2007-01-1855

8/12/2019 2007-01-1855

1/9

JSAE 20077035SAE 2007-01-1855

Operation strategy of a Dual Fuel HCCI Engine with VGT

Carl Wilhelmsson, Per Tunest al and Bengt JohanssonDivision of Combustion Engines, Faculty of engineering, Lund University

Copyright c 2007 Society of Automotive Engineers of Japan, Inc. and Copyright c 2007 SAE International

ABSTRACT

HCCI combustion is well known and much results re-garding its special properties have been published.Publications comparing the performance of differentHCCI engines and comparing HCCI engines to con-ventional engines have indicated special features ofHCCI engines regarding, among other things, emis-sions, efciency and special feedback-control require-ments. This paper attempts to contribute to the com-mon knowledge of HCCI engines by describing an op-erational strategy suitable for a dual-fuel port-injectedHeavy Duty HCCI engine equipped with a variable ge-ometry turbo charger. Due to the special properties ofHCCI combustion a specic operational strategy hasto be adopted for the engine operation parameters (inthis case combustion phasing and boost pressure).The low exhaust temperature of HCCI engines limitsthe benets of turbo charging and causes pumpinglosses which means that the more the merrier prin-ciple does not apply to intake pressure for HCCI en-gines. It is desirable not to use more boost pressurethan necessary to avoid excessively rapid combustionand/or emissions of N Ox . It is also desirable to select

a correct combustion phasing which, like the boostpressure, has a large inuence on engine efciency.The optimization problem that emerges between theneed for boost pressure to avoid noise and emis-sions and, at the same time, avoiding an extensive de-crease of efciency because of pumping losses is thetopic of this paper. The experiments were carried outon a 12 liter Heavy Duty Diesel engine converted topure HCCI operation. Individually injected natural gasand n-Heptane with a nominal injection ratio of 85%natural gas and the rest n-Heptane (based on heat-

ing value) was used as fuel. The engine was underfeedback combustion control during the experiments.

INTRODUCTION

Despite intense research efforts in recent years theHCCI engine is still less known than its relatives theOtto and the Diesel engine. Nevertheless it seems tobe one of the most promising engine concepts of thefuture, combining high efciency with ultra low emis-sions of nitrogen oxides. The rst report of the con-cept of HCCI combustion was presented by Onishi etal. [1] and was primarily an attempt to reduce emis-sions of unburned hydrocarbons (HC) and improvepart load efciency for two stroke engines. After On-ishi et al. published their results it took a while beforefurther HCCI publications emerged. When they didthe publication rate increased. Important early publi-cations are for example [2]-[4].

The working principle of the HCCI engine is best un-derstood as a hybrid between the Otto and Diesel en-gines. The HCCI engine operates with a premixedcharge as the Otto engine, and it operates unthrottled

with compression ignition, controlling the load with theglobal air/fuel equivalence ratio ( ) like the diesel en-gine. This operation principle makes it possible toincrease the efciency compared to the Otto enginedue to the avoidance of throttling losses. At the sametime the high soot and nitrogen oxide emissions of thediesel engine are avoided. Soot emission is avoidedbecause of the homogeneity of the mixture and theabsence of locally rich combustion zones. Nitrogenoxide emission is avoided because of the decreasedpeak in-cylinder temperature due to the diluted oper-

-356-

THIS DOCUMENT IS PROTECTED BY U.S. AND INTERNATIONAL COPYRIGHTIt may not be reproduced, stored in a retrieval system, distributed or transmitted, in whole or in part, in any form or by any means.

Downloaded from SAE International by Brought to by the J. Robert Van Pelt Library / Michigan Technological Univ. , Thursday, January 24, 2013 05:32:57 PM

8/12/2019 2007-01-1855

2/9

ation of the engine and the absence of stoichiometricundiluted zones.

