Atkinson - Miller

Energy Conversion and Management 79 (2014) 59–65

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier .com/ locate /enconman

The Miller cycle effects on improvement of fuel economy in a highlyboosted, high compression ratio, direct-injection gasoline engine: EIVCvs. LIVC

0196-8904/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.enconman.2013.12.022

⇑ Corresponding author. Address: School of Mechanical Engineering, ShanghaiJiao Tong University, 800 Dong Chuan Rd., Shanghai 200240, PR China. Tel./fax: +8621 3420 3749.

E-mail address: [email protected] (T. Li).

Tie Li ⇑, Yi Gao, Jiasheng Wang, Ziqian ChenSchool of Mechanical Engineering, Shanghai Jiao Tong University, PR China

a r t i c l e i n f o

Article history:Received 18 October 2013Accepted 5 December 2013Available online 29 December 2013

Keywords:Fuel economyGasoline enginesIntake boostingMiller cycleKnockPumping loss

a b s t r a c t

A combination of downsizing, highly boosting and direct injection (DI) is an effective way to improve fueleconomy of gasoline engines without the penalties of reduced torque or power output. At high loads,however, knock problem becomes severer when increasing the intake boosting. As a compromise, geo-metric compression ratio (CR) is usually reduced to mitigate knock, and the improvement of fuel econ-omy is discounted. Application of Miller cycle, which can be realized by either early or late intakevalve closing (EIVC or LIVC), has the potential to reduce the effective CR and suppress knock. In this paper,the effects of EIVC and LIVC on the fuel economy of a boosted DI gasoline production engine reformedwith a geometric CR of 12.0 are experimentally compared at low and high loads. Compared to the originalproduction engine with CR 9.3, at the high load operation, the brake specific fuel consumption (BSFC) isimproved by 4.7% with CR12.0 and LIVC, while the effect of EIVC on improving BSFC is negligibly small. Atthe low load operation, combined with CR12.0, LIVC and EIVC improve the fuel economy by 6.8% and7.4%, respectively, compared to the production engine. The mechanism behind the effects of LIVC andEIVC on improving the fuel economy is discussed. These results will be a valuable reference for enginedesigners and researchers.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Requirement for better fuel economy and reduction in exhaustgas emissions in automobiles are driving the introduction of a widerange of new gasoline engine technologies. One of the most prom-ising technologies is the downsized, highly boosted, direct-injec-tion spark-ignition (DISI) engine [1–7]. In automotive gasolineengines, the most frequent operations are at low loads and the loadcontrol is usually implemented by using a throttle valve to restrictairflow into cylinders, which results in pumping loss during gas ex-change strokes. With keeping the engine torque output constant, adecrease in the engine displacement will lead to a shift in the typ-ical engine operating range to higher loads, which will help reducethe pumping loss and thus the fuel consumption. To meet vehiclerequirements for the maximum power output, intake boost is gen-erally necessary for downsized engines at high loads, causingknock a more severe problem compared with their naturallyaspirated (NA) counterparts. A reduction in geometric compressionratio (CR) is the most typical solution to mitigate the knock prob-

lem in downsized gasoline engines at high loads, but this strategyis at the expanses of reduced fuel economy benefit from enginedownsizing.

A strategy of shortening compression stroke relative to expan-sion stroke, called Miller cycle, is effective to reduce the tempera-ture and pressure at the end of compression stroke, and thereforesuppress knock. Miller cycle can be realized by either early or lateintake valve closing (EIVC or LIVC). The original Miller engine em-ployed a compression control valve on the cylinder head to releasepart of the charge to the exhaust port during the compressionstroke to reduce the effective CR [8,9]. With the thermodynamicanalysis of ideal Miller cycle, Wu et al. [10] concluded that theMiller cycle has no inherent efficiency advantage over the Otto cy-cle because of the shorten compression stroke. However, theyemphasized that the engine knock problem of the Otto enginecould be suppressed by the lowered temperature and pressure ofthe air–fuel mixture before combustion with LIVC or EIVC, whichshould be beneficial for supercharging. Analysis of air-standardMiller cycle with considerations of heat transfer loss, friction andvariable specific heats of working fluid demonstrated that Millercycle is more efficient than Otto cycle and the effects of heattransfer loss, friction and specific heat ratio on fuel efficiency aresignificant and should be considered in practical cycle analysis[11–14].

