on the motoring valve motion rig (measurements shown obtained at 750 rpm, equivalent to 1500 rpm crank).

Table 2

Key fuel properties.

95 RON ULG i100 C8H18 E100 C2H5OH E85 Bu100 C4H9OH

Lower heating value (MJ/kg) 43.2 44.2 26.6 28.9 33.2

Latent heat (kJ/kg) 349 305 837 532RON 95 100 110 109 94

MON 86 100 90 90 80Octane sensitivity 9 20 19 14

(R+ M)/2 91 100 100 100 87

Density @ 15 C (g/m3) 0.73 0.69 0.79 0.78 0.81

H/C Ratio 1.90 2.25 3.00 2.66 2.50

O2 (% weight) 0 0 35 31 22



6/12

0.12 bar i.e. at the conservative end of the part-load limits used bycar manufacturers. The lower pumping loop area in Fig. 6a and

higher valve lift illustrated inFig. 6b clearly demonstrate the re-duced throttling at the valves themselves. In turn, fuel economywas improved by an additional 3%. The problem of throttling atthe valves is due to the need to maintain acceptable kinematicforces, which is ultimately the reason for the S-shaped curve in

Fig. 6b.

3.2. EGR tolerance effects

One objective was to quantify how ethanol and butanol fuelsinfluence part-load EGR tolerance, fuel economy and emissionswhen used in tandem with EIVC operation. The worst case engineoperating condition for SI engine EGR tolerance is normally engine

idle, where the low speed and load lead to minimum in-cylinder

gas pressures and temperatures, reduced in-cylinder charge mo-tion levels, decreased burning velocities and (relatively) high cylin-der wall heat transfer losses. Hence the purpose of these tests was

to set the engine to EIVC operation at idle and then attempt to in-crease the valve overlap until the acceptable combustion stabilitylimit was approached. The fuels tested included gasoline, E25g75,E100, Bu25g75 and Bu100 fuel and the fuel injection timing re-

mained fixed to allow complete injection on to the back of a warmclosed valve, with the End of Injection (EOI) timing at 400bTDC.The experiments were performed at warm idle engine conditions(850 rpm/1.85 bar IMEPn, 90 C coolant) with Maximum Brake

Torque (MBT) spark timing used at all sites. Throughout the testsopen loop fuelling control was employed, with the relative air-to-fuel ratio (k) allowed to vary within the mapping limits of

0.981.00. Hence the engine was nominally rich of stoichiometric(k= 0.99) at the reported sites. The relatively high load for idlewas necessary to sustain engine speed due to the relatively highfriction of the single cylinder unit.

Set out inFig. 7are key results. During the low valve overlapcondition tests the overlap remained set at 15crank as illustratedin Fig. 7a. Such overlap levels can be considered to be an acceptableminimum for a modern SI engine [37]. In Fig. 7b it can be seen that

for the gasoline, butanol and E25 fuels it was not possible to signif-icantly increase the valve overlap as the engine was already oper-ating near the acceptable combustion stability limit. However,with higher ethanol content more overlap was possible, increased

to 33 crank with E100 fuel and resulting in small additional ISFCsavings of 2.7% compared to EIVC-only operation. One reason

for the improved fuel consumption was reduced pumping work,with values of Pumping Mean Effective Pressure (PMEP) shown

in Fig. 7d. This benefit was again associated with reduced throttlingat the inlet valve itself, with the valve lift increased from 1.64 mm

to 1.84 mm from low to high overlap with E100 fuel (hence push-ing the lift further away from the tail of the S-shaped valve open-ing curve inFig. 5).

Shown inTable 3is a summary of key net indicated thermalefficiency values for the four valvetrain settings (where the base-

line refers to throttled operation with a fixed timing valvetrain).The E100 fuel exhibited slightly higher values of thermal efficiency,which may have been partly associated with higher H/C ratio andratio of specific heats compared to the non-oxygenated baseline.