HCCI combustion can be achieved in numerous ways,both in two and four stroke engines. High compres-sion ratio can be applied [5], the inlet air can be preheated [6]. HCCI combustion can be induced by un-conventional valve strategies that retain hot residuals

[7] and the octane number of the fuel can be altered[8] to modulate the ignition temperature. Since theHCCI combustion process in many operating pointsis unstable, feedback combustion control is neededto operate an HCCI engine in parts of its operatingrange. Such combustion control can be performed innumerous ways using different actuators and sensors.

Even though it has many good features, the HCCI en-gine also has some limitations besides the, just de-scribed, need for feedback combustion control. Theoperational principle unfortunately suffers from veryhigh combustion rates, causing noise as well as wearof engine hardware. Another issue with the HCCIprinciple is low combustion efciency at low load.This causes high emissions of unburned hydrocar-bons and carbon monoxide.

Besides the development of the actual HCCI principlemuch work has been carried out addressing the greattask of designing a feedback combustion control sys-tem for the HCCI engine. The highly non-linear na-ture of HCCI combustion imposes a great challengeto researchers in the eld and many interesting resultshave been published [8]-[12]. Many of these publica-

tions share the property of model assisted controllers,i.e. a model of the combustion process is maintainedin the control system and it is continuously updated toreect the state the engine. Such models are used bythe controllers to predict adequate control outputs.

This work addresses the control issue from a strategyperspective rather than from a single control variableor model perspective. A setup (described in [8]) wasalready available to the authors and slight modica-tions to this setup were carried out. The main fuel wasaltered to Natural Gas (NG). N-Heptane was addedin small quantities as an ignition improver. Besidesthe change of fuels the turbo charger was altered,instead of the turbocharger used by Olsson a Vari-able Geometry Turbocharger was used. The changeof turbo is to adopt the setup for NG operation. NG,from an HCCI perspective, is a poorly suited fuel! Aspreviously well known, NG has a very high octanerating (the exact octane number depends on the ori-gin of the gas) typically close to 120 which representsboth RON and MON for methane (NG contains mainlymethane). The high octane rating of NG makes it verydifcult to compression ignite. When combustion is

initiated on the other hand the rate of heat releaseduring the combustion event is explosive. The reasonfor the high rate of heat release is left for the chem-istry and the kinetics people to explain, but empiri-cally this is a sound fact. For successful operation itis absolutely essential to operate an NG-fuelled HCCIengine leaner than if the same engine were operatedwith e.g. ethanol. Successful, in this case, means

free of NOx emissions and with a low enough pres-sure derivative to avoid noise and engine damage.At the same time the exhaust gas temperature of anHCCI engine is very low, which means that it is difcultto obtain high boost pressure from the turbo withoutadding pumping losses. An operating strategy thatprovides enough boost pressure to avoid noise andemissions, and at the same time avoids excessive ef-ciency penalty due to pumping losses is hence de-sirable. One such strategy will be developed in thispaper.

EXPERIMENTAL APPARATUS

EXPERIMENTAL ENGINE

The engine system that was the basis for the inves-tigation is presented in this section. It should bepointed out that numerous results have been pre-sented from the same setup before, by for exampleOlsson et al. [8]. Since the last publication by Ols-son et al. the laboratory setup has however beensomewhat modied, for example the turbo chargerhas been altered and the fuel selection has beenchanged. The fuel combination has been changedfrom n-heptane/ethanol to n-heptane/NG. This affectsthe performance of the engine, particularly the engineout N Ox emissions and maximum achievable engineload are affected by the altered fuel properties.

The engine system used in these experiments (seeFigure 1) is based on a 12 l turbo charged Sca-nia Euro III Heavy Duty Diesel engine originally in-tended for truck applications. The geometrical prop-erties are accounted for through Table 1. The enginewas converted to a port fuel injected HCCI enginethrough the removal of the direct injectors and modi-

cations of the intake in order to mount sequential portfuel injectors. The engine is run on two different fuels,n-Heptane and Danish NG. Two independent fuel sys-tems were needed in order to supply the engine withthe two fuels and combustion phasing was controlledthrough the injection ratio of the different fuels. Eachfuel was, for practical reasons, introduced in a sep-arate port. This does, however, not have any majorsignicance to the combustion according to previousresults by Olsson et al. [8].