Nomenclature

A constant (–)a constant (–)b constant (–)Cv isochoric specific heat (J/(mol K))E activity energy (J)m molar number (mol)ne engine speed (rpm)P pressure (bar)Q apparent heat release (J)R universal gas constant (J/(mol K))Rc the ratio of connecting rod length to crank radius (–)T temperature (K)U internal energy (J)V volume (cm3)Vc the clearance volume (cm3)X parameter related to fuel or mixture properties (–)e compression ratio (–)c ratio of specific heat (–)h crank angle (CAD ATDC)

s ignition delay of end gas (ms)

AcronymsATDC after top dead centerBMEP brake mean effective pressure (MPa)BSFC brake specific fuel consumption (kg/kW h)CA50 combustion phasing defined by the crank angle of 50%

accumulative heat release (�CA)CAD crank angle degreeCR compression ratio (–)DISI spark-ignited direction-injectionEIVC early intake valve closingLIVC late intake valve closingMBT maximum brake torqueNA naturally aspiratedPMEP pumping mean effective pressure (MPa)ROHR rate of heat releaseTDC top dead center

60 T. Li et al. / Energy Conversion and Management 79 (2014) 59–65

In practical spark-ignition engines, as aforementioned, pumpingloss significantly limits the engine efficiency at low loads, whileknock restricts the compression ratio and spark advances that re-sult in low thermal efficiency at high loads [15–17]. Cleary et al.[18] investigated the part-load fuel consumption potential ofunthrottled operation of a single cylinder engine using the EIVCvariable valve actuation. They reported that a 7% fuel consumptionimprovement is achieved through the optimization of the intakevalve lift, duration and timing while maintaining the conventionalexhaust valve event. Not as they expected, however, the enhancedin-cylinder charge motion owing to the unthrottled operation withEIVC did not lead to significant combustion performance improve-ment. Miklanek et al. [19] pointed out that due to the shortenedcompression stroke with Miller cycle, the lowered in-cylindercharge temperature at the beginning of the combustion may leadto a slower combustion speed compared to a standard Otto cycle,which could negatively influence the engine efficiency. They pro-posed a strategy combining Miller cycle and mixture heating, andwith one dimensional simulation they demonstrated a significantimprovement in the fuel economy. Anderson et al. [20] conductedthe first and second law analysis and compared a NA, Miller cycle,SI engine with LIVC to a conventionally-throttled Otto cycle coun-terpart at light load conditions. The first law analysis revealed thatas much as 6.3% improvements in the indicated thermal efficiencyare obtained with the Miller cycle, and these improvements areprimarily due to reductions in pumping and compression workwhile heat transfer losses are comparable. The second law analysisshowed that the throttling process in the conventional engine de-stroys up to 3% of the fuel availability and that the higher pressurein the LIVC intake manifold leads to a notable thermo-mechanicaladvantage over the conventional throttled engine.

In the original design of Miller engine, a supercharger was at-tempted to remedy the inadequacy of power output due to theshortened compression stroke [8,9]. From another viewpoint, theshortened compression stroke can be actively utilized to mitigateknock in boosted or high CR engines. In the development of MazdaV6 Miller cycle engine equipped with a LIVC mechanism and a Lys-holm supercharger, the ram-pulsation effects of the Miller cyclewere well utilized to enhance the anti-knocking performance[21–23]. Wang et al. [24] examined the effects of the Miller cycleon a high CR SI engine, and concluded that LIVC with a high valvelift should be implemented at high loads to reduce the effective CR

and knock tendency while EIVC with a low valve lift should beadopted at low loads to improve the thermal efficiency. Gottschalket al. [25] investigated the effect of the Miller cycle with EIVC onknock control in a turbocharged, downsized DISI engine at the mid-dle loads. They demonstrated that the Miller cycle strategies offerhigh potential in cyclic variations of combustion, fuel economy andexhaust gas temperature.

From the above paragraphs, it can be seen that there have beenextensive researches on Miller cycle including both theoretical cy-cle analysis and experimental studies on practical engines. A com-bination of Miller cycle and modern highly boosted downsizedengine should have great potentials in further improvement of fueleconomy in SI engines. Either EIVC or LIVC is the most practical ap-proach to realize the Miller cycle, thanks to technical advances invariable valve actuations and electronic controls. However, a sys-tematic comparison of the performances of EIVC and LIVC onimproving the fuel economy of a modern highly-boosted, down-sized DI gasoline engine with high CR at low and high loads israrely reported, although such information is very valuable for en-gine designers.

In this study, the Miller cycle effects of EIVC and LIVC on the fueleconomy of a boosted DI gasoline production engine modified witha geometric CR of 12.0 are experimentally compared at low andhigh loads, meanwhile the mechanism of these effects is discussed.The details are reported as follows.

