The Influence of Boost Pressure on Emissions and Fuel Consumption ...

SAE TECHNICAL PAPER SERIES

1999-01-0840

The Influence of Boost Pressure on Emissions and Fuel Consumption of a Heavy-Duty Single-Cylinder D.I. Diesel Engine K. V. Tanin, D. D. Wickman, D. T. Montgomery, S. Das and R. D. Reitz University of Wisconsin-Madison

Reprinted From: In-Cylinder Diesel Particulate and NOx Control 1999 (SP-1427)

International Congress and Exposition Detroit, Michigan March 1-4, 1999 400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A.

Tel: (724) 776-4841 Fax: (724) 776-5760

The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the paper may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay a $7.00 per article copy fee through the Copyright Clearance Center, Inc. Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. SAE routinely stocks printed papers for a period of three years following date of publication. Direct your orders to SAE Customer Sales and Satisfaction Department. Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department. To request permission to reprint a technical paper or permission to use copyrighted SAE publications in other works, contact the SAE Publications Group.

All SAE papers, standards, and selected books are abstracted and indexed in the Global Mobility Database

No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission of the publisher. ISSN 0148-7191 Copyright 1999 Society of Automotive Engineers, Inc. Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group. Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA

1999-01-0840

The Influence of Boost Pressure on Emissions and Fuel Consumption of a Heavy-Duty Single-Cylinder D.I. Diesel Engine K. V. Tanin, D. D. Wickman, D. T. Montgomery, S. Das and R. D. Reitz University of Wisconsin-Madison Copyright © 1999 Society of Automotive Engineers, Inc.

ABSTRACT

INTRODUCTION

An electronically controlled Caterpillar single-cylinder oil test engine (SCOTE) was used to study diesel combustion. The SCOTE retains the port, combustion chamber, and injection geometry of the production six cylinder, 373 kW (500 hp) 3406E heavy-duty truck engine. The engine was equipped with an electronic unit injector and an electronically controlled common rail injector that is capable of multiple injections.

The capability of direct-injection (DI) diesel engines to produce efficient power is well known. However, diesel engines exhibit a propensity for high nitrogen oxide (NOx) and particulate matter (particulate) emissions, and most strategies to reduce either NOx or particulate emissions cause an increase in the other emission. Since 1990, progress in engine technology has allowed a significant reduction in all forms of emissions from diesel engines. However, demands for further emissions reduction, quieter operation, and improved performance still provide serious challenges.

An emissions investigation was carried out using a sixmode cycle simulation of the EPA Federal Transient Test Procedure. The results show that the SCOTE meets current EPA mandated emissions levels, despite the higher internal friction imposed by the single-cylinder configuration. NOx versus particulate trade-off curves were generated over a range of injection timings for each mode and results of heat release calculations were examined, giving insight into combustion phenomena in current “state of the art” heavy-duty diesel engines.

Advanced fuel injection technology can help reduce NOx and particulate emissions. More precise control of the rate, pressure, and timing of fuel injection, while using a high-pressure injection system, not only permits simultaneous reductions in NOx and particulate formation during combustion, but can also reduce engine noise. The electronic unit injectors (EUIs) found on most commercially available heavy-duty engines are capable of precise highpressure injection [1].

Next, a study of the effects of varying boost pressure levels was conducted. For fixed brake specific NOx levels, with low-pressure (90 MPa) single injections, particulate was found to reduce monotonically as the boost pressure was increased. Interestingly, with low pressure double injections and with high pressure (>90 MPa) single injections, particulate was found to decrease at first and then to increase as the boost pressure was increased beyond an optimum level. A minimum value for particulate with respect to boost level was found for all cases, including the low-pressure single injections, when a correction for the six-cylinder turbocharger efficiency was applied. Computer modeling confirms that this is due to a reduction in the spray penetration and mixing that occurs as the engine gas density is increased. BSFC was generally reduced with increasing boost pressure. These results suggest that variable geometry turbochargers or other enhanced boosting methods will aid in the reduction of emissions and fuel consumption from heavy-duty truck engines.

Another technology used to gain more control over combustion processes in diesel engines is the use of a variable boost pressure system. A variable boost system allows flexible control, and thus optimization, of boost pressure for different load and speed conditions. In addition to the original power and efficiency goals of variable boost systems, these systems have proven to improve emissions and transient response as well [2]. Thus, variable boost systems are expected to be an important component for future low emissions heavy-duty diesel engines [1]. When a fixed geometry, un-wastegated turbocharger is used to provide boost for an engine, a compromise must be made. Midrange torque and efficiency must be sacrificed in order to not over-boost at high loads and speeds. Wastegates allow some exhaust gas to bypass the turbocharger turbine at high loads and speeds, thus reducing

1

experiments, however, duration was determined by required fuel flow rate for a given running condition. Table 2 gives the specifications of the EUI system. The EUI system has a rising injection rate profile with SOI (needle valve opening) at 36.5 MPa and the end at 29.0 MPa (needle valve closing).

boost pressure and avoiding an over-boost situation. The drawback to wastegates is that bypassed gas’ availability is wasted, leading to reduced overall efficiency at high loads and speeds [2]. In an attempt to provide increased boost at lower loads and speeds while not overboosting or operating at reduced efficiency at high loads and speeds, variable boosting systems such as VGTs and variable speed superchargers have been developed.

Specifications of the common rail fuel injection system are given in Table 3. The common rail system gives square injection rate profiles and is capable of single, double, triple, and quadruple injection schemes with injection pressures from 20 to 100 MPa, however, only single and double injection with a maximum of 90 MPa were used to ensure reliability. Timing and duration can be varied independently for each injection pulse.

There are three types of VGTs; those with variable area turbines, those with variable geometry nozzles, and those with axially movable vanes [2,3]. One type of VGT that appears especially promising is the axially movable vane turbocharger. They have a high degree of flexibility and fewer moving parts than other designs.

A PC based data acquisition and analysis system was used to take a 50 cycle average of cylinder pressure (and injection pressure for the EUI) at 1/2 crank angle degree increments. The cylinder pressure data, along with other operating conditions were then analyzed using the First Law of Thermodynamics to calculate the apparent heat release rate (AHRR).

Another variable boost system that promises to provide even more flexibility than VGTs is sequential hydrosupercharging (SHS). SHS systems use turbocharger compressor impellers driven by hydraulic turbines. An engine driven hydraulic pump would provide hydraulic power for the hydraulic turbines. The hydraulically driven compressors are used to increase the pressure of air supplied to the engine’s normal turbocharger compressor. This type of system would allow very flexible control of boost levels and retain the availability recovery of the turbocharger [4].

Baffle

Mixing Orifice

T.C.

Exhaust

Primary Dilution Air Compressed Air

Heater

Clearly, with the present variety of available variable boost systems, effects of boost pressure should be studied in more detail. Thus, a significant portion of the present research is dedicated to studying effects of boost pressure.

EGR Cooler

T.C.

EGR Filters Emissions Analyzers

Vacuum Pump

Exhaust

EXPERIMENTAL SETUP

Bellows Meter

T.C.

T.C.

The test engine is a fully instrumented Caterpillar singlecylinder oil test engine (SCOTE). The SCOTE is a singlecylinder version of the Caterpillar 3400 series heavy-duty diesel engine. It retains the injector, combustion chamber, and much of the port geometry of Caterpillar's 500 hp 3406E heavy-duty over-the-road truck engine. The engine laboratory setup used in the present experiments is shown schematically on Fig. 1 and the details of the engine are given in Table 1.

T.C.

Secondary Dilution Air

Intake

Caterpillar Single Cylinder Engine

Exhaust T.C.

Thermocouple Gate Valve Metering Orifice

Figure 1. Engine Laboratory Setup Table 1. Engine Specifications

The engine is capable of producing 62 kW at a rated speed of 1800 rev/min. Simulated supercharging and/or turbocharging with intercooling is accomplished by metering temperature controlled, compressed air into a intake surge tank and controlling back pressure in the exhaust surge tank. Engine fluid temperatures are monitored using type K thermocouples.