The influence of faster rates of mass burning on thermal efficiencymust also be considered, with values of 010% mass fractionburned duration set out inFig. 7e. Equivalent 1090% values weresimilar for all cases, ranging between 21and 23crank. The obser-

vation of faster burning with near-stoichiometric levels of ethanol

is in good agreement with laminar burning velocity correlations inthe literature [42,43] and also previous observations by the authorsof improved part-load EGR tolerance during conventional throttled

SI operation[37]. Ultimately, due to such effects, small fuel con-sumption benefits were noted with E100 but these obviously onlyprovided a minor offset compared to the high calorific deficit ofsuch fuel. Otherwise it was clear that the internal EGR tolerance

of a gasoline fuel is not effected by 1-butanol content, regardlessof butanol blend volume. Shown in Figs. 7fh are the effects ofthe fuels and varied overlap settings on the engine-out emissions.Ethanol exhibits lower adiabatic flame temperature compared to

gasoline [3436]. However, the faster rate of mass burning maystill lead to higher gas temperatures in the remaining unburnedcharge, which together with potential improved global oxygen dis-

tribution was in agreement with the higher values of NOx mea-sured as the alcohol content was increased. Despite theseobservations, the use of high ethanol content fuels at idle enabledmoderate reductions (up to 20%) in engine-out NOxto be made,

which was primarily associated with the ability to run higher valveoverlap. For emissions of CO, a trade-off between alcohol contentand EGR level was apparent. Values of CO and unburned hydrocar-bons (uHC) are well-known to increase with internal EGR under

part-load conditions, the result of the EGR reducing peak in-cylinder gas temperatures. This is related to reduced heat releaseper unit mass, dissociation effects and lower combustion efficiency[14,37,38]. However, some caution is recommended when

reviewing absolute unburned hydrocarbon levels of alcohol fuelsmeasured via the standard Flame Ionisation Detection method

(calibrated in the automotive industry with non-oxygenatedhydrocarbons). This is because the oxygenates reduce the response

Fig. 6. (a) Logp logVdiagram comparing EIVC and EIVC + VVT operation and (b) inlet valve operating curve with the corresponding points of operation superimposed.



7/12

of the unit in direct proportion to the oxygen content in the fuel, aspreviously reported by the current authors[37,38]and elsewhere[44]. Nonetheless, for each ethanol fuel, increasing the valve over-

lap reduced the combustion efficiency (with qualitative reductionof 1.7% combustion efficiency for E100 fuel, for example, when

calculating the inefficiency based upon the measured engine-outemissions).

(b)(a)

(d)(c)

(f)(e)

(h)(g)

ValveOverlap[degCA]

10

15

20

25

30

35

Alcohol content [%]

0 20 40 60 80 100

IMEP

n[bar]

0.070

0.080

0.090

0.100

0.110

0.120

0.130

Alcohol content [%]

0 20 40 60 80 100

Minimum overlap (ethanol)Highest possible overlap (ethanol)Highest possible overlap (butanol)

ISFC

Benefit[%]

9.5

10.0

10.5

11.0

11.5

12.0

12.5

13.0

Alcohol content [%]

0 20 40 60 80 100

0-10

%MFB[degCA]

34

3536

37

38

39

40

41

42

Alcohol content [%]

0 20 40 60 80 100

uHC[ppmC]

5200

5400

5600

5800

6000

6200

6400

Alcohol content [%]

0 20 40 60 80 100

PMEP[bar]

0.36

0.38

0.40

0.420.44

0.46

0.48

0.50

0.52

0.54

0.56

Alcohol content [%]

0 20 40 60 80 100

CO[%]

0.65

0.70

0.75

0.80

0.85

0.90

0.95

1.00

Alcohol content [%]

0 20 40 60 80 100

NOx[ppm]

750

800

850

900

950

1000

Alcohol content [%]

0 20 40 60 80 100

Fig. 7. Key performance and emissions measurements during fixed low valve overlap and maximum overlap tests with varied alcohol content fuels (hot idle engine

conditions).

Table 3

Comparison of net indicated thermal efficiencies for the different valvetrain

conditions considered (850 rpm/1.85 bar IMEPn).

Fuel Baseline (%) VVT-only (%) EIVC-only (%) EIVC + VVT (%)