2. Experimental approaches and procedure

A 2.0 liter 4-cylinder DISI engine was operated at low and highload conditions: 2000 rpm 0.4 MPa brake mean effective pressure(BMEP) and 1000 rpm 1.32 MPa BMEP. Here the condition of2000 rpm 0.4 MPa BMEP is one of the frequently operating condi-tions of the engine, representative of typical partial load in thisstudy, while 1.32 MPa BMEP is wide open throttle (WOT) conditionat 1000 rpm for the engine, and it was used as a representative oflow-speed high-load operation in this study. Engine performancewas compared at the geometric compression ratios (CR) of 9.3and 12.0 configured with the base camshaft and CR12.0 with theEIVC and LIVC camshafts. A summary of the engine specificationsare provided in Table 1, where the rated power and torque are ta-ken from the original production engine. The lift profiles of the in-

Table 1Engine specifications, fuel and test conditions.

Cylinder number In-line 4Bore � stroke Ø86 � 86 mmDisplacement 1998 cm3

Connecting rod length 145.5 mmCrank radius 43.0 mmCompression ratio 9.3, 12.0Intake and exhaust systems Duel VVTValve number 16Fuel delivery system Direct injectionFuel #97 gasoline (RON97)Rated powera 190 kW @ 5300 rpmRated torquea 353 Nm @ 2000 � 5250 rpmIntake boosting Turbocharger + superchargerTest conditions 1000 rpm 1.32 MPa BMEP

2000 rpm 0.4 MPa BMEPExcess air ratio 0.9 at 1000 rpm 1.32 MPa BMEP

1.0 at 2000 rpm 0.4 MPa BMEP

a For the original production engine.

Lift

Crank Angle

Base CamLIVC CamEIVC Cam

TDC BDC

Fig. 1. A schematic of valve lift profiles for the different camshafts.

T. Li et al. / Energy Conversion and Management 79 (2014) 59–65 61

take valve for the original, EIVC and LIVC camshafts are schemati-cally shown in Fig. 1. LIVC is obtained by extending the duration ofthe maximum lift of the base camshaft, while EIVC is realized witha reduced lift compared to the original profile. The high-pressureinjectors were located below the two intake ports (side injection)with an operating pressure up to 15.0 MPa. Both a superchargerand a turbocharger were used to provide sufficient intake air quan-tity to achieve the desired torque levels when using different cam-shafts. ETAS hardware and INCA software were used tocommunicate with the engine controller to optimize the engineperformance at each operating condition. The commercial #97 gas-oline (research octane number 97) in Chinese market was used asfuels for all the tests.

Thrott

Dyno

ExhaustTurbine

Fig. 2. A schematic of experimental

Fig. 2 shows a schematic of the experiment0al setup. An AVL ACdynamometer system was operated in speed-control mode tomaintain the desired engine speed. The engine coolant and oil tem-peratures were maintained at 90 ± 1.5 �C using the conditioningsystems. An inlet air conditioning system along with an intercoolerconditioning system was used to maintain the temperature,humidity, and pressure of the inlet air at desired values. Becausethe temperature at the compressor outlet varied in a range of44–52 �C at the low load and 73–112 �C at the high load for the dif-ferent strategies, an in-house made intercooler was employed tokeep the intake temperature at 34 ± 1.0 �C for all the tests in thisstudy. Accelerometer-type knock sensors were used to monitorknocking combustion. A fuel flow meter (AVL733S) with a preci-sion of 0.12% and a high-precision (0.01%) torque sensor (HBM)were utilized to measure the fuel flow rate, _mf and the torque, Tand then the brake specific fuel consumption (BSFC) can be calcu-lated as

BSFC ¼_mf

T � 2pne=60¼ 30 _mf

T � pneð1Þ

where ne is the engine speed in revolution per minute.Before each test, at first, the engine was well conditioned by

warming-up operation from low to high loads for 30 min to ensurethe data repeatability. Then, after the engine was set at the desiredoperating conditions, the engine was optimized in terms of fueleconomy by tuning the fuel injection timing and pressure, the ex-cess air ratio as well as the phasing of intake and exhaust valveevents. Finally, spark sweeps were carried out, and the points ofknock limit at the high load operation and maximum brake torque(MBT) at the low load operation were identified and the data wererecorded by the sampling system.

Piezoelectric in-cylinder pressure transducers (Kistler 6125A)with flame arrestors were flush-mounted through the cylinderhead between the intake and exhaust valves to collect the in-cylin-der pressure. The pressure signals were processed using the chargeamplifiers (Kistler 5064B) and a high-speed data acquisition sys-tem at a resolution of 0.2 crank angle degrees (CAD) triggered bya crankshaft encoder. For an internal combustion engine with amechanism of connecting rod and crank shaft, the cylinder volumeV at any crank position hcan be written as

V=Vc ¼ 1þ 12ðe� 1Þ Rc þ 1� cos h� ðRc

2 � sin2 hÞ1=2h i

ð2Þ

where Vc is the clearance volume, e the compression ratio and Rc theratio of connecting rod length to crank radius.