-357-



8/12/2019 2007-01-1855

3/9

Figure 1: The test engine setup used in the experi-ments.

Furthermore the standard turbo was changed fromthe standard diesel sized turbo to a Variable Geom-etry Turbine (VGT) turbo charger There is a need fora VGT turbo in order to supply adequate boost pres-sure regardless of engine load without inferring exten-sive pumping losses at the high load operation points.Upstream the turbo the charge air is either cooled orheated depending on the current need for intake aircondition.

The engine was, of course, fully instrumented withcylinder individual cylinder pressure sensors (CPS),thermocouples and pressure sensors in exhaust portsand at various locations in the intake system. Theengine out emissions of O2 ,CO,CO 2 , NO x and NO

were all measured with relevant instruments.

The system used for control of the engine was a PCbased system accounted for and developed by Ols-son et al. [8]. The software was written in-housein Pascal/Delphi and implements an operator graph-ical interface which allows the user to monitor theengine operation as well as command fuel ratio, to-tal fuel amount and intake heating/cooling. The con-trol variables can be commanded open-loop by theoperator, or by feedback controllers (gain scheduledPID controllers) controlling the combustion phasing bychanging the cylinder individual ratios of n-Heptane toNG. Other variables that can be feedback controlledare intake temperature (by changing the power to theelectric heaters) and IMEP n (through the total fuelamount).

RESULTS AND DISCUSSION

The rst part of the results deals with engine opera-tion with (almost) no boost pressure applied, we callthis Minimum Intake Pressure (MIP) operation. Dur-ing MIP operation the VGT was, of course, set for min-

Table 1: Engine geometric properties.Number of cylinders 6Displaced Volume 11705 cm3

Bore 127 mmStroke 154 mmConnecting Rod Length 255 mmNumber of Valves 4Compression Ratio 18 : 1Exhaust valve open 82 BBDCExhaust valve close 38 ATDCInlet valve open 39 BTDCInlet valve close 63 ABDCFuel Supply dual PFI

imum boost pressure which also results in minimumexhaust back pressure. The reason for performingMIP tests was to provide reference data to comparethe boosted operation with. MIP operation data wereobtained prior to developing a boost strategy and con-clusions drawn from the MIP data set were kept inmind when developing the boost strategy.

The engine setup used in this study has a large num-ber of operating parameters and it is hence essentialto apply a strategy for selecting these parameters inorder to operate the engine in a good way and re-duce the number of number of experiments that haveto be performed. It is important to remember that themain objective with the VGT is to extend the operatingrange of the engine and the boosting strategy servesas a tool for selecting boost pressure to maximize ef-ciency and minimize emissions of NOx throughout

the extended operating range.

In order to limit the number of pages in this papereach explanation or statement can not be motivatedby a diagram.

LIMITATIONS

In order to dene the operational area of the en-gine limits on certain operational parameters were se-lected. An NOx emission level of 0.2 g/kWh wasregarded as the maximum tolerated, based on theUS2007 emission regulations. Peak Cylinder Pres-sure (PCP) was limited to 180 bar , ad-hoc ,. This is adesign parameter set by the engine manufacturer and180 bar is a reasonable PCP in HD applications. Maxpressure derivative was limited to 15 bar/CAD , againad-hoc , based on the emission of noise of the actualengine used. Fuel composition was limited as well.N-Heptane is more expensive than NG and the tar-get consumption of n-heptane thus has to be limited.Since the experiments were run in a multi cylinder en-gine the fuel ratio condition had to be within a range,due to cylinder to cylinder variations and variation

-358-



8/12/2019 2007-01-1855

4/9

Table 2: limitations of engine operation.CA 50max = arg max nte (CA50)