95 RON ULG 29 29.2 32.5 32.5

E100 29.4 31 32.8 33.7



8/12

3.3. Fuel effects (2000 rpm/6 bar IMEP)

Another key objective of the work was to quantify the effects of

n-butanol and ethanol on variable valvetrain settings at moderateengine speeds and loads where the engine is still partially throttledbut less sensitive to EGR tolerance. The tests were therefore per-formed at 2000 rpm/6 bar IMEPn to represent load where the en-

gine was still throttled in baseline low valve overlap mode (withan intake plenum pressure of 0.62 bar when using gasoline). Forall tests the engine was operated with MBT spark timing and nearstoichiometric (k= 0.99) fuelling levels, with the fuel injection tim-

ing again fixed at EOI = 400bTDC. The valve overlap settingsadopted are identified below inTable 4. The corresponding intakevalve settings for the EIVC cases are set out in Table 5and thesesettings allowed full unthrottled operation. In terms of fuels, for

ethanol 25% alcohol was either splash blended with the 95 RONunleaded gasoline (E25g75) or with the isooctane (E25i75). Inthe case of n-butanol, an isooctane blend was investigated(Bu25i75), equivalent in terms of volume percent alcohol but obvi-ously then incurring a lower oxygen mass weighting due to the

higher carbon count.Set out inFig. 8 are key engine performance and combustion

characteristics for all cases. Firstly, a common prior observationof EIVC operation is that the early intake valve closure slows down

the combustion [8]. Observing Fig. 8ac this prior observation heldtrue but, at this elevated speed and load, it was interesting to notethat the increase in combustion duration under EIVC-only condi-

tions (Case 3) was actually less than that during VVT-only opera-tion (Case 2). When then combining EIVC and VVT the totalcombustion duration increased substantially, with significantly ad-vanced spark timing required. This observation insinuates that theinternal EGR rate was increased although the combustion re-

mained of remarkably similar stability. Such preservation of com-bustion stability when increasing the internal EGR is quite typicalof SI engines[37], where the COV of IMEP remains reasonably flatuntil a certain EGR rate is reached, after which an exponential-like

deterioration is observed due to excessive combustion duration.Shown inFig. 8d is the angle of peak in-cylinder pressure. The

optimum phase in this engine at these conditions was found tobe 14aTDC. Some of the variation shown was believed to be

mainly due to the resolution and control of the ignition (the reso-

lution of the ignition timing was 0.75). Due to the arising oscilla-tion in this phase, corresponding variation is apparent in theaveraged peak in-cylinder pressure and combustion stability.

In terms of fuel effects, the ethanol again resulted in marginallyfaster burning rates compared to gasoline, whereas the isooctanewas slightly slower. The isooctane also produced higher PMEPand lower indicated fuel consumption than the commercial gaso-

line case. These iso-octane effects remained apparent when com-paring the two 25% ethanol cases. Such results help highlight thewell-known difficulties with use of a single component as a surro-gate for gasoline but this was considered a necessary limitation in

order to allow comparisons with both prior and future optical en-gine results; with optical work currently underway at Brunel toquantify flame stretch effects with such fuels and varying levelsof EGR. Finally, the butanol blend would be expected to incur lower

ISFC than its ethanol equivalent due to being matched on a volumebasis and hence incurring lower oxygen mass content and highercalorific value. The PMEP for the butanol case was remarkably sim-

ilar to the gasoline and lower ethanol content fuel.It is well known that combining unthrottled operation with EGR

has the ability to minimise pumping losses and reduce peak heattransfer rates in the SI engine [610]. However, with the valvetrain

system studied other fuel economy benefits may be gained. Thisfact is apparent in Fig. 8 when comparing the PMEP, gross andnet ISFC values for gasoline in Cases 3 and 4. InFig. 8g little extra

reduction in PMEP was produced when additionally adopting VVT.However, inFig. 8h the net ISFC continues to improve, ultimatelydue to improvement in gross ISFC. To understand these effects itis useful to consider thepVdiagrams for Cases 3 and 4 when usinggasoline (as an example of this common observation) as shown in

Fig. 9. The valve timing events are identified on the figure, wherethe hollow markers denote Case 3 events, solid markers Case 4,squares indicate exhaust events and dots intake valve events. Thefigure clearly demonstrates an increase in useful expansion work

with the retarded exhaust event. However, the trade-off is in-creased pumping work during the early part of the exhaust stroke,where the exhaust valves initially throttle the escaping gas. Apartfrom this, the area enclosed within the pumping loop is still lower

during the remainder of the breathing which suggests further de-

throttling at the intake valves did actually occur albeit not detect-able in the PMEP values (due to the new exhaust pumping loss).Otherwise, reductions in compression loss [610] were also

apparent as marked on the figure but these were small in compar-ison to the other effects. From these results it is clear the ability toincrease valve overlap during EIVC operation still pays dividends atmoderate loads and is not only associated with reduced throttling

at the valves but the ability to further improve the high pressurepart of the four-stroke engine cycle.