Based on the first law of thermodynamics, the released heat atthe crank angle h is

Fresh air

Supercharger

Intercooler

compressor

le valve

Air cleaner

setup for the engine test bench.

BSFC

(g/k

w·h

)

5

CR 9.3 CR 12.0 CR 9.3 CR 12.01000 rpm

1.32 MPa BMEP2000 rpm

0.4 MPa BMEP

2.2%

4.5%

Fig. 3. Effect of the higher compression ratio (CR) on the brake specific fuelconsumption (BSFC) at the high and low load operating conditions.

0

5

10

15

20

25

30

35

CR 9.3 CR 12.0

CA5

0 (C

AD A

TDC

)

Fig. 4. Effect of the higher compression ratio on the combustion phasing (CA50) atthe knock limited point (ne = 1000 rpm, BMEP = 1.32 MPa).

62 T. Li et al. / Energy Conversion and Management 79 (2014) 59–65

dQdh¼ dU

dhþ P

dVdh¼ d

dhðmCvTÞ þ P

dVdh

ð3Þ

where U is the internal energy, m the molar number, T the averagetemperature of the working gas, and Cv the isochoric specific heatgiven by

Cv ¼R

c� 1ð4Þ

where R is gas constant and c the specific heat ratio that can be cal-culated using the data in the JANAF table [26,27]. Combining the Eq.(4) and the equation of state of ideal gas PV = mRT into Eq. (3), therate of heat release is obtained as

dQdh¼ 1

c� 1V

dPdhþ P

dVdh

� �� PV

ðc� 1Þ2dcdh

ð5Þ

BSFC

(g/k

w·h

)

5

CR 9.3Base Cam

CR 12.0Base Cam

CR 12.0LIVC

CR 12.0EIVC

2.2%

6.9%

2.0%

Fig. 5. Effects of EIVC and LIVC on improving BSFC at the knock limited point underthe high load condition (ne = 1000 rpm, BMEP = 1.32 MPa).

3. Results and discussion

3.1. Effect of higher compression ratio on fuel economy

Tests were conducted at geometric compression ratios of 9.3and 12.0 with the base camshaft to compare the effect of imple-menting higher CR piston on the fuel economy. An operating con-dition of 2000 rpm 0.4 MPa bar in brake mean effective pressure(BMEP) was chosen to represent the part load and 1000 rpm1.32 MPa BMEP for the high load.

Fig. 3 shows the fuel consumption results for the geometric CRsof 9.3 and 12.0 at the MBT point for the low load condition and theknock limited point for the high load condition. A 4.5% improve-ment in the brake specific fuel consumption (BSFC) is observedwhen increasing the geometric CR from 9.3 to 12.0 for the low loadcondition.

Although the higher CR improves the part-load fuel consump-tion, knock becomes severer at the high load condition. HereCA50, the crank angle where 50% of the inducted fuel energy hasbeen released, was used to characterize the engine’s knock perfor-mance. When increasing CR from 9.3 to 12.0, the engine becomesmore knock limited, where CA50 is retarded by about 11.0 CADat 1000 rpm 0.4 MPa BMEP engine operating condition, as shownin Fig. 4. As a result, as shown in Fig. 3, compared to CR of 9.3,implementation of the CR 12.0 increases BSFC by 2.2% at the knocklimited spark timing (KLSA) due to the CA50 penalty.

Knock is one of the major constraints to improve the thermalefficiency of spark-ignition (SI) engines. Moderate knock, usuallydefined by the ‘‘clanging’’ or ‘‘pinging’’ sound, increases the enginevibration and noise, while heavy knock may directly lead to the en-gine failure [28]. It is generally accepted that knock is attributed tothe oscillation of pressure wave resulted from the rapid heat

release of auto-ignited unburned mixture (so called end gas) aheadof the propagating flame front. The ignition delay s of end gas canbe expressed as an Arrhenius-type correlation [29–31].

s ¼ AðXÞaPb expE

RT

� �ð6Þ

where A, a and b are constants, X is parameters related to fuel ormixture properties, and E is related to activity energy. It is clearfrom Eq. (6) that an increase in the intake pressure leads to higherP of the end gas, increasing the knock propensity, and a higher CRdeteriorates the knock problem by increasing both P and T. There-fore, despite the potential of fuel economy improvement at lightloads, a relatively low CR is usually employed for boosted engine.