Engine Type

Caterpillar SCOTE (single-cylinder oil test engine) - single-cylinder - direct injection - 4 valve

Bore x Stroke

137.2 mm x 165.1 mm

Compression Ratio

15.6 : 1

Two fuel injection systems were used in the present experiments: A conventional electronic unit injector (EUI) system and a common rail system. The EUI system is the standard fuel injection system from Caterpillar for use with SCOTE and 500 hp 3406E heavy-duty diesel engines. With the EUI, start of injection (SOI) and duration can be varied independently. During the present

Displacement

2.44 liters

Combustion Chamber

Quiescent

Piston

Articulated Steel/Aluminum Mexican Hat Sharp Edge Crater

2

Emissions data recorded during experiments include total hydrocarbons (THC), CO, CO2, NOx, and particulate . SOF was measured from the particulate using Soxhlet extraction. Emissions data was obtained using the instrumentation summarized in Table 5. Emission levels were calculated according to US EPA CFR 40 specifications [6].

Table 2. EUI System Specifications Injector Type

Electronically Controlled Unit Injector (EUI)

Injection Pressure

Up to 190 MPa

Number of Nozzle Holes

6

Nozzle Hole Diameter

0.214 mm

Spray Angle (included)

130°

For measurements of particulate matter, a full dilution tunnel designed according to EPA 40CFR [6] recommendations is used. The purpose of the dilution tunnel is to simulate particulate growth due to hydrocarbon adsorption that is seen in the atmosphere as exhaust is expelled into air. The dilution tunnel is a mixing device that is designed to ensure that the exhaust is thoroughly mixed with dilution air. The dilution air is compressed air from the laboratory supply that is blown into the dilution tunnel through diffusers, which enhance mixing. A portion of diluted sample is drawn from the tunnel for analysis. The sample portion can be further diluted in the secondary dilution tunnel before being drawn, in series, through two Pallflex T60A20 filters. The filters are weighed before and after particulate deposition so that a total particulate measurement can be made.

Table 3. Common Rail System Specifications Injector Type

Electronically Controlled Common Rail Injector (see reference [5])

Injection Pressure

90 MPa

Number of Nozzle Holes

6

Nozzle Hole Diameter

260 µm

Spray Angle (included)

125°

COMPUTATIONAL SETUP

Table 4. Fuel Analysis Results Carbon

87.10%

Hydrogen

12.68%

C/H Ratio

6.869

Heat of Combustion Gross Net

45.19 MJ/kg 42.49 MJ/kg

Sulfur

0.048%

Cetane Rating

44.3

API Gravity @ 60° F

32.7

Viscosity @ 40° C

2.43 cSt

Flash Point, PMCC

67.2° C

THE MODEL – The basic features of the KIVA-II CFD code used in this study are well documented [7,8] and will not be repeated here for brevity. However, the original KIVA code has been extensively modified at the University of Wisconsin-Madison's Engine Research Center and a number of new and improved sub-models have been incorporated to improve its ability to predict diesel engine combustion with improved accuracy. These models include a modified RNG (Re-Normalization Group) ke turbulence model [9], a Kelvin-Helmholtz Raleigh- Taylor (KHRT) droplet breakup model [10], a wall-impingement model [11], and a dynamically varying drop-drag coefficient model to account for drop distortions from the spherical shape [12]. The ignition model is based on the Shell auto ignition model [13-15]. The earlier specification of a minimum cell temperature of 1000 K or product concentration of 0.1 was found to work well with no or low concentration of exhaust gas recirculation (EGR). With high EGR cases, however, the model was unable to predict the ignition delay correctly. Furthermore, if a high cell temperature is used as a switch from cool flame to main combustion, a detrimental effect was observed for retarded injection cases. This problem was rectified using a modified version, which incorporates the concentration of intermediate species, "q" as one of the criteria for switching from cool flame ignition to main combustion based on the suggestion of Schäpertöns and Lee [16].

Table 5. Emissions Instrumentation NOx

Chemiluminescent Analyzer / Thermo Environmental Instruments Inc., Model 10S

Particulate

Full Dilution Tunnel (EPA 40CFR Design)

THC

Flame Ionization Detector / Gow-Mac Model 23-500

CO

Infrared Gas Analyzer / California Analytical Instruments Model 3300A

CO2

Infrared Gas Analyzer / Horiba, Model VIA-510

The combustion model is based on the characteristictime model originally developed for spark ignition engines [17]. Finally, the emission models used are the Extended Zeldovich mechanism to predict NO formation [18] and the Hiroyasu model [19] for the soot formation. The soot oxidation models used in the present study are primarily

The fuel used during the present testing was obtained from a commercial fuel vendor, so it represents a valid sample of what is currently available to the trucking industry. A sample of the fuel was analyzed and the results are given in Table 4. 3

Tint is the intake surge tank temperature. Since this simplified code does not take into account heat transfer effects, some freedom was reserved to adjust the initial temperature and initial species densities within a reasonable range (< 5%) so that the computed initial pressure at IVC matched the experimental pressure.

based on the kinetically controlled Nagle and StricklandConstable (NSC) oxidation model [20]. However, the Arrhenius rate equation's pre-exponential factor was optimized to match experimental soot results. INITIAL AND BOUNDARY CONDITIONS – The prediction of diesel engine combustion using a CFD code strongly depends on the accurate specification of initial and boundary conditions in the combustion chamber. For computational efficiency, simulation starts from intake valve closing (IVC). A methodology developed at ERC for estimating species density and temperature at IVC as inputs for the combustion predictions was used in the present study [21]. In this prediction routine, it is assumed that complete combustion occurs at each cycle. The experimental data used are fuel flow rate, intake air flow rate, engine speed, volume at IVC, intake temperature, exhaust temperature, and the EGR percentage computed from the zero-dimensional residual gas fraction code. The initial gas temperature at IVC, Tivc, is determined by the following formula: 

At the solid wall surfaces, constant wall temperature with logarithmic law-of-the-wall velocity boundary conditions were used. Turbulence model parameters are the same as those reported by Han and Reitz [9]. In the computational study, the symmetry of the six nozzle holes symmetrically placed around the axis of the cylinder was exploited to minimize the computational time by considering only one-sixth of the engine combustion chamber in the computations. A perspective view of the computational grid with an arbitrary spray plume is shown on Fig. 2.



mr* Tr + ma, f +ma * EGR * Tint 100   Tivc = ma, f +ma * EGR + mr 100 Where ma,f, and ma are the mass of fresh air charge, and the total mass of intake charge respectively. Tr is the temperature of the residual gas (assumed to be the same temperature as the measured exhaust temperature), and

Figure 2. Computational Mesh

100

Mode 4

90 80

Mode 3

Load - %

70 60

Mode 5

50 40 30

Mode 2

20

Mode 6

10

Mode 1

0 600

800

1000

1200

1400

1600

Engine Speed (rev/min) Figure 3. Loads and Speeds for the Six-Mode FTP Simulation

4

1800

Table 6. Test Conditions (nominal) Running Condition

Mode 1

Mode 2

Mode 3

Mode 4

Mode 5

Mode 6

700

821

993

1672

1737

1789

0

25

75

95

57

20

Fuel Rate (kg/hr)

0.32

1.50

6.28

11.50

6.97

3.80

Intake Temp (C)

29

28

28

40

32

27

Intake Press (kPa), {psia}

100 {14.5}

103 {14.9}

161 {23.4}

267 {38.8}

184 {26.7}

121 {17.5}

Exhaust Press (kPa), {psia}

100 {14.5}

106 {15.4}

136 {19.8}

247 {35.9}

181 {26.2}

135 {19.6}

Speed (rev/min) % Load

between intake pressure and exhaust pressure that was used to set the exhaust pressure in the boost study, there is an enthalpy imbalance. The available energy in the exhaust at the exhaust pressures and temperatures used in the experiments was not great enough to accomplish the compression of the intake charge to the level used in the experiments if reasonable turbocharger efficiencies were assumed. Thus, a turbocharger model was developed to estimate the additional power required to compress the intake charge to the levels used in the experiments. The corrected six-cylinder data presented was attained by using a corrected engine power that is the difference between the measured power of the engine and the additional required power estimated using the turbocharger model. In the turbocharger model, the changes in enthalpy for an isentropic compression of the intake gases and an isentropic expansion of the exhaust gases are calculated. When these differences in enthalpies are calculated for the known six cylinder operating condition, an overall turbocharger (and manifolding) efficiency can be calculated as the ratio of the differences of compression and expansion enthalpies. The efficiency can then be used to calculate the additional power (above that which is available from the exhaust) which would be required to compress the intake charge. In the results presented on Figs. 18, 19, 21, 22, 27, 28, 30, 31, 33, 34, 36, and 37, both the actual experimental data and the “enthalpy balance corrected” data are presented (the corrected data is labeled “with 6-cyl. correction”). Table 7 shows the efficiencies used at each mode for the corrections (efficiencies were calculated from baseline runs on the single-cylinder engine using intake and exhaust pressures from six-cylinder engine data).