Finally, shown inFig. 10 are the corresponding measured en-gine-out emissions. At these elevated loads there are less gains

to be had in terms of reducing pumping losses, as such losses nat-urally decrease with load in the SI engine. Comparing Cases 1 and 2for gasoline, the exhaust-to-inlet pressure ratio (PR) reduced from

PR= 1.65 to PR= 1.3 at maximum valve overlap. For VVT-only oper-ation, engine-out emissions of NOx were reduced by 35%. It isinteresting to note that Case 2 provided the highest NOxreduction

Table 4

Summary of the test cases.

Case number Valve conditions

1 Low overlap (15crank) and maximum lift

2 High overlap (63crank) and maximum lift3 Low overlap (15crank) and EIVC

4 High overlap (63crank) and EIVC

Table 5

Summary of intake valve settings during the EIVC cases with low and high valve overlap.

Fuel Case number 3 Case number 4

Inlet duration (crank) Inlet lift (mm) Inlet duration (crank) Inlet lift (mm)

95 RON ULG 125 3.04 149 4.87I100 123 2.89 146 4.61

E25i75 124 2.96 147 4.64

E25g75 122 2.83 145 4.52

Bu25i75 126 3.11 148 4.77



9/12

of all three cases compared to the baseline. During unthrottled en-

gine conditions the exhaust-to-inlet pressure ratio drops to unity.It is this pressure ratio that drives internal EGR, where a higher va-

lue corresponds to greater backflow of residuals into the intakemanifold. As a result, the ability to drive increased EGR and mini-

mise NOx diminishes when using EIVC (for a given valve overlap

setting). The observation of maximum NOx reduction using aVVT-only strategy is in good agreement with previously published

speed-load maps when using gasoline in this engine [45]. However,from the current work it would also appear that the observation

0-10%

MFB[crank]

24

26

28

30

32

34

36

38

Case Number

1 2 3 4

95 RON ULGi100E25i75E25g75Bu25i75

10-90%

MFB[crank]

17

18

19

20

21

22

23

Case Number

1 2 3 4

An

gleofPmax[aTDC]

13.4

13.6

13.8

14.0

14.2

14.4

14.6

Case Number

1 2 3 4

ISFC[g/kW.h

]

200

210

220

230

240

250

260

Case Number

1.0 1.5 2.0 2.5 3.0 3.5 4.0

I

MEP[bar]

0.0350.040

0.045

0.050

0.055

0.060

0.065

0.070

0.075

Case Number

1 2 3 4

Pmax[bar]

31.532.0

32.5

33.0

33.5

34.0

34.5

35.0

35.5

36.0

36.5

37.0

Case Number

1 2 3 4

PMEP[bar]

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.55

Case Number

1.0 1.5 2.0 2.5 3.0 3.5 4.0

SparkTiming[bTDC]

28

30

32

34

36

38

40

42

44

46

Case Number

1 2 3 4

(b)(a)

(d)(c)

(f)(e)

(h)(g)

Fig. 8. Key performance parameters obtained at 2000 rpm/6 bar IMEPn including (a) 010% mass fraction burned (b) 1090% mass fraction burned (c) ignition timing (d)

angle of peak in-cylinder pressure (e) peak in-cylinder pressure (f) standard deviation in gross IMEP (g) PMEP and (h) Indicated Specific Fuel Consumption (solid linesare net

values, dashed line gross for gasoline only). Thex-axis refers to the case numbers indentified inTable 4.



10/12

holds true regardless of the fuel types studied. Of course, it could

be argued that this result was subjective and that the NOx mayhave been further reduced by the adoption of wider range camphasers. However the ranges used in the current work were typicalof current production technology. Otherwise, in terms of fuel ef-

fects it was interesting to note relatively similar emissions forthe moderate alcohol level ethanol and butanol blends. Again theiso-octane indicated marginally lower NOx and CO which mayagain have been indicative of the increased throttling compared

to gasoline (and hence lower peak pressures and temperatures).

4. Conclusions

The effects of combining EIVC with internal EGR were studiedduring port fuel injection of various alcohol blended fuels in a ther-

modynamic single cylinder research engine. Under warm idle en-gine conditions the following conclusions were made:

When adopting higher ethanol content fuels reasonable

improvements in residual gas tolerance can be made. Atthese lowest load conditions this allows increased valveoverlap to be tolerated which, with such a fully variablevalvetrain, reduces the throttling of the gases at the

intake valves themselves during EIVC operation. Forexample, when using E100 fuel, the overlap was increasedfrom 15 to 33 crank, resulting in small additional ISFCsavings of 2.7% compared to EIVC-only operation. Such

improvement was at least partially associated with thefaster laminar burning velocities of ethanol and arisingfaster rates of mass burning.