P CP 180 bardP/dCAD (PCDP ) 15 bar/CAD

NO x 0.2 g/kWh80/ 20 Q(NG )/Q (n Heptane ) 85/ 15

T in 180C

over time of the combustion. The fuel ratio was henceconditioned to be within 80% - 85% NG based on theheating value (the rest is n-Heptane). N-Heptane isnot a commercially viable fuel. The fuel limitation ishowever still valid based on the assumption that youhave one main fuel which is difcult to compressionignite (NG, ethanol, methanol or gasoline) by whichmost of the energy is supposed to be delivered. Tothe main fuel a fuel which is easy to compression ig-nite (DME, Diesel or n-Heptane) is added. The taskof the secondary fuel is hence not to supply a ma-

jor part of the energy, only to promote ignition. In acase where the secondary fuel is prized same as theprimary fuel user convenience could be one reasonto limit the usage of the secondary fuel. A situationwhere the secondary fuel is supplied at service onlywould e.g. be ideal. The intake temperature should,due to limitations of the hardware, not exceed 180 C .Finally a limitation of the combustion phasing, CA50,was employed. In this work the crank angle degreewhere 50 % of the heat release has occurred (CrankAngle 50, CA50) is used as a measure of combus-tion phasing. CA50 should be set for Maximum netIndicated Efciency (MIE). The reason for not choos-

ing the standard way of determining best CA50 fromthe torque (Maximum Brake Toque, MBT) is that thetorque sensor did not accurately enough resolve thedifference in torque close to MIE. The set of condi-tions (rules of engine operation) are summarized inTable 2.

OPERATION WITH MINIMUM BOOST

Figure 2 shows net indicated efciency as a functionof CA50 for all speeds and loads during MIP opera-tion. From the gure it is clear that there, for eachload but the lowest one, exists a certain CA50 thatyields maximum indicated efciency, i.e. the MIEpoint. MIE occurs later and later in the cycle as theengine load increases. There are two phenomenathat, when combined, result in this behavior. From anideal-cycle point of view it would be benecial to havethe combustion occurring as close to Top Dead Cen-ter (TDC) as possible. This represents the best trade-off of minimized compression work without sacric-ing expansion work. However as combustion movescloser to TDC the convective heat losses to the cylin-der walls increase. The reason for increased heat

loss is twofold. Both the maximum in-cylinder temper-ature and the maximum pressure derivative increasewhen the combustion moves closer to TDC. An in-creased pressure derivative causes pressure ringingin the cylinder, breaking down the thermal bound-ary layer close to the cylinder walls, which in turn in-creases the convective heat loss even further. Thosetwo phenomena are responsible for the distinct max-

imum efciency at a certain CA50 for all operationpoints but the very lowest load points. The efciencyat low load actually increases monotonically as thecombustion is phased earlier. The reason for this be-havior is that combustion efciency is very sensitiveto combustion phasing at low load. The increase ofmaximum temperature with earlier combustion phas-ing causes a drastic increase of the combustion ef-ciency. Since the low loads normally have very poorcombustion efciency, the effect of increased com-bustion efciency is the most powerful phenomenonand hence we never nd the local maxima for the low-est load points.

It is well known that combustion phasing is a very im-portant parameter for the formation of NOx (throughpeak temperature). Figure 3 shows NOx emissionsfor the same operating points presented in Figure 2.It is easily noticed that the two major variables thatdetermine the NOx emissions are engine load andCA50. NOx emissions increase with decreasingCA50 and increasing load. From Figure 4, showingNO x emissions for the full operating range, it is ob-vious that the operating range with MIP operation isvery limited. Maximum achievable engine load is just

above 4 bar net Indicated Mean Effective Pressure(IMEP net ). This is anything but impressive, but withthe selected operating strategy this is the maximumengine load. Interesting to note is also that it is thelimitation of engine out NOx that sets the maximumload limit, not any other operational parameter. Theemission of noise (pressure derivative) is for examplefar below its limit, it is never more than half the limitvalue within the engine operating range. The samething can be said about the other operating parame-ters.