TEST CONDITIONS ENGINE TEST CONDITIONS – Testing was done at six operating conditions that constitute a six-mode FTP simulation. The running conditions are shown on a load versus speed map on Fig. 3. The controlled parameters for the running conditions are presented in Table 6. The test conditions held constant for all testing at a given mode were engine speed, intake temperature, intake pressure, and fuel flow rate.The running conditions, except for the fuel flow rate at mode 1, were taken from the operating map of a six cylinder production version of the Caterpillar 3400 series engine (373 kW (500 hp) 3406E). The fuel flow rate at mode 1 was established by varying injection duration, and thus the fuel flow rate, to give a power output near but slightly above zero. The assumption made was that the additional frictional mean effective pressure imparted due to the use of a single-cylinder test engine would be equivalent to the accessory load on an “as installed” six cylinder engine. A constant pressure drop across the engine was selected as the exhaust pressure determinant during the boost pressure experiments because it was thought that this would yield a nearly constant residual, so that intake boost pressure effects would not be confounded with “internal EGR” effects from the increased residual. This assumption was verified using a “1-D” unsteady flow model [22]. The model was run for four engine cycles to ensure that a steady-state condition had been achieved. When operating at mode 5, the residual gas fraction in the fourth cycle’s intake charge changes from 4.0% to 4.7% of the total intake charge when going from the base condition (+0% boost) to the +55% boost condition. If, for the +55% boost condition, the additional work required for an enthalpy balance (modeling a real turbocharger) were coming from piston compression work, the exhaust pressure and temperature would have increased to 369kPa and 673K respectively, (compared to 282kPa and 627K for the constant delta P used in the experiment). Under this elevated backpressure condition, the 1-D model predicts that the residual gas fraction would climb to 6.3%.

Table 7. Efficiencies Used for Six-Cylinder Corrections

The explanation for the six-cylinder correction seen in the results of the boost pressure experiments is as follows. With the assumption of a constant pressure drop 5

Running Condition

Overall Turbocharger/ Manifolding Efficiency

Common Rail, Mode 4

47.25%

Common Rail, Mode 5

47.83%

EUI, Mode 3

52.71%

EUI, Mode 4

49.58%

EUI, Mode 5

45.55%

COMPUTATIONAL SIMULATION – In the computational work, results are given on a fuel specific basis rather than on a brake specific basis, so the enthalpy imbalance does not affect the computational results. Two different load conditions (Modes 4 and 5) with varying boost pressures were studied in the present work. The intake boost pressures were varied, while keeping SOI at the value used in the corresponding engine experiment.

0.8

-9.5 ATDC

Particulate (g/hr)

0.7

NOMENCLATURE

0.6 0.5

-0.5 ATDC

0.4 0.3

-6.5 ATDC

-3.5 ATDC

Start of injection (SOI) timing data will be given in crankshaft degrees after top dead center (ATDC), thus -6 ATDC means 6 degrees before top dead center. Emissions data are given in either on an emission rate basis (g/hr), a brake power specific (brake specific) basis (g/ kW-hr or g/bhp-hr), or on a fuel specific basis (g/kg-fuel). “Soot” should be taken to mean the insoluble portion of particulate emissions.

0.2 11

12

13

14

15

16

17

18

19

20

NOx (g/hr)

Pressure (bar) and Rate of Injection (mg/CA)

RESULTS OF BASELINE ENGINE EXPERIMENTS One objective of the present study was to explore the emissions and performance characteristics of a current state-of-the-art heavy-duty diesel engine using the standard production EUI injection system. As mentioned previously, a multi-mode FTP simulation was used. The results for each mode will be discussed and then the results of the cycle estimate will be reviewed. MODE 1 – Mode 1 is a simulation of idle. The peak injection pressure at mode 1 is 40 MPa. Idle is an important running condition for heavy-duty engines because this type of engine often sees extended idling periods while in operation.

60

0.5

50

0.4

40

0.3

30

0.2

20

0.1

10

0.0

0 -40 -30 -20 -10

0

10

20

30

40

AHRR (normalized)

Figure 4. NOx versus Particulate Trade-off Curve for Mode 1

-0.1 50

Crank Shaft Angle (degrees ATDC)

Figure 5. Cylinder Pressure, Rate-of-injection, and Apparent Heat Release Rate for Mode 1, -6.5 ATDC SOI

Figure 4 shows the NOx versus particulate trade-off curve for mode 1. This trade-off was produced by varying the SOI from -9.5 degrees ATDC to -0.5 degrees ATDC. The units used on Fig. 4 are g/hr rather than the brake specific g/bhp-hr units, as will be used in the discussion of the other modes. Emission rate units are used at mode 1 because the power is quite low and emissions levels given with brake specific units are hard to interpret because they are so large, owing to the fact that they are a quotient with a very small denominator (power).

Figure 5 is a plot showing cylinder pressure, rate-of-injection, and apparent heat release rate (AHRR) for mode 1 when using an SOI of -6.5 ATDC. The -6.5 ATDC is the SOI that results in the lowest combined estimate of emissions and fuel consumption when used in the calculation of the six-mode-cycle estimate. On Fig. 5, notice that all of the fuel is injected during the ignition delay. Injecting fuel during the ignition delay causes a large premixed burn fraction, which is often associated with high NOx emissions. At mode 1, where a very small amount of fuel is injected, much fuel is premixed before start of ignition. The heat release curve on Fig. 5 shows no appreciable diffusion burn which confirms that fuel burns at a high kinetically controlled rate (premixed burn), rather than at a slower diffusion controlled rate (diffusion burn).

On Fig. 4, notice that when operating at the most advanced timing (-9.5 ATDC), there is sharp increase in particulate mass. This is evidence of over-penetration of the fuel spray. At earlier injection timings, the in-cylinder temperature and air density are relatively low. Therefor, much of the injected fuel collects on the walls and the piston and doesn’t burn completely. As a result, the particulate exhibits a high percentage of soluble fraction (~90%) at the idling conditions represented by mode 1.

6

MODE 2 – Mode 2 is a low speed (821 rev/min), low load (25%) operating condition. Peak injection pressure for mode 2 was 75 MPa. Figure 6 shows the NOx versus particulate trade-off for mode 2. This trade-off was produced by varying the SOI from -6.0 degrees ATDC to 2.0 degrees ATDC. On Fig. 6, at the most retarded timing there is a sharp drop in particulate. The unexpected sharp drop occurs because under these conditions, injection occurs when the piston is past TDC. Consequently, much heat transfer from the compressed air has already occurred, and the pressure and temperature are dropping rapidly due to expansion. The lower in-cylinder temperature causes a longer ignition delay, and consequently a large premix burn. The large premixed burn, in turn, causes high temperatures late in the cycle. The late cycle high temperature improves soot oxidation but also increases NOx levels. Even though retarded timings give interesting emissions results, BSFC is high, making this type of operation unacceptable.

0.080 -3.5 ATDC

0.05

0.067

0.04

-6.0 ATDC 0.03 5.5

0.054

6.5

7.0

7.5

8.0

8.5

-0.05 0 -30 -20 -10

0

10

20

30

40

50


Particulate (g/bhp-hr)

0.18

4.5 ATDC

0.16

0.24 0.21

-1.5 ATDC

0.19

0.14

0.16

0.12

Particulate (g/kW-hr)

0.27

-7.0 ATDC

NOx (g/bhp-hr)


0.05 0.04 0.03 0.02 0.01

AHRR (normalized)

100 90 80 70 60 50 40 30 20 10 0 -30 -20 -10

0.00 0

10

20

30

40

50


Figure 9. Cylinder Pressure, Rate-of-injection, and Apparent Heat Release Rate for Mode 3, -4.5 ATDC SOI

0.040 6.0

0.00

10

0.13 0.10 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00



0.06

0.05

-4.5 ATDC

-0.5 ATDC 0.094

20

1.5 ATDC

0.107

2.0 ATDC

0.10

0.20

9.39 10.06 10.73 11.40 12.07

0.07

0.15 30

5.36 5.70 6.04 6.37 6.71 7.04 7.38 7.71 8.05 0.30 0.22


8.72

0.20 40

NOx (g/kW-hr)

NOx (g/kW-hr) 8.05

0.25

50

Figure 7. Cylinder Pressure, Rate-of-injection, and Apparent Heat Release Rate for Mode 2, –0.5 ATDC SOI

Figure 7 shows pressure, apparent heat release rate, and instantaneous rate-of-injection for the –0.5 ATDC SOI point seen on Fig. 6. Similar to the -6.5 ATDC at mode 1, the –0.5 ATDC SOI point seen on Fig. 6 is the mode 2 emissions level that results in the lowest emissions when used in the calculation of the six-mode-cycle estimate. Figure 7 shows that at mode 2, similar to mode 1, almost all of the fuel is injected during the injection delay. Thus, high NOx emissions are expected at this mode due to the large premixed burn fraction.