It would appear that the EGR tolerance of a gasoline fuel is

not influenced by 1-butanol content, regardless of thepercentage of butanol within the blend

The use of high ethanol content fuels at idle enabled mod-erate reductions (up to 20%) in engine-out NOx to be

EVO

IVC

IVO

EVCIn-cylinderPressure[bar]

0

5

10

15

20

25

30

35

40

In-cylinder Volume [cm3]

50 100 150 200 250 300 350 400 450 500 550

Case 3 (EIVC)Case 4 (EIVC+VVT)

Increasedexh pumping

Increasedexpansion

Reducedcompression

lossIn-cylinderPressure[bar]

0

1

2

3

4

In-cylinder Volume [cm3]

50 100 150 200 250 300 350 400 450 500 550

(a) (b)

Fig. 9. (a) In-cylinder pressurevolume (pV) diagram for the average in-cylinder pressure data and (b) zoomed-in view on the pumping loop with key valve events shown.

(b)(a)

(c)

CO[%

]

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

Case

1 2 3 4

NOx[p

pm]

1500

2000

2500

3000

3500

4000

4500

Case

1 2 3 4

uHC[ppmC]

1200

1300

1400

1500

16001700

1800

1900

2000

2100

Case

1 2 3 4

95 RON ULGi100E25i75E25g75Bu25i75

Fig. 10. Engine out emissions of (a) NOx(b) CO and (c) unburned hydrocarbons (2000 rpm/6 bar IMEPn).



11/12

made, which was primarily associated with significantincrease in the residual mass tolerated.

For all ethanol fuels, increasing the valve overlap reducedthe combustion efficiency (with qualitative reduction of1.7% combustion efficiency for E100 fuel, for example,when calculating the inefficiency based upon the mea-sured engine-out emissions).

Under moderate speed and load conditions (2000 rpm/6 bar IM-EPn) the following additional conclusions were made:

At these higher loads the combination of VVT and EIVCstill facilitated good reduction in throttling at the intakevalves at all valve lifts studied. At the most retarded valveoverlap conditions such throttling was minimised, how-

ever the late opening of the exhaust valves resulted inincreased exhaust pumping losses early on during theexhaust stroke. It can therefore be concluded that, whenemploying a retarded valve overlap strategy, a compro-

mise exists between throttling at the intake versus atthe exhaust valves.

The benefit of increasing valve overlap during EIVC oper-

ation at moderate loads is not only associated withreduced valve throttling but also the ability to furtherimprove the high pressure part of the cycle, withincreased expansion work and reduced compressionlosses noted when using the retarded valve overlap

strategy. During unthrottled engine conditions the exhaust-to-inlet

pressure ratio drops to unity. As a result, the ability todrive increased EGR and minimise NOxdiminishes when

using EIVC at higher loads. For maximum NOxreduction,VVT-only operation proved to offer the highest benefit,regardless of the type of alcohol blended in the fuel atup to 25% volume. The observation of minimum NOxwith

a VVT-only strategy is in good agreement with compre-

hensive speed-load maps previously produced forgasoline.

In terms of fuel blending effects, the ethanol again

resulted in faster burning rates compared to gasoline,whereas the iso-octane was slightly slower. The iso-octane also produced higher PMEP and lower indicatedfuel consumption than the commercial gasoline case.

These differences remained apparent when comparingsimilar levels of alcohol blended with gasoline or iso-octane. Such results help highlight the well-known diffi-culties with use of a single component as a surrogate for

gasoline but this was considered a necessary limitationin order to allow comparisons with planned opticalengine tests.

Acknowledgements

Thanks to Hermann Hoffman and Oliver Fritz from MAHLEGmbH and Neil Fraser (formerly MAHLE Powertrain) for theirsupport.

References

[1] Kramer U, Phlips P. Phasing strategy for an engine with twin variable camtiming. Society of Automotive Engineers (SAE) paper no. 2002-01-1101; 2002.

[2] Moro D, Minelli G, Serra G. Thermodynamic analysis of variable valve timinginfluence on SI engine efficiency. SAE paper no. 2001-01-0667; 2001.