3.2. Effects of EIVC and LIVC at high loads

Shown in Fig. 5 are the effects of EIVC and LIVC on improvingBSFC at the knock limited point under the high load condition.For comparison purpose, the data at the same operating conditionin Fig. 3 are also plotted. In comparison to the case of base cam-shaft and CR 12.0, both EIVC and LIVC result in significant improve-ments of BSFC. Particularly, BSFC is improved by 6.9% by using LIVCcompared to the case of base camshaft and CR 12.0, and it is evenbetter than the case of base camshaft and CR 9.3. However, noimprovement in BSFC is observed by using the EIVC strategy, com-pared to the original production engine of CR 9.3. Here it should bementioned that the intake valve closure was swept in a range of161–191 and 231–251 CAD ATDC during gas exchange strokesfor EIVC and LIVC, respectively, and the data in Fig. 5 are the opti-mized results in terms of the intake camshaft phasing for the low-est fuel consumption.

Fig. 6 shows the effects of EIVC and LIVC on the combustionphasing (CA50) at the knock limited point under the high load con-dition. Compare to the case of base camshaft and CR 12.0, the com-bustion phasing is advanced by 3 crank angle degrees (CAD) for

CA5

0 (C

AD A

TDC

)

CR 9.3Base Cam

CR 12.0Base Cam

CR 12.0LIVC

CR 12.0EIVC

0

5

10

15

20

25

30

35

Fig. 6. Effects of EIVC and LIVC on the combustion phasing (CA50) at the knocklimited point under the high load condition (ne = 1000 rpm, BMEP = 1.32 MPa).

0.5

1

2

4

8

16

32

64

Cyl

inde

r pre

ssur

e (M

Pa)

Volume (cm3)

Base Cam CR12.0LIVC Cam CR12.0EIVC Cam CR12.0

0.05

0.1

0.2

0.4

0.8

1.6

3.2

6.4

40 80 160 320 640

Fig. 7. P–V diagrams of the whole engine cycle with EIVC and LIVC at the knocklimited point under the high load condition (ne = 1000 rpm, BMEP = 1.32 MPa).

Cyl

inde

r pre

ssur

e (M

Pa)

Volume (cm3)

LIVC Cam CR12.0EIVC Cam CR12.0

Intake valve closure

40 80 160 320 6400.07

0.14

Fig. 8. P–V diagrams of the exhaust and intake strokes of the engine with EIVC andLIVC at the knock limited point under the high load condition (ne = 1000 rpm,BMEP = 1.32 MPa).

0

500

1000

1500

2000

2500

-30 0 30 60 90

Cyl

inde

r tem

pera

ture

(K)

Crank Angle (ATDC)

Base Cam CR12.0LIVC Cam CR12.0EIVC Cam CR12.0

Fig. 9. History of in-cylinder averaged temperature for various camshafts at theknock limited point under the high load condition (ne = 1000 rpm,BMEP = 1.32 MPa).

T. Li et al. / Energy Conversion and Management 79 (2014) 59–65 63

EIVC, while it is advanced by up to 10 CAD for LIVC to a value com-parable to the case of CR 9.3 and base camshaft. While both strat-egies suppress the engine knock, LIVC exhibits higher potentials inthe suppression of knock than EIVC, despite the higher effective CRfor LIVC than for EIVC as will be shown below.

Figs. 7 and 8, respectively, show the P–V diagrams of the wholecycle and during the intake and compression strokes of the enginewith EIVC and LIVC at the knock limited point under the operatingcondition of 1000 rpm 1.32 MPa BMEP. In this study, the effectiveCR is defined as the volume ratio of the in-cylinder charge at themoment of intake valve closure and at the top dead center (TDC).As shown in Fig. 7, despite the higher effective CR of 9.9 for LIVCthan the CR 9.3 for EIVC, the pressure at the end of compressionstroke for LIVC is lower than for EIVC. This may be attributedmostly to the heat transfer from the cylinder to the charge. ForEIVC, since the intake valve closes early before the bottom deadcenter (BDC), the charge is actually expanded as the piston moves

downward. As a result, there would be significant heat transferfrom the hot cylinder wall to the charge, which can be deducedfrom the pressure trace after BDC, as shown in Fig. 8. For LIVC,the pressure at the closure of intake valve is higher than for EIVC.However, for LIVC, since the charge is continually inducted untilthe valve closure after BDC, both the temperature and heat capac-ity of the charge around BDC could be higher than for EIVC. There-fore, there could be less heat transfer from the cylinder to thecharge and thus the smaller temperature rise for LIVC than forEIVC. In addition, after BDC some charge would be pushed outthe cylinder by the piston upward motion owing to the late valveclosure, taking away some heat from the cylinder. As a conse-quence, as shown in Figs. 7 and 9, both the pressure and tempera-ture during the compression stroke are lower for LIVC than forEIVC, resulting in the better anti-knock performance for LIVC thanfor EIVC as have been shown in Fig. 6.