7.38 0.08

0.30

60

AHRR (normalized)


Also, the study by Montgomery et al. [23] showed that injection pressure has a significant influence on total particulate emissions at engine idling. During those injection pressure variation experiments it was found that 30 MPa injection produced the lowest particulate. In the present study, the peak injection pressure for the mode 1 is 40 MPa. Which appears to be appropriate, except at more advanced timings.

9.0

NOx (g/bhp-hr)

Figure 6. NOx versus Particulate for Mode 2

7

MODE 3 – Mode 3 is an operating condition at 75% load with an engine speed of 993 rev/min. Peak injection pressure at mode 3 was 95 MPa. Mode 3 is a simulation of engine operation near the torque peak.

NOx (g/kW-hr)


Figure 8 shows the NOx versus particulate trade-off curve for mode 3. Again, notice that the trade-off is normal except for the “hook” at late timing that was also seen for mode 2 and discussed in that section. Figure 9 shows pressure, apparent heat release rate, and instantaneous rate-of-injection for the –4.5 ATDC SOI point seen on Fig. 8 (the point used in the six-mode cycle evaluation). This is the first mode where fuel is sprayed for a period of time longer than the ignition delay. In fact, the injection at this high load mode lasts well into the diffusion burn. Thus, the heat release displays the three normal characteristics of diesel combustion; ignition delay, premixed burn, and diffusion burn.

0.08

0.107

0.07

0.094

0.06

0.080

3.0 ATDC 0.05

0.067

0.04


3.8 4.0 4.3 4.6 4.8 5.1 5.4 5.6 5.9 6.2 0.10 0.134 -6.0 ATDC 0.09 0.121

0.054

0 ATDC

-3.0 ATDC

0.03 0.040 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6

NOx (g/bhp-hr)


MODE 4 – Mode 4 an operating condition at 95% load with an engine speed of 1672 rev/min. Peak injection pressure at mode 4 was 170 MPa. Mode 4 is a simulation of high load engine operation between the torque and power peaks. Mode 4 is an especially interesting mode for two reasons. First, at this mode the engine operates at high load and speed conditions, where normal emissions control schemes are less effective. Second, mode 4 is the largest contributor to the cycle estimate of NOx, power, and fuel consumption. Figure 10 shows a NOx versus particulate trade-off generated by varying SOI at mode 4. As can be seen from Fig. 10, most of the emission curve for mode 4 is within the region below the 1998 EPA emission levels of 5.36 g/ kW-hr (4.00 g/bhp-hr) NOx with 0.134 g/kW-hr (0.100 g/ bhp-hr) particulate.

140

0.08

120

0.07

AHRR (normalized)


0.06 100

0.05

80

0.04

60

0.03 0.02

40

0.01 20

0.00

0 -30 -20 -10

0

10

20

30

40

-0.01 50


Figure 11. Cylinder Pressure, Rate-of-injection, and Apparent Heat Release Rate for Mode 4, 3.0 ATDC SOI

Figure 11 shows pressure, rate-of-injection and apparent heat release rate versus crankshaft angle for a 3.0 ATDC SOI injection at mode 4. The 3.0 ATDC SOI point was the one used for mode 4 in the cycle emissions estimate. It is interesting to note that there is little premixed burn observed under these running conditions.

NOx (g/kW-hr) 2.95 3.22 3.49 3.75 4.02 4.29 4.56 4.83 5.10 5.36 0.087 0.065


This means that in-cylinder temperatures are sufficiently high that the fuel starts burning almost immediately after SOI. In support of this, the pressure trace is seen to be smooth relative to other modes, further indicating a small premixed burn fraction.

0.081

0.055

0.074

0.050

-4.0 ATDC 0.067

0.045

0.060

2.0 ATDC -1.0 ATDC

0.040 0.035 0.030 2.2

0.047

4.5 ATDC 2.4

2.6

2.8

0.054


7.5 ATDC

0.060

3.0

3.2

3.4

3.6

3.8

0.040 4.0

NOx (g/bhp-hr)


8

Figure 12 shows the NOx versus particulate trade-off curve for mode 5. Notice that the emissions levels at mode 5 comply easily with the 1998 emission standards. Figure 13 shows pressure, rate-of-injection and apparent heat release rate versus crankshaft angle for a 4.5 ATDC SOI injection at mode 5. The 4.5 ATDC SOI point was the one chosen to represent mode 5 in the cycle emissions estimate. Similar to mode 3, the heat release for mode 5 shown on Fig. 13 indicates distinct periods of ignition delay, premixed burn, and diffusion burn.

70

0.14

60

0.12 0.10

50

0.08

40

0.06 30

0.04

20

0.02

10

AHRR (normalized)


MODE 5 – Mode 5 is a running condition at 57% load and an engine speed of 1737 rev/min. The peak injection pressure for this mode was found to be 145 MPa.

0.00

0 -30 -20 -10

0

10

20

30

40

-0.02 50

Crank Shaft Angle (degrees)

Figure 15. Cylinder Pressure, Rate-of-Injection, and Apparent Heat Release Rate for Mode 6, -3.5 ATDC SOI

As discussed earlier, the magnitude of the unburned HC (soluble fraction) for the light-load modes depends largely upon the amount of fuel injected during the ignition delay. It also depends on the mixing rate with the air during this period, which is influenced by the swirl flow in the cylinder.

Similar to modes 2 and 3, at the most retarded injection timing (2 ATDC on the NOx versus particulate trade-off shown on Fig. 14) there is a decrease in particulate due to the longer ignition delay and the subsequent large premixed charge which causes high in-cylinder temperatures. This results in lower particulate and higher NOx than one might expect.

100 90 80 70 60 50 40 30 20 10 0 -30 -20 -10

0.09

SIX MODE CYCLE EMISSIONS ESTIMATE – In order to judge the effect of each mode on the overall cycle emissions, the emission rates and power from each mode is multiplied by a weighting factor that reflects the amount of time that the FTP test spends near the respective mode’s load and speed condition. For instance, the emissions at mode 1 (idle) are quite low, however mode 1 is weighted heavily in the FTP and is correspondingly weighted heavily in the six-mode simulation. In calculating the cycle estimate, each mode is represented by a point (SOI) with the most advantageous emission levels for that mode. For example, at mode 4, a -3.5 ATDC SOI yielded the lowest cycle estimate of emissions of all injection timings used at this mode.

0.08 0.07 0.06 0.05 0.04 0.03 0.02

AHRR (normalized)


MODE 6 – Mode 6 is a light load mode (20%) at high engine speed (1789 rev/min). The peak injection pressure for mode 6 was 93 MPa.

0.01 0.00 0

10

20

30

40

50


Figure 13. Cylinder Pressure, Rate-of-injection, and Apparent Heat Release Rate for Mode 5, 4.5 ATDC SOI

Table 8 summarizes the results of the six-mode analysis. As can be seen, the engine complies with 1998 emission levels and is thus a useful research tool for combustion and emissions research. The cycle NOx was 5.10 g/kWhr (3.80 g/bhp-hr), particulate was 0.107 g/kW-hr (0.079 g/bhp-hr), and the BSFC was 321 g/kW-hr (173 g/bhphr). Note that these values are not corrected for the single-cylinder engine’s frictional losses, which are higher than the six-cylinder version.

NOx (g/kW-hr) 5.36

6.03

6.71

7.38

8.05

-0.5 ATDC

0.322

0.24


8.72 0.349

0.22

0.295

-3.5 ATDC

0.20

-6.5 ATDC

0.268

0.18

0.241

0.16

0.215

2.0 ATDC

0.14 0.12 3.5

4.0

4.5

5.0

-9.0 ATDC 0.188 5.5

6.0


4.69 0.26

The bars on Fig. 16 represent each mode’s weighted contribution to the six-mode cycle’s estimate of emissions, fuel consumption, and power. Clearly, the modes where the emissions contributions are high and the power contribution is low are modes that have high brake specific emissions.