A strategy for retarding combustion phasing to limitNO x emissions could be applied but it comes to alarge cost in net indicated efciency (3-4%). Sincecombustion has to be phased very late to decreaseNO x below the 0.2g/kWh limit, (CA50 between 12and 15 CAD after TDC) the cycle to cycle variationsof engine load and combustion phasing will hence besignicantly larger compared to if CA50 is set for MIE.In some cases it might not even be practically feasibleto run the engine with such late combustion phasing.A better way seems to be to apply boost pressure.Even though some boost was present even with the

-359-



8/12/2019 2007-01-1855

5/9

Table 3: Color and sign encoding belonging to NOxand indicated efciency gures of the unboosted en-gine experiments (Figure 2 and Figure 3).

Load [bar]6

5 4 + + + + + +3 2

Speed [rpm] 800 1000 1200 1400 1600 1800

5 0 5 10 15 2020

25

30

35

40

45Net indicated efficiency

CA50 [CAD]

n e

t [ % ]

Figure 2: Net indicated efciency as a function ofcombustion phasing (MIP operation). Legend can befound in Table 3.

VGT set for MIP (intake pressure and gas exchangeefciency are shown in Figure 5 and Figure 6), this

level of boost pressure is negligible and does not inferany net indicated efciency decrease due to pumpinglosses in any of the operating points. It does, however,contribute slightly to the increased maximum IMEP at 1800 rpm . The option of further increasing the in-take pressure to a greater extent using the VGT isdealt with in the following section.

BOOSTED ENGINE OPERATION

For each combination of load, speed and combustionphasing the minimum boost pressure that resulted inacceptable NOx emission was applied. In practicethis meant that the boost pressure was gradually in-creased, maintaining constant combustion phasing,until the NOx criterion was met. The combustionphasing controllers adjusted the fuel composition au-tomatically to maintain the combustion phasing at thesetpoint. Finally intake temperature, being a slow ac-tuator, was used to bias the fuel composition de-cided by the combustion phasing controllers. Thecombustion phasing controllers have, as stated be-fore, the capability to affect CA50% by altering thefuel composition on a cycle to cycle basis. However

5 0 5 10 15 200

2

4

6

8

10

Specific NO x emissions

CA50 [CAD]

N O x [ g / k W h ]

Figure 3: NOx emissions as a function of combus-tion phasing (MIP operation). Legend can be found inTable 3.

Speed [rpm]

L o a d

[ b a r

]

NOx [g/kWh]

800 1000 1200 1400 1600 18002

3

4

5

6

7

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Figure 4: NOx emissions (MIP operation). Reddashed line indicate limit of 0.2 g/kW h .

Speed [rpm]

L o a d

[ b a r

]

P in [bar]

800 1000 1200 1400 1600 18002

3

4

5

6

7

1

1.02

1.04

1.06

1.08

1.1

1.12

1.14

1.16

Figure 5: Boost pressure during operation with mini-mum boost pressure.

-360-



8/12/2019 2007-01-1855

6/9

Speed [rpm]

L o a d [ b a r

]

ge [%]

800 1000 1200 1400 1600 18002

3

4

5

6

7

91

92

9394

95

96

97

98

99

Figure 6: Gas exchange efciency during operationwith minimum boost pressure.

to maintain the fuel composition criterion in all of thesix cylinders simultaneously intake temperature wasused to force each of the phasing controllers to usea fuel composition within the limits of the fuel crite-rion. Since the controllers adjust the fuel ratios tothe cylinders to maintain combustion phasing at thesetpoint, a change of intake temperature will resultin changed fuel ratios. In reality the intake temper-ature was increased until all of the cylinders fullledthe fuel composition criterion. In this way the mini-mum intake temperature needed was used (meaningmaximum thermal efciency). The process was iter-ated until suitable conditions were obtained (meaningthat the limitations from Table 2 are fullled). All of the

measurements were of course carried out when theengine had reached steady state conditions. In thisway the lowest possible boost pressure was used inorder to suppress the NOx formation. This strategyensures the minimum efciency penalty, in the formof pumping losses, for controlling the N Ox emissions.Combustion phasing was varied from the earliest pos-sible to the latest possible in the same manner as dur-ing MIP conditions. For each load and speed the bestefciency point (MIE) was again used for putting to-gether performance maps. MIE combustion phasingare not exactly same during boosted and MIP con-ditions. This is due to the fact that MIE now alsodepends on the gas exchange efciency and boostpressure. The suggested boost strategy represents aconstructive way of utilizing the extra degree of free-dom that the VGT offers.