[3] Roberts CE, Stanglmaier RH. Investigation of intake timing effects on the coldstart behaviour of a spark ignition engine. SAE paper no. 1999-01-3622; 1999.

[4] CairnsA, Irlam G, Blaxill H. Exhaust gas recirculation for improvedpart andfull

load fuel economy in a turbocharged gasoline engine. SAE paper no. 2006-01-0047; 2006.

[5] Schfer J, Balko J. High performance electric camshaft phasing system. SAEpaper no. 2007-01-1294; 2007.

[6] Tuttle J. Controlling engine load by means of late intake valve closing. SAEpaper no. 800794; 1980.

[7] Tuttle J. Controlling engine load by means of early intake valve closing. SAEpaper no. 820408; 1982.

[8] Cleary D, Silvas G. Unthrottled engine operation with variable intake valve lift,duration and timing. SAE paper no. 2007-01-1282; 2007.

[9] Hannibal W, Flierl R, Stiegler L, Meyer R. Overview of current continuouslyvariable valve lift systems for four-stroke spark-ignition engines and the

criteria for their design ratings. SAE paper no. 2004-01-1263; 2004.[10] Sauer C, Kulzer A, Rauscher M, Hettinger A. Analysis of different gasoline

combustion concepts with focus on gas exchange. SAE paper no. 2008-01-0427; 2008.

[11] Allen J, Law D. Advanced combustion using a lotus active valve train: internalexhaust gas recirculation promoted auto-ignition. In: Proceedings of the IFPinternational congress; 2001. p. 85100.

[12] Koopmans L, Strom H, Lundgren S, Backlund O, Denbratt I. Demonstrating a SI-HCCI-SI mode change on a Volvo 5-cylinder electronic valve control engine.SAE paper no. 2003-01-0753; 2003.

[13] Unger H, Schneider J, Schwarz C, Kock K-F, VALVETRONIC experience from 7years of series production and a look into the future. In: 29th InternationalWiener Motorensymposium; 2008.

[14] Shimizu K, Fuwa N, Yoshihara Y, Hori K. The new toyota variable valve timingand lift system. Aachener Kolloquium Fahrzeug und Motorentechnik; 2007.

[15] Sellnau M, Kunz T, Sinnamon J, Burkhard J. 2-Step variable valve actuation:system optimisation and integration on an SI engine. SAE paper no. 2006-01-0040; 2006.

[16] Stansfield PA, Wigley G, Garner CP, Patel R, Ladommatos N, Pitcher G, et al.Unthrottled engine operation using variable valve actuation: the impact on theflow field, mixing and combustion. SAE paper no. 2007-01-1414; 2007.

[17] Patel R, Ladommatos N, Stansfield P, Wigley G, Garner P, Pitcher G., et al.Comparison between unthrottled, single and two-valve induction strategiesutilising direct gasoline injection: emissions, heat-release and fuelconsumption analysis. SAE paper no. 2008-01-1626; 2008.

[18] Palma A, Del Core D, EspositoC. The HCCI concept and control, performedwithmultiair technology on gasoline engines. SAE paper no. 2011-24-0026; 2011.

[19] WhiteTL. Alcohol asa fuel forthe automobilemotor. SAEpaperno.070002;1902.[20] Schelp H. Alcohol blends in gasoline. SAE paper no. 360058; 1936.[21] Porter JC. Alcohol as a high octane fuel. SAE paper no 510086; 1951.[22] Kremer FG, Fachetti A. Alcohol as automotive fuel Brazilian experience. SAE

paper no. 2000-01-1965; 2000.[23] Griffin WM, Lave LB,MacLeanHL. Promise andcostof cellulosic ethanol for the

U.S. light-duty fleet. SAE paper no. 2001-01-2474; 2001.[24] West BH, Lopez AJ,Theiss TJ, Graves RL, StoreyJM, Lewis SA. Fuel economyand

emissions of the ethanol-optimised Saab 95 BioPower. SAE paper no. 2007-01-3994; 2007.

[25] Bergstrm K, Melin S-A, Jones C. The new ECOTEC turbo BioPower engine fromGM powertrain utilising the power of natures resources. In: 28thinternationales wiener motorensymposium; 2007.

[26] Dohmel K. Future mobility from a fuels perspective. In: 29th internationaleswiener motorensymposium; 2008.