0.161 6.5

NOx (g/bhp-hr)

Figure 14. NOx versus Particulate Trade-Off Curve for Mode 6 9

45 40 35 30

% of total NOx

25

% of total Part

20

% of total Pow er

15

% of total Fuel

10 5 0 Mode 1

Mode 2

Mode 3

Mode 4

Mode 5

Mode 6

Figure 16. Modes’ Relative Contributions to the Cycle Estimate Table 8. Modes’ Emissions and Cycle Emissions Estimate Mode 1

Mode 2

Mode 3

Mode 4

Mode 5

Mode 6

SOI (deg. ATDC)

-6.5

-0.5

-4.5

3.0

4.5

-3.5

NOx (g/kW-hr) {g/bhp-hr}

43.61 {32.52}

7.98 {5.95}

6.52 {4.86}

4.03 {3.01}

3.62 {2.70}

5.90 {4.40}

Part. (g/kW-hr) {g/bhp-hr}

0.722 {0.538}

0.098 {0.073}

0.153 {0.114}

0.067 {0.050}

Power (kW) {hp}

0.4 {0.5}

7.5 {10.0}

32.8 {44.0}

53.33 {71.5}

30.0 {40.2}

15.1 {20.2}

Fuel (g/kW-hr) {g/bhp-hr}

888 {662}

208 {155}

194 {145}

218 {163}

231 {172}

252 {188}

Cycle NOx

5.10

Cycle Particulate

0.107

BSFC

231

(g/kW-hr)

3.80

(g/bhp-hr)

(g/kW-hr) 0.079

(g/bhp-hr)

(g/kW-hr)

(g/bhp-hr)

173

0.048 {0.036} 0.280 {0.209}

results on the other side of the inflection point from other boost levels’ results.

INTAKE AIR BOOST PRESSURE STUDY A second objective of the present study was to characterize effects of varying boost pressure level on diesel emissions and performance. Both fuel injection systems were used for this study; the EUI and the common rail system. Unlike the EUI system, which has a rising rate-of-injection, the common rail system produces a nearly square rate-of-injection shape. During the boost pressure experiments, the engine was run at modes 3, 4, and 5 with the EUI system and at modes 4 and 5 with the common rail system. Also, with the common rail system two sets of experiments were conducted at mode 5; one with single injections and the other with double injections. It is important to reiterate that during the experiments, SOI was varied in an attempt to keep brake specific NOx emissions constant. Therefore, changes in particulate emissions and BSFC versus boost pressure are isolated with respect to NOx emissions.

Table 9. NOx Levels Used During Boost Experiments Modes

Nox (g/bhp-hr)

Nox (g/kW-hr)

EUI. Mode 3 single injection

4.92

6.60


3.58

4.80


3.51

4.71

CR. Mode 4 single injection

2.78

3.73

CR. Mode 5 single injection

3.24

4.35

CR. Mode 5 split injection

3.30

4.42

The variation of NOx from run-to-run at any given mode was within 5%. At mode 5 the intake air pressure was increased up to +65% (304 kPa, 44.1 psia) in increments of 5 psia, where 184 kPa (184 kPa, 26.7 psia) was taken as +0% boost pressure. At mode 4 where +0% boost corresponds to 267 kPa (268 kPa, 38.8 psia), the intake air pressure was increased only up to +20% (321 kPa, 46.6 psia), to respect load limits on the engine. At mode 3 the boost pressure was increased up to +85% (299 kPa, 43.3 psia), with 161 kPa (23.4 psia) taken as the +0% boost.

Table 9 presents the mean NOx levels that were chosen and kept constant during the boost pressure experiments at each given mode. The NOx levels in Table 9 were chosen because they were on the smooth part of the NOx versus particulate trade-off curve, away from inflection points. Inflection points might make data interpretation difficult if the SOI required for a given boost level gave 10

the SOI needed to keep brake specific NOx constant during the experiments. Notice on Fig 17a that equivalence ratio decreases inversely and air/fuel ratio increases linearly with increasing boost pressure, just as one would expect for an increase in air flow with a constant fuel flow rate. The SOI points on Fig. 17 first proceed toward more retarded timings, then toward more advanced timings with increasing intake air pressure. This indicates that small increases in intake air pressure cause higher in-cylinder temperatures, requiring more retarded injection timings to hold a constant NOx level. However, as the intake air pressure is increased further, in-cylinder temperatures are decreased, allowing more advanced injection timings.

In order to correct the engine power for the increased intake pressure, the exhaust pressure was also increased so that difference between the intake and exhaust pressures was kept constant during all experiments, as described earlier. The results and discussion of the boost pressure experiments with the EUI system will be presented first followed by a discussion the common rail system performance. Intake Air Pressure (kPa) 172.4

206.8

241.3

275.8

310.3 -8.0

SOI Equiv.Rto.

0%

Equivalence Ratio

10%

20%

-4.0

0.60

35%

0.50

-2.0

45% 65%

55%

0.40

0.0

85% 20.0

Figure 18 shows the results of BSFC versus the intake air pressure for mode 3. The data on Fig. 18 includes both the constant delta P data from the experiments and data that has been corrected for the increased power required to compress the intake charge of a six-cylinder engine operating at elevated boost levels. Both sets of data show that the fuel consumption at mode 3 decreases drastically with increasing boost pressure until the boost pressure reaches +55% (250.1 kPa, 36.3 psi) at which point, increases in boost pressure were less beneficial.

-6.0

0.70

25.0

30.0

35.0

40.0

SOI (degrees ATDC)

137.9 0.80

2.0 45.0

Intake Air Pressure (psia) Intake Air Pressure (kPa) 137.9 150

0%

172.4

145

172.4

206.8

241.3

275.8

SOI A/F Ratio

0% Air/Fuel Mass Ratio

20

310.3 -8.0 -6.0

20%

10% 25

-4.0

45% 35%

30

-2.0

55%

65%

35 40 20.0

0.0

85% 25.0

30.0

35.0

40.0

SOI (degrees ATDC)

137.9 15

BSFC (g/bhp-hr)

Intake Air Pressure (kPa)

206.8

10% 20%

241.3

275.8

310.3 201.2

single cylinder data with 6-cyl. correction 194.4

140

187.7

135

35%

181.0

45%

130

174.3

55% 125 120 20.0

65% 25.0

30.0

35.0

40.0

BSFC (g/kW-hr)

Figure 17a. EUI, Mode 3, Equivalence Ratio and Injection Timing versus Boost

167.6

85% 160.9 45.0

Intake Air Pressure (psia)

Figure 18. EUI, Mode 3, BSFC versus Boost for constant NOx of 4.92 g/bhp-hr (6.60 g/kW-hr)

2.0 45.0


Figure 19 shows the effect of intake boost pressure on particulate emissions. Although the particulate emissions were still decreasing with increasing intake pressure, a tremendous reduction, more than 70%, was observed as the boost pressure was increased up to +45%, (233.9kPa, 33.9psia). Also, notice that at mode 3 the sixcylinder correction had little effect on the brake specific particulate emissions.

Figure 17b. EUI, Mode 3, Air/Fuel Mass Ratio and Injection Timing versus Boost BOOST PRESSURE EFFECTS WITH EUI – Mode 3 – At mode 3 experiments were conducted with intake air pressures from 161.3kPa (23.4 psia) for baseline (+0% boost), to 298.5kPa (43.3 psia) for +85% boost. Mode 3 results are shown on Figs. 17 to 19. Figure 17 shows the equivalence ratio (Fig. 17a) or air/fuel ratio (Fig. 17b) and

11

Mode 4 – As previously mentioned, Mode 4 is a high load (61kW) operating condition. At the baseline +0% boost pressure level, the engine’s power output just reached the maximum absorption capability of the dynamometer, so experiments at this mode were limited to decreases in boost pressure. The intake air pressure was decreased from 267.5kPa (38.8psia) at +0%, to 213.7kPa (31.0psia) at –20%. The SOI timings required to maintain constant brake specific NOx are shown on Fig. 20. Also shown on Fig. 20 is the equivalence ratio (Fig. 20a) and air/fuel ratio (Fig. 20b). Notice on Fig. 20 that unlike mode 3, at mode 4, injection timing must be steadily advanced with increasing boost pressure in order to maintain a constant brake specific NOx level.