The results from this operating principle are shown inFigure 7. The principal behavior is similar to MIP op-eration with efciency decreasing for early and latecombustion phasing except at very low loads. Athird effect of CA50 on the efciency is added us-ing the turbo to gain boost. With early combustion,

Table 4: Color and sign encoding belonging to N Oxand indicated efciency gures of the boosted engineexperiments (Figure 7 and Figure 8).

Load [bar]7 6

5 4 + + + + + +3 2

Speed [rpm] 800 1000 1200 1400 1600 1800

the exhaust temperature is reduced which increasesthe back pressure requirement for maintained boost.Thus early combustion phasing means higher pump-ing losses. The pumping losses with early combustionare further increased by the fact that early combus-tion generates higher maximum temperature and thusNO x which further increases the boost requirement.Because of this it is necessary to apply more boostpressure in order to lower the maximum temperature,hence net indicated efciency is decreased slightly forearly combustion phasings by the increased pumpinglosses. As a result MIE is phased later with boost thanwith MIP operation.

The reason for having the turbo charger is, as men-tioned previously, to extend the operating range ofthe engine with the limitations from Table 2 in mind.As visible from the 0.2 g/kWh NO x line in Fig-ure 9 the operating range was expanded by almost2 bar IM EP net with the use of the turbo charger. Thelimiting factor of engine load is as with MIP opera-

tion the engine out emissions of NOx . A compari-son of Figure 3 and Figure 8 reveals that the VGTturbo makes it possible to move the NOx forma-tion curves signicantly to the left. This means that,for all the loads that were possible to operate withoutboost, N Ox formation requires much earlier combus-tion phasing with boost than without. It is hence pos-sible to maintain MIE combustion phasing at higherloads without violating the NOx limit. Almost all ofthe indicated operating points in Figure 8 have beenrun with a signicant amount of boost to reduce N Oxemissions below the limit. The boost pressure usedand the corresponding gas exchange efciency areshown in Figure 10 and Figure 11. In the event thatit was not possible to reduce N Ox below the limit theturbo was set for maximum boost pressure (as is thecase with some high load and low speed curves).

NET INDICATED EFFICIENCY

The most important result of this investigation is thecomparison of the overall net shown in Figure 12 andFigure 13 for MIP and boosted operation respectively.It is found that net for the two cases are equal up to

-361-



8/12/2019 2007-01-1855

7/9

5 0 5 10 15 2020

25

30

35

40

45

50

55Net indicated efficiency

CA50 [CAD]

n e

t [ % ]

Figure 7: Net indicated efciency as a function ofcombustion phasing. Intake pressure according toboost strategy. Legends can be found in Table 4.

5 0 5 10 15 200

2

4

6

8

10

12Specific NO x emissions

CA50 [CAD]

N O

x [ g / k W h ]

Figure 8: N Ox emissions as a function of combustionphasing. Intake pressure according to boost strategy.Legends can be found in Table 4.

Speed [rpm]

L o a d [ b a r

]

NOx [g/kWh]

800 1000 1200 1400 1600 18002

3

4

5

6

7

0.5

1

1.5

2

2.5

Figure 9: NOx emissions. Engine run according tooperational strategy. Red dashed line indicates limitof 0.2 g/kW h .

Speed [rpm]

L o a d

[ b a r

]

Pin

[bar]

800 1000 1200 1400 1600 18002

3

4

5

6

7

1

1.1

1.2

1.3

1.4

1.5

Figure 10: Boost pressure with boosted operation ac-cording to strategy.

Speed [rpm]

L o a d

[ b a r

]

ge [%]

800 1000 1200 1400 1600 18002

3

4

5

6

7

82

84

86

88

90

92

94

96

98

Figure 11: Gas exchange efciency with boosted op-eration according to strategy.