Note that in Fig. 8 while the pressure during the intake strokebefore the valve closure is similar for both EIVC and LIVC, the pres-sure during the exhaust stroke for LIVC is significantly lower thanfor EIVC. This may be attributed to the fact that LIVC advanced thecombustion phasing as has been shown in Fig. 6, reducing the pres-sures during the expansion stroke, after the blow down as well asduring the exhaust stroke as can be seen in Fig. 7. As a result, LIVCleads to a larger pumping work than EIVC.

Fig. 10 shows the comparison of pumping work for the differentcases at the knock limited point under the high load condition. Thedata in the case of base camshaft and CRs of 9.3 and 12.0 are alsoplotted in the figure for reference. To maintain the load of 1.32 MPaBEMP at 1000 rpm, the intake charge was aggressively boosted sothat the intake pressure is higher than the exhausting pressure.Therefore, here the pumping work plays a positive role in the fueleconomy. In comparison to the base camshaft, while EIVC reducesthe pumping work, LIVC significantly increases the pumping work.

Therefore, here the better performance of LIVC in improvementof fuel consumption over EIVC can be attributed to the advancedcombustion phasing as well as the increased pumping work.

3.3. Effects of EIVC and LIVC at low loads

Fig. 11 shows the effects of EIVC and LIVC on improving BSFC atMBT point under the operating condition of 2000 rpm 0.4 MPa barBMEP. Both EIVC and LIVC exhibit good performance: a 2.9% reduc-tion in BSFC for EIVC and a 2.3% reduction for LIVC are obtainedcompared to the case of base camshaft and CR 12. Compared tothe case of base camshaft and CR 9.3, the improvements are 7.2%and 6.7% for EIVC and LIVC, respectively. Here the better fuel econ-omy performance of EIVC than LIVC is consistent with the earlierstudies [24,32].

PEM

P (M

Pa)

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

CR 9.3Base Cam

CR 12.0Base Cam

CR 12.0LIVC

CR 12.0EIVC

Fig. 10. Effects of EIVC and LIVC on the pumping work at the knock limited pointunder the high load condition (ne = 1000 rpm, BMEP = 1.32 MPa).

BSFC

(g/k

w·h

)

5

CR 9.3Base Cam

CR 12.0Base Cam

CR 12.0LIVC

CR 12.0EIVC

6.8%

4.5%

7.4%

Fig. 11. Effects of EIVC and LIVC on improving BSFC at the MBT point under the lowload condition (ne = 2000 rpm, BMEP = 0.4 MPa).

Cyl

inde

r pre

ssur

e (M

Pa)

Volume (cm3)

Base Cam CR12.0LIVC Cam CR12.0EIVC Cam CR12.0

Intake valve closure

40 80 160 320 6400.04

0.08

0.16

Fig. 12. P–V diagrams of the exhaust and intake strokes of the engine with EIVC andLIVC at the MBT point under the low load condition (ne = 2000 rpm,BMEP = 0.4 MPa).

PEM

P (M

Pa)

-0.05

-0.045

-0.04

-0.035

-0.03

-0.025

CR 9.3Base Cam

CR 12.0Base Cam

CR 12.0LIVC

CR 12.0EIVC

Fig. 13. Effects of EIVC and LIVC on pumping work at the MBT point under the lowload condition (ne = 2000 rpm, BMEP = 0.4 MPa).

Rat

e of

hea

t rel

ease

(J/d

eg)

Norm

alized cumulative heat release (%

)

0102030405060708090100

0

5

10

15

20

25

30

35

40

-60 -40 -20 0 20 40 60Crank angle degree (CAD ATDC)

CR9.3 Base CamCR12 Base CamCR12 LIVCCR12 EIVC

Fig. 14. Effects of EIVC and LIVC on rate of heat release at the MBT point under thelow load condition (ne = 2000 rpm, BMEP = 0.4 MPa).

64 T. Li et al. / Energy Conversion and Management 79 (2014) 59–65

Fig. 12 shows the P–V diagrams during the intake and compres-sion strokes of the engine with EIVC and LIVC at the MBT point un-der the conditions of 2000 rpm 0.4 MPa BMEP. Here the effectiveCRs are 9.7 and 11.4 for EIVC and LIVC, respectively. It should bepointed out that the variable valve timing mechanisms were usedfor both the intake and exhaust systems, and here the valve tim-ings have been optimized for the best fuel economy for the differ-ent camshafts. At the low load operation, without the intake boostand with the throttle, the in-cylinder pressure during the intakestroke is lower than that during compression stroke. Therefore,the pumping work plays a negative role in the fuel economy.