Intake Air Pressure (kPa) 172.4

0.10

0%


0.09

206.8

241.3

275.8

310.3 0.148

single cylinder data with 6-cyl. correction

0.134 0.121

10%

0.08

0.107

0.07

0.094

20%

0.06

0.080

0.05

0.067

35%

0.04

0.054

45%

0.03

0.040

0.02 0.01 20.0

25.0

30.0

35.0

65%

85% 0.027

40.0

0.013 45.0


137.9 0.11


Intake Air Pressure (kPa) 206.8 156 154

220.6

234.4

248.2

262.0

275.8 -6.5

-20%

-6.0

Equivalence Ratio

0.60

0.56

-5.5

-15%

0.58

-5.0

SOI Equiv.Rto.

-10%

-4.5 -4.0

0.54

-3.5

0.52

-5%

0.50 0.48 30.0

BSFC (g/bhp-hr)

220.6

-3.0

0% 32.0

34.0

36.0

SOI (degrees ATDC)

206.8 0.62

262.0

275.8 209.2

203.8

-15%

150

201.1

-10%

148

198.5

146

195.8

-5%

144

140 30.0

193.1

0%

142 32.0

34.0

36.0

190.4

187.7 40.0

38.0


-2.5

Figure 21. EUI. Mode 4. BSFC versus Boost for constant NOx of 3.58 g/bhp-hr (4.80 g/kW-hr)

-2.0 40.0

38.0

248.2

single cylinder data 206.5 with 6-cyl. correction

-20%

152


234.4

BSFC (g/kW-hr)

Figure 19. EUI, Mode 3, Particulate versus Boost for constant NOx of 4.92 g/bhp-hr (6.60 g/kW-hr)

Intake Air Pressure (psia) Intake Air Pressure (kPa) 206.8


0.06


234.4

248.2

262.0

-20%

275.8 -6.5 -6.0

Air/Fuel Mass Ratio

24 25 26

-5.5

-15%

-5.0

SOI A/F Ratio

-10%

-4.5 -4.0

27

-3.5

-5%

28

-3.0 29 30 30.0

0% 32.0

34.0

36.0

38.0

SOI (degrees ATDC)

206.8 23

220.6

234.4

248.2

262.0

275.8 0.087


-20%

0.081

0.06 0.074

-15% 0.05

0.067

0.05

0.060

-10% -5%

0.04 0.04 30.0

0%

32.0

34.0

36.0

38.0

0.054



0.047 40.0


-2.5


-2.0 40.0


Figure 22 shows the relationship between the particulate and the boost pressure. The particulate emission rises rapidly (see Fig. 22) as the boost is decreased, providing a minimum of 0.052 g/kW-hr (0.039 g/bhp-hr) in particulate at –5% (254.4 kPa, 36.9 psia) boost pressure. Since experiments were not conducted at higher boost pres-

Figure 20b. EUI, Mode 4, Air/Fuel Mass Ratio and Injection Timing versus Boost

12

From the experimental results, it appears that the turbocharger selected for this specific engine has operating conditions, which match well with the engine at high load conditions. The intake air pressure at +0% boost provided by the turbocharger is close to the optimum pressure at mode 4. Mode 5 – As discussed earlier, mode 5 represents a high speed, intermediate load mode with a peak injection pressure of 145 MPa. Figures 23, 24, and 25 present cylinder pressure, rate-of-injection, and apparent heat release at mode 5 at the boost pressures of +0% (184 kPa, 26.7 psia), +35% (249 kPa, 36.1 psia), and +65% (304 kPa, 44.1 psia) respectively. In order to keep NOx constant, the injection timing had to be advanced somewhat as the boost pressure was increased (see Fig. 26). The requirement of earlier SOI is an interesting result because one might easily expect the opposite. However, it can be explained by the diluting effect of the intake mixture with extra air, which provides leaner air/fuel ratios and results in decreased in-cylinder temperatures and lower NOx levels. Also, as can be seen from the graphs, the premix burn decreases along with the ignition delay as the boost pressure increases. This also explains the influence of the boost on NOx formation since high NOx levels are generally thought to accompany large premixed burns. This behavior of NOx, premix burn, and the ignition delay versus the boost pressure was true for all cases, so graphs of the pressure, rate-of-injection and apparent heat release for the other modes will not be plotted.

0.08

100

0.07 80

0.06 0.05

60

0.04 0.03

40

0.02 0.01

20

AHRR (normalized)

Pressure (bar) and Rate of injection (mg/CA)

sures, it is difficult to judge whether this was an absolute minimum or not. However, the computational work, which was also conducted at this mode, gives an opportunity for further investigation as discussed below.

0.00 0 -30

-20 -10

0

10

20

30

40

-0.01 50


0.08

120

0.07

100

0.06 0.05

80

0.04

60

0.03 0.02

40

0.01

20

AHRR (normalized)


Figure 23. Cylinder Pressure, Rate-of-Injection, and Apparent Heat Release Rate for Mode 5, +0% Boost, 0.5 ATDC SOI

0.00

0 -30 -20 -10

0

10

20

30

40

-0.01 50



Figures 26 through 28 present equivalence ratio, air/fuel ratio, BSFC, and particulate versus the boost pressure, respectively, at mode 5 with the EUI fuel injection system. As can been seen from Fig. 28, particulate decreases at first and then increases as the boost pressure is increased up to +35%, 249 kPa (36.1 psia). This appears to be due to a reduction in the spray penetration and mixing that ultimately outweighs the benefits of leaner mixtures as the engine gas density is increased. The 35% increase in the intake air boost pressure results in almost 40% reduction in the particulate which corresponds to 0.058 g/kW-hr (0.043 g/bhp-hr). BSFC was reduced monotonically with increasing boost pressure for all cases. At 35% increased boost, the minimum for particulate, the BSFC was found be reduced by 7% for the corrected six-cylinder configuration.

160

0.08

140

0.07

120

0.06 0.05

100

0.04

80

0.03

60

0.02

40

0.01

20

0.00

0 -30 -20 -10

0

10

20

30

40

AHRR (normalized)

Figure 24. Cylinder Pressure, Rate-of-Injection, and Apparent Heat Release Rate for Mode 5, +35% Boost, -1.5 ATDC SOI

-0.01 50


Figure 25. Cylinder Pressure, Rate-of-Injection, and Apparent Heat Release Rate for Mode 5, +65% Boost, -4.0 ATDC SOI

13


207 224

241 259

276 293

310 -5.0

172 0.07

0%

30%

-2.0

40%

0.33

-1.0

50%

0.30

0.0

0.28


SOI Equiv.Rto.

SOI (degrees ATDC)

Equivalence Ratio

-3.0

20%

0.38

65%

259

276

293

310 -5.0

20%

45

SOI A/F Ratio

-3.0

30% -2.0

40% -1.0

50% 50

0.0

SOI (degrees ATDC)

Air/Fuel Mass Ratio

-4.0

10%

40

310 0.094

65% 45%

0.06

0.081 0.074

20%

10% 0.05

0.067

0.05

0.060

35%

0.054 30.0 32.5 35.0 37.5 40.0 42.5 45.0

BOOST PRESSURE EFFECTS WITH COMMON RAIL – Mode 4 with, single injection – At this mode the experiments were done for intake air pressures from 267.5kPa (38.8psia) at +0%, up to 321.0kPa (46.6psia) at +20%. Further increase in boost pressure was not considered pertinent since most existing turbochargers have a maximum pressure ratio of about 3:1.

0% 35

293


Intake Air Pressure (kPa) 241

276



207 224

259

0.06


190

241


0.04 25.0 27.5

0.25 1.0 25.0 27.5 30.0 32.5 35.0 37.5 40.0 42.5 45.0

172 30

224

0.07

-4.0

10%

0.35

207

0%

0.43 0.40

190


Intake Air Pressure (kPa) 172 190 0.45

The results of the experiments conducted at mode 4 with the common rail system, using single injections are shown on Figs. 29 to 31. Comparing the results with those of the EUI system at mode 4, it can be seen that trends of the curves are very similar, however the difference in the particulate is a factor of 3. The injection pressure for the EUI system was higher (170 MPa at mode 4) than for the common rail (90 Mpa), so the difference in the results confirms that there is an optimum injection and intake air pressure combination.