-362-



8/12/2019 2007-01-1855

8/9

8/12/2019 2007-01-1855

9/9

REFERENCES

[1] S. Onishi, S.H. Jo, K. Shoda, P.D. Jo, S.Kato, Active Thermo-Atmosphere Combustion(ATAC)-A New Combustion Process for InternalCombustion Engines, SAE 790501

[2] P. Najt, D.E. Foster, Compression-Ignited Ho-mogeneous Charge Combustion, SAE 830264

[3] R.H. Thring, Homogeneous-ChargeCompression-Ignition (HCCI) Engine, SAE892068

[4] M. Stockinger, H. Sch apert ons, P. Kuhlmann,Versuche an einem gemischansaugenden mitSelbszundung, MTZ 53 (1992).

[5] G. Haraldsson, J. Hyv onen, P. Tunest al, B. Jo-hansson, HCCI Combustion Phasing in a MultiCylinder Engine Using Variable Compression

Ratio, SAE 2002-01-2858.[6] J. Martinez-Frias, S. Aceves, D. Flowers, J.R.

Smith, R. Dibble, HCCI Engine Control by Ther-mal Management, SAE 2000-01-2869

[7] N. Milovanovic, R. Chen, J. Turner, Inuence ofthe Variable Valve Timing Strategy on the Con-trol of a Homogeneous Charge Compression(HCCI) Engine SAE 2004-01-1899

[8] J-O. Olsson, P. Tunest al, B. Johansson Closed-Loop Control of an HCCI Engine, SAE 2001-01-1031

[9] R. Johansson, A. Rantzer (Eds.), Nonlinearand Hybrid Systems in Automotive Control,Springer-Verlag London Berlin Heidelberg, ISBN1-85233-652-8, 2003.

[10] J. Bengtsson, R. Johansson, P. Strandh, B. Jo-hansson and P. Tunest al, Closed-Loop Com-bustion Control of Homogeneous Charge Com-pression Ignition (HCCI) Engine Dynamics In-ternational journal of adaptive control and signalprocessing, 2004, Vol. 18, No. 2, pp 167-179,Wiley 2002; 00:1-6.

[11] G.M. Shaver, J.C. Gerdes, M. Roelle, Physics-Based Closed-Loop Control of Phasing, PeakPressure and Work Output in HCCI Engines Uti-lizing Variable Valve Actuation, Proceeding ofthe 2004 American Control Conference, June-July, 2004, Vol. 1, pp 150-155, IEEE 0-7803-8335-4/0.

[12] J. Bengtsson, P. Strandh, R. Johansson, P.Tunest al, B. Johansson, Hybrid Control ofHomogeneous Charge Compression Ignition

(HCCI) Engine Dynamics, International Journalof Control, Vol. 79, No. 5, 2006, pp 422-448, Tay-lor&Francis.

[13] P. Tunest al, Estimation of the In-CylinderAir/Fuel Ratio of an Internal Combustion Engineby the Use of Pressure Sensors, Doctoral The-sis, Lund University.

CONTACT

Carl Wilhelmssone-mail: [email protected]

Per Tunest ale-mail: [email protected]

Bengt Johanssone-mail: [email protected]

APPENDIX

DEFINITION OF CA50%

The denition of CA50% (the crank angle where 50%of the total heat release is obtained) is evident fromgure 14. CA50% is best understood as the instancewhere half of the total heat which will be released dur-ing the cycle are released. CA50% were in this workas well as in much other research, used as an mea-surement on where during the engine cycle combus-tion actually occur.

20 15 10 5 0 5 10 15 20500

0

500

1000

1500

2000

2500

C u m u l a t

i v e

H e a

t R e l e a s e

[ J ]

CA 10%

CA 50%

CA 90%

max Qch

min Qch

10%

50%

90%

Qtot

=max Qch

min Qch

HRD = CA 90% CA 10%

Figure 14: A typical accumulated heat release curve(adopted from [13]). The denition of CA50% is indi-cated in the gure, the unit of the horizontal axis areCrank Angle Degrees (CAD).

-364-



Documents

2007-01-1855