Fig. 13 shows the effects of EIVC and LIVC on the pumping effec-tive mean pressure (PMEP) at 2000 rpm 4.0 bar BMEP. The EIVCstrategy results in the smallest absolute PMEP, followed by LIVC,and the base camshafts for the two CRs shows similar performance.An explanation can be found from Fig. 12. With EIVC, a relativelylarge open angle of the intake throttle valve could be implemented,

resulting in relatively higher intake pressure. With the similarpressure during the exhaust stroke, the higher intake pressurewould lead to a less pumping loss. Therefore, the reduction ofpumping loss for EIVC should be one of the most important con-tributors to the improved fuel economy. The reason for the supe-rior performance of EIVC over LIVC is not yet fully understood,and one explanation goes to the higher lift for LIVC that could re-quire more throttled operation to restrict the air flow and thus re-sult in more pumping loss than for EIVC.

With EIVC, besides the reduced pumping loss, improved fuel–air mixing and enhanced in-cylinder turbulence strength were re-ported, which could contribute to combustion [18,32]. As shown inFig. 14, however, EIVC shows the lowest rate of heat release, fol-lowed by LIVC, despite the almost identical CA50 for all the cases.As mentioned in reference [19] and shown in Fig. 12 in this study,the lowered pressure and temperature at the end of compressionstroke could be the reason for the slower rate of heat release inthe case of EIVC. However, also shown in Fig. 12, LIVC leads to ahigher pressure during the compression stroke than the cased ofbase camshaft and CR12.0. Here the reason is not clear at presentand further study is needed to address this issue.

4. Conclusions

Part- and high-load evaluations have been conducted with re-spect to the performance of EIVC and LIVC with a high CR of 12.0for the boosted DISI engine. The results are summarized as follows:

1. At the part-load operation, a fuel consumption improvement isobtained with increasing the geometric CR from 9.3 to 12.0.However, at the high-load condition, the higher CR increasesknock tendency and the fuel consumption deteriorates.

T. Li et al. / Energy Conversion and Management 79 (2014) 59–65 65

2. Improved anti-knock performance is achieved when imple-menting either EIVC or LIVC strategies for the CR 12.0 at thehigh load condition. A fuel economy even better than the caseof base camshaft and CR9.3 is obtained with a combination ofCR12 and LIVC.

3. LIVC is advantageous over EIVC in the fuel consumptionimprovement at the high load, primarily owing to the betterknock resistance and greater pumping work of LIVC than EIVC.

4. At the low load, the fuel economy can be improved with eitherEIVC or LIVC, primarily owing to the decreased pumping loss.

5. The reduction in the pumping loss with EIVC is greater thanwith LIVC. However, the rate of heat release for EIVC is slowerfor LIVC, probably owing to the lowered pressure and tempera-ture at the end of compression stroke.

Acknowledgments

The Supports by Doctoral Fund of Ministry of Education of Chi-na (20120073120059), Shanghai Pujiang Program (13PJ1404300),and the Recruitment Program of Global Youth Experts of ChineseGovernment are gratefully acknowledged.

References

[1] Lecointe B. Monnier G. Downsizing a gasoline engine using turbo-chargingwith direct injection. SAE Paper; 2003 [2003-01-0542].

[2] Pallotti P, Torella E. New application of an electric boosting system to a small,four-cylinder S.I. engine. SAE Paper; 2003 [2003-32-0039].

[3] Lake T, Stokes J, Murphy R, Osborne R, Schamel A. Turbo-charging concepts fordownsized DI gasoline engines. SAE Paper; 2004 [2004-01-0036].

[4] Königstein A, Hock C, Frensch M. Comparison of advanced turbochargingtechnologies under steady-state and transient conditions. SAE Paper; 2006[2006-05-0364].

[5] Zaccardi JM, Pagot A, Vangraefschepe F. Optimal design for a highly downsizedgasoline engine. SAE Paper; 2009 [2009-01-1794].

[6] Alger T, Gingrich J, Roberts C, Mangold B. Cooled exhaust-gas recirculation forfuel economy and emissions improvement in gasoline engines. Int J Engine Res2011;12:252–64.

[7] Li T, Wu D, Xu M. Thermodynamic analysis of EGR effects on the first andsecond law efficiencies of a boosted spark-ignited direction-injection gasolineengine. Energy Convers Manage 2013;70:130–8.