65% 55 1.0 25.0 27.5 30.0 32.5 35.0 37.5 40.0 42.5 45.0


Figure 26b. EUI, Mode 5, Air/Fuel Mass Ratio and Injection Timing versus Boost

Intake Air Pressure (kPa) Intake Air Pressure (kPa) 0.50

0%

221.3

163

217.9

160

214.6

158

10%

30%

40%

50%

20%

211.2

155

207.9

153

204.5

150 148

201.2


BSFC (g/kW-hr)

BSFC (g/bhp-hr)

165

65%

Equivalence Ratio

172 190 207 224 241 259 276 293 310 168 224.6

275.8

289.6

303.4

317.2

0%

-1.5

5%

0.48

-1.0

0.46

10%

0.44

SOI Equiv.Rto. -0.5 15%

0.42 0.40 38.0

197.8

330.9 -2.0

0.0

20% 0.5 40.0

42.0

44.0

46.0

SOI (degrees ATDC)

262.0 0.52

1.0 48.0


145 194.5 25.0 27.5 30.0 32.5 35.0 37.5 40.0 42.5 45.0

Figure 29a. Common Rail, Mode 4, Single Injection, Equivalence Ratio and Injection Timing versus Boost


Figure 27. EUI, Mode 5, BSFC versus Boost for constant NOx of 3.51 g/bhp-hr (4.71 g/kW-hr)

14


289.6

303.4

317.2

137.9 0.50

330.9 -2.0

172.4

241.3

-1.5

0.45

15%

0.0

34

20% 0.5

Equivalence Ratio

32

SOI (degrees ATDC)

-1.0

SOI A/F Ratio -0.5

10%

-6.0

0.40

10%

-5.0

20%

0.35

-4.0

0.30

40%

-3.0

0.25 36 38.0

40.0

42.0

44.0

1.0 48.0

46.0

65%

20.0

25.0

152

5% 15%

148 146 38.0

206.5

201.2

10%

198.5

20% 40.0

42.0

44.0

46.0

206.8

275.8

310.3 -7.0 -6.0

0%

40

10%

-5.0

20%

45

-4.0 50

40% -3.0

55

65%

60 20.0

195.8 48.0

241.3

SOI A/F Ratio

-10%

203.8

150

172.4

35

Air/Fuel Mass Ratio

154

137.9 30

330.9 209.2


0%

BSFC (g/bhp-hr)

317.2

BSFC (g/kW-hr)

303.4

-2.0 45.0

40.0



35.0

Figure 32a. Common Rail, Mode 5, Single Injection, Equivalence Ratio and Injection Timing versus Boost

Figure 29b. Common Rail, Mode 4, Single Injection, Air/ Fuel Mass Ratio and Injection Timing versus Boost

275.8

30.0



262.0 156

310.3 -7.0

0%

5%

30

275.8

SOI Equiv.Rto.

-10%

0% Air/Fuel Mass Ratio

206.8

25.0

30.0

35.0

40.0

SOI (degrees ATDC)

275.8

SOI (degrees ATDC)

Intake Air Pressure (kPa) 262.0 28

-2.0 45.0



Figure 30. Common Rail, Mode 4. Single Injection, BSFC vs. Boost, NOx at 2.78 g/bhp-hr (3.73 g/kW-hr)

Figure 32b. Common Rail, Mode 5, Single Injection, Air/ Fuel Mass Ratio and Injection Timing versus Boost


0.18

0.241

5%

0.16

0.215

0.14

0.188

10%

0.12

15%

0.10 0.08 38.0

40.0

42.0

44.0

46.0


330.9 0.295

0.20

0% Particulate (g/bhp-hr)

303.4


0.161

20%

0.134

137.9 166

172.4

164 162 158 154 152

146 20.0

Figure 31. Common Rail Mode 4, Single Injection Particulate versus Boost for constant NOx of 2.78 g/bhp-hr (3.73 g/kW-hr)

310.3 222.6 219.9 217.2 214.6

0%

211.9

156

148


241.3 275.8 single cylinder data with 6-cyl. correction

-10%

160

150

0.107 48.0

206.8

209.2

10%

206.5

20%

203.8 201.1

40%

198.5

65% 25.0 30.0 35.0 40.0 Intake Air Pressure (psia)

BSFC (g/kW-hr)

289.6

BSFC (g/bhp-hr)

275.8


262.0 0.22

195.8 45.0

Figure 33. Common Rail, Mode 5, Single Injection, BSFC versus Boost for constant NOx of 3.24 g/bhphr (4.35 g/kW-hr) 15

Mode 5, single injection – Figures 32 and 33 show the effect of boost air pressure on equivalence ratio, air fuel ratio and BSFC, respectively. Figure 34 shows the particulate effects. It can be seen that the tendency of the curve is almost the same as that of mode 4 with single injection (see Fig. 31), but the values decrease less rapidly with the boost pressure.


Air/Fuel Mass Ratio

40


275.8

310.3 0.603

0%

0.469

0.35 0.30

0.402

10%

0.25

0.335

20%

0.20

20.0

0.268

40%

0.15 25.0

30.0

35.0

310.3

344.7 -8.0

0%

-7.0 -6.0

SOI Equiv.Rto. -5.0

20% 30%

0.30 -4.0 0.25

45% 65%

0.20 25.0

30.0

35.0

40.0

45.0

75%

30%

-6.0

50

SOI A/F Ratio -5.0

55

45%

60

-4.0

65% 30.0

35.0

40.0

45.0

75% -3.0 -2.0 50.0

Intake Air Pressure (kPa) 172.4 180

206.8

-3.0

-2.0 50.0


Figure 35a. Common Rail, Mode 5, Split Injection, Equivalence Ratio and Injection Timing versus Boost

241.3

275.8

310.3

344.7 241.4


175

BSFC (g/bhp-hr)

10% 0.35

SOI (degrees ATDC)

Equivalence Ratio

0.40

275.8

20%

Mode 5, split injection – At mode 5, experiments were conducted with split injections using the same amount of fuel as for the single injection but 50% of the fuel was injected in the first pulse and 50% after a 10 degree crankshaft angle dwell. Figures 35, 36, and 37 show results for the common rail system at mode 5 with split injection. As can be seen on Fig. 37, the particulate exhibits a strong minimum at an optimal boost pressure of about +45%. In this case, comparable particulate emissions are reached with the low pressure injection system as those obtained with the higher injection pressure EUI system (e.g. compare to Fig 28 at +0% boost). Confirming the benefit of split injection seen previously (e.g. [23]).

0.201 45.0


45

-7.0

Figure 35b. Common Rail, Mode 5, Split Injection, Air/ Fuel Mass Ratio and Injection Timing versus Boost

Figure 34. Common Rail, Mode 5, Single Injection, Particulate versus Boost for constant NOx of 3.24 g/bhp-hr (4.35 g/kW-hr)

206.8

344.7 -8.0



172.4 0.45

310.3

0% 10%

70 25.0

65% 40.0

275.8

65


-10%

0.40

241.3

241.3

234.7

170

228.0

165

221.3

160

214.6

0% 155 150 145 25.0

10% 20% 30.0

45% 65% 30% 35.0

40.0

45.0

207.9

75%

BSFC (g/kW-hr)


0.45

172.4


137.9

206.8

SOI (degrees ATDC)

172.4 35

201.2 194.4 50.0


Figure 36. Common Rail, Mode 5, Split Injection, BSFC versus Boost for constant NOx of 3.30 g/bhphr (4.42 g/kW-hr)

16

Table 10. Simulation Parameters used in Mode 4



0.11

206.8

275.8

310.3


0%

0.10

241.3

344.7 0.161

Parameters 0.148

10%

0.134

75% 20%

0.09

0.121

0.08

65%

0.07 0.06 25.0

30% 30.0

35.0

Boost Pressure

0.107 0.094


172.4 0.12

45% 40.0

45.0

0.080 50.0

Baseline

-15%

+20%

+40%

Intake Temp, K

313

313

313

313

Intake Pressure, kPa

268

228

321

375

Exhaust Pressure, kPa

248

208

301

355

Air Flow Rate, kg/min

5.67

4.84

6.81

7.94

Fuel Flow Rate, kg/min

0.192

0.192

0.192

0.192


14

Figure 37. Common Rail, Mode 5, Split Injection, Particulate versus Boost for constant NOx of 3.30 g/bhp-hr (4.42 g/kW-hr)

Exp KIVA

Cylinder Pressure, MPa

12

COMBUSTION AND EMISSION MODELLING As described in the experimental study, results were obtained using six modes of operation. In this study, only the Mode 5 and Mode 4 cases were simulated with varying boost pressures. In the experimental study there was a limitation in the power output that could be safely absorbed by the existing engine dynamometer. Therefore, Mode 4 test runs could not be carried out beyond a certain boost pressure. Accordingly, the effect of increased boost pressure was further investigated in the simulation study. The airflow rate and SOI timing for constant NOx emissions corresponding to a given boost pressure level was determined using trends observed in the experiment and used as input in the simulation study.

10 8 6 4 2

0 -100

-50 0 50 Crank Angle, degree ATDC

100

(a) Baseline


12

SIMULATION OF MODE 4 – Mode 4 represents approximately 95 percent full load operating condition. A maximum boost pressure of 268 kPa could be used for the baseline experiment. Further experimental studies were carried out at reduced boost pressures of -5% to -20%. In the simulation four boost pressures were considered for Mode 4: the baseline (268 kPa), -10% boost (241 kPa), 15% boost (228 kPa) and -20% boost (214 kPa). The parameters used for the simulation are given in Table 10.

8 6 4 2

0 -100

Combustion and Emission Characteristics – Figures 38 and 39 show the comparison of cylinder pressure and rate-of-heat release, respectively, for the baseline and 15% boost conditions. In this case the cylinder pressure curves were found to match very well with the experimentally measured data. In the rate-of-heat-release curves shown on Fig. 39, the ignition delay is also well simulated.

Exp KIVA

10


100

(b) -15% Boost Pressure Figure 38. Comparison of Simulated and Measured Cylinder Pressure

17

Extrapolation Studies – For mode 4, the highest boost pressure cases were +20% and +40%. The intake pressure, air mass flow rate initial conditions were estimated by extrapolating the measured data in the experiments. Table 10 shows the extrapolated data used in the simulations for the +20% and +40% boost pressure cases. Also shown are the baseline and -15% boost pressure data obtained from the experiment.

Normalized Heat Release Rate

0.04 Exp KIVA

0.03 0.02 0.01

The predicted cylinder pressure and rate-of-heat-release curves are compared with the baseline condition on Fig. 41. The accumulated heat release data obtained from the simulations show that at a boost pressure of +20%, there is an increase in heat release by 2% at EVO but the accumulated heat release drops as the boost pressure is increased to 40%. This could again be due to the fact that at very high boost pressure some regions in the combustion chamber might have very low equivalence ratios and the low temperature inhibits combustion in those regions.

0.00

-0.01

-20

-10

0

10

20

30

40

50

60

Crank Angle, degree ATDC (a) Baseline

Exp KIVA

0.03

206.8 1.2

0.01 0.00

-0.01 -20 -10

Intake Air Pressure, kPa 241.3 275.8 310.3 344.7

0 10 20 30 40 50 Crank Angle, degree ATDC

379.2

Exp KIVA

1.0

0.02 Soot, g/kg-fuel


0.04

0.8 0.6 0.4 0.2

60

0.0 30

(b) -15% Boost Pressure Figure 39. Comparison of Predicted and Measured Heat Release

35

40 45 50 Intake Air Pressure, psia

55

Figure 40. Effect of Boost Pressure on Particulate Emission for Mode 4

Soot emission as a function of boost pressure is shown on Fig. 40. The measured data for particulate emissions includes both soot and soluble organic fraction (SOF) in the exhaust. In the simulation study only in-cylinder soot (the insoluble fraction of particulate) is predicted. Therefore, in the comparisons, the measured SOF was subtracted from the particulate emission data. The increasing trend of soot with lowering of boost pressure observed in the experiment is well maintained in the simulation study. However, the predicted soot levels are higher than those measured. The overall equivalence ratio is the main factor controlling soot emissions. In this operating condition, no minimum value of soot was observed, however, the slope of the curve was found to reduce with further increase in the boost pressure. This suggests that there is a cut-off boost pressure condition beyond which the advantage of increasing boost pressure on soot reduction no longer exists.

The emission results obtained with +20% and +40% boost pressure (Fig. 40) indicate that significant additional soot reduction could be obtained using a boost pressure in the range of +20 to +30%. Further increase in boost pressure may not be economical. With a modification in the turbocharger design, it may be possible to optimize an existing engine for lower soot emission while keeping the NOx emissions within acceptable limits. From the simulation study it is also evident that by increasing boost pressure the thermal loading of the engine remains unaffected as the fueling rate remains the same in this case. SIMULATION OF MODE 5 – Four different boost pressure conditions were considered: baseline, 20, 40 and 60 percent increase in boost pressure. The details of these conditions are listed in Table 11.

18

10

15

Baseline 20% Boost 40% Boost



20

10

5

0 -100


2


100

14

0.04 Baseline 20% Boost 40% Boost

0.03 0.02 0.01 0.00


Exp KIVA

12 Cylinder Pressure, MPa


4

(a) Baseline

(a) Average Cylinder Pressure

-0.01 -20 -10

6

0 -100

100

Exp KIVA

8

10 8 6 4 2

0 -100

60


100

(b) Normalized Heat Release

(b) +40% Boost Pressure

Figure 41. Effect of Boost Pressure on Cylinder Pressure and Heat Release for Mode 4

Figure 42. Comparison of Simulated and Measured Cylinder Pressure for Mode 5 Combustion and Emissions Characteristics – A comparison of the experimental and simulated average cylinder pressure is shown on Fig. 42 for the baseline and +40% boost pressure operating conditions. Figures 42a and 42b show reasonable agreement between the predicted and measured pressures, although the cylinder pressure is overestimated. The ability to predict correct ignition delay is one of the main features of the modified ignition model used in this study. The rates-of-heat-release for the above two cases are shown on Figs. 43a and 43b respectively. It is evident from Fig. 43 that the ignition delay is well predicted using the modified ignition model and the experimental and simulation results match with reasonable accuracy. The diffusion burn is also well reproduced in the simulation study.

Table 11. Simulation Parameters used for Mode 5 Boost Pressure Parameters

Baseline

+20%

+40%

+60%

Intake Temp, K

305

305

305

305

Intake Pressure, kPa

184

221

258

294

Exhaust Pressure, kPa

181

217

254

291

Air Flow Rate, kg/min

4.16

4.98

5.82

6.61

Fuel Flow Rate, kg/min

0.125

0.125

0.127

0.125

Equivalence Ratio

0.440

0.37

0.32

0.28

19

formation in the combustion chamber which is a function of in-cylinder flow pattern. With a high boost pressure, the spray penetration is reduced and the mixing phenomenon in the cylinder is modified. Therefore, the soot formation region in this case could be entirely different than for low boost pressure cases. In the next section, the mechanism of soot reduction with boost pressure is explained using the computational results.


0.06 Exp KIVA

0.05 0.04 0.03 0.02 0.01

Mechanism of Soot Reduction with Increased Boost Pressure – With increased boost pressure, the overall equivalence ratio in the combustion chamber decreases. This requires the use of more advanced SOI timings for the same level of NOx emissions. This is mainly due to the shortened ignition delay period and the lean burn conditions associated with increased boost pressures. In the case of Mode 5, an injection timing of -5°ATDC with +60% boost pressure increase gave the same level of NOx as 0° ATDC in baseline condition.

0.00

-0.01 -20

-10

0

10

20

30

40

50

60

Crank Angle, degree ATDC

(a) Baseline


0.06 0.05

Exp KIVA

0.04

Table 12. Fuel specific emission results for Mode 5 – experimental and computationally predicted

0.03

Experiment (g/kg-fuel)

0.02

Simulation (g/kg-fuel)

Boost Pressure

NOx

Soot

NOx

Soot

0.01

Baseline

21.49

0.310

20.41

0.194

0.00

+20%

22.82

0.177

20.85

0.103

+40%

23.24

0.210

23.99

0.062

+60%

24.04

0.247

25.74

0.057

-0.01 -20 -10


60

(b) +40% Boost Pressure 0.35 180

Figure 43. Comparison of Predicted and Measured Heat Release for Mode 5

0.30 Soot, g/kg-fuel

The effect of boost pressure on soot emissions is compared on Fig. 44. The decreasing trend of soot with increase in the boost pressure is also seen in the simulation study and could be the result of the lean-burn condition in the combustion chamber. However, the minimum soot value observed experimentally at about +30% boost pressure was not seen in the simulation. Instead, the predicted soot was found to be nearly constant at high boost pressure. The deviation from the experiment could be due to an inaccurate specification of the timing of SOI. This is supported by the increased NOx concentration observed in the simulation results at +60% boost pressure (Table 12). Nevertheless, the simulation study provides the same conclusion that increased boost pressure is effective in reducing soot up to certain percent only, beyond which there is no significant further soot reduction.

0.25

Intake Air Pressure, kPa 200 220 240 260 280

300

Exp KIVA

0.20 0.15 0.10 0.05 0.00 25

30 35 40 Intake Air Pressure, psia

45

Figure 44. Effect of Boost Pressure on Particulate Emission Due to the increased density in the combustion chamber, comparatively low