[27] Turner J, Pearson R, Purvis R, Dekker E, Johansson K, Bergstrm K. GEMternaryblends: removing the biomass limit by using iso-stoichiometric mixtures ofgasoline, ethanol and methanol. SAE paper no. 2011-24-0113; 2011.

[28] Turner J, Pearson R, McGregor M, Ramsay J, Dekker E, Iosefa B, et al. GEMternary blends: testing iso-stoichiometric mixtures of gasoline, ethanol andmethanol in a production flex-fuel vehicle fitted with a physical alcoholsensor. SAE paper no. 2012-01-1279; 2012.

[29] Lumsden G, OudeNijeweme D, Fraser N, Blaxill H. Development of aturbocharged direct injection downsizing demonstrator engine. SAE Int JEngines 2009;2(1):142032.

[30] OudeNijeweme D, Stansfield P, Bisordi A, Bassett M, MAHLE Powertrain UK,Williams J, et al. Significant CO2reductions by utilising the synergies betweena downsized SI engine and biofuels. In: IMechE performance, fuel economyand emissions conference. London; 2011.

[31] Kapus P, Fuerhapter A,Fuchs H,FraidlGK. Ethanoldirect injectionon turbochargedSI engines potential and challenges. SAE Paper No. 2007-01-1408; 2007.

[32] Brewster S. Initial development of a turbocharged direct injection e100combustion system. SAE paper no. 2007-01-3625; 2007.

[33] Kar K, Last T, Haywood C, Raine R. Measurement of vapor pressures andenthalpies of vaporisation of gasoline and ethanol blends and their effects onmixture preparation in an SI engine. SAE Int J Fuels Lubr 2009;1(1):13244.

[34] Gautam M, Martin DW. Combustion characteristics of higher alcohol/gasolineblends. Proc Instn Mech Eng Part A 2000;214.

[35] Yacoub Y, Bata R, Gautam M. The performance and emission characteristics ofC1C5 alcohol-gasoline blends with matched oxygen content in a singlecylinder spark ignition engine. Proc Inst Mech Eng Part A 1998;212.

[36] Bata RM, Elrod C, Lewandowski TP. Butanol as a blendingagent forI.C. engines.SAE paper no. 890434; 1989.

[37] Cairns A, Stansfield P, Fraser N, Blaxill H. A study of gasolinealcohol blendedfuels in an advanced turbocharged DISI engine. SAE Int J of Fuels Lubr2009;2(1):4157.

[38] Cairns A, Todd A, Fraser N, Aleiferis P, Malcolm J. A study of alcohol blended

fuels in an unthrottled single cylinder spark ignition engine. SAE paper no.2010-01-0618; 2010.



12/12

[39] Brusstar M, Stuhldreher M, Swain D, Pidgeon W. High efficiency and lowemissions from a port injected engine with neat alcohol fuels. SAE paper no.2002-01-2743; 2002.

[40] Lancefield T, Lawrence T, Ahmed A, Hamouda HBH. VLD a flexible, modular,cam operated VVA system giving variable valve lift and duration andcontrolled secondary valve openings. SIA conference on variable valveactuation. IFP Rueil; 2006.

[41] Bunsen E, Grote A, Willand J, Hoffmann H, Fritz O, Senjic S. Fuel consumptionimprovements by variation of intake valve lift and timing a mechanicallyfully variable valvetrain system for a 1.6l gasoline engine with direct injection.

Haus der technik variable valvetrain congress; 2007.

[42] Glder L. Burning velocities of ethanolisooctane blends. Combust Flame1984;56:2618.

[43] Beeckmann J, Rohl O, Peters N. Experimentaland numerical investigation of iso-octane, methanol and ethanol regarding laminar burning velocity at elevatedpressure and temperature. SAE paper no. 2009-01-1774; 2009.

[44] Dec JE, Davisson ML, Sjberg M, Leif RN, Hwang W. Detailed HCCI exhaustspeciation and the sources of hydrocarbon and oxygenated hydrocarbonemissions. SAE Int J Fuels Lubr 2009;1(1):5067.

[45] Cairns A, Todd A, Aleiferis P, Malcolm J. A comparison of inlet valve operatingstrategies in a single cylinder spark ignition engine. In: IMechE performance,

fuel economy and emissions conference. London; 2009.


Documents

Bai 2. Qua trinh -ánh l-a --ng c- có acohol.pdf