[8] Miller R. Supercharge and internally cooling for high output. ASME Trans1947;69:453–64.

[9] Miller R. Supercharged engine. Patent US2817322 A; 1957.[10] Wu C, Puzinauskas PV, Tsai JS. Performance analysis and optimization of a

supercharged Miller cycle Otto engine. Appl Therm Eng 2003;23:511–21.[11] Ge Y, Chen L, Sun F, Wu C. Effects of heat transfer and friction on the

performance of an irreversible air-standard miller cycle. Int Commun Heat &Mass Trans 2005;32(8):1045–56.

[12] Hou SS. Comparison of performances of air standard Atkinson and Otto cycleswith heat transfer considerations. Energy Convers Manag2007;48(5):1683–90.

[13] Lin JC, Hou SS. Performance analysis of an air-standard Miller cycle withconsiderations of heat loss as a percentage of fuel’s energy, friction andvariable specific heats of working fluid. Int J Therm Sci 2008;47(2):182–91.

[14] Al-Sarkhi A, Jaber JO, Probert SD. Efficiency of a Miller engine. Appl Energy2006;83:343–51.

[15] Kutlar OA, Arslan H, Calik AT. Methods to improve efficiency of four stroke,spark ignition engines at part load. Energy Convers Manage2005;46(20):3202–20.

[16] Alkidas AC. Combustion advancements in gasoline engines. Energy ConversManage 2007;48:2751–61.

[17] Caton JA. The thermodynamic characteristics of high efficiency, internal-combustion engines. Energy Convers Manage 2012;58:84–93.

[18] Cleary D, Silvas G. Unthrottled engine operation with variable intake valve lift,duration, and timing. SAE Paper; 2007 [2007-01-1282].

[19] Miklanek L, Vitek O, Gotfryd O, Klir V. Study of unconventional cycles(Atkinson and Miller) with mixture heating as a means for the fuel economyimprovement of a throttled SI engine at part load. SAE Int J Engines2012;5(4):1624–36.

[20] Anderson M, Assanis D, Filipi Z. First and second law analyses of a naturally-aspirated, Miller cycle, SI engine with late intake valve closure. SAE Paper;1998. p. 980889.

[21] Choshi M, Asanomi K, Abe H, Okamoto S, Shoji M. Development of V6 Millercycle engine. JSAE Rev 1994;15(3):195–200.

[22] Hatamura K, Hayakawa M, Goto T, Hitomi M. A study of the improvementeffect of Miller-cycle on mean effective pressure limit for high-pressuresupercharged gasoline engines. JSAE Rev 1997;18(2):101–6.

[23] Iwata N, Miyagoshi K, Nakatani S, Hitomi M. Improvement of anti-knockingperformance by supercharged Miller-cycle engine – ram-pulsation effects onanti-knocking performance by Miller-cycle. JSAE Rev 1995;16(3):313–8.

[24] Wan Y, Du A. Reducing part load pumping loss and improving thermalefficiency through high compression ratio over-expanded cycle. SAE Paper;2013 [2013-01-1744].

[25] Gottschalk W, Lezius U, Mathusall L. Investigations on the potential of avariable Miller cycle for SI knock control. SAE Paper; 2013 [2013-01-1122].

[26] Lemmon E, McLinden M, Friend D. Thermo Physical Properties of FluidSystems. NIST, Gaithersburg, MD. <Http://webbook.nist.gov.entro> [accessedon 18/10/2012].

[27] Kroenlein K, Muzny C, Kazakov A, Diky V, Chirico R, Magee J, Abdulagatov I,Frenkel M. NIST/TRC Web Thermo-Tables (WTT), Technical report version 2-2011-4-Pro, NIST; 2011. <Http://wtt-pro.nist.gov/wtt-pro> [accessed on 18/10/2012].

[28] Heywood JB. Internal combustion engine fundamentals. McGraw-Hill; 1988.[29] Livengood JC, Wu PC. Correlation of autoignition phenomena in internal

combustion engines. In: 5th Symp (Int) on Combust; 1955. p. 347–56.[30] Soylu S. Prediction of knock limited operating conditions of a natural gas

engine. Energy Convers Manage 2005;46(1):121–38.[31] Verhelst S, Sheppard CGW. Multi-zone thermodynamic modeling of spark-

ignition engine combustion – an overview. Energy Convers Manage2009;50(5):1326–35.

[32] Diana S, Iorio B, Giglio V, Police G. The Effect of valve lift shape and timing onair motion and mixture formation of DISI engines adopting different VVAactuators. SAE Paper; 2001 [2001-01-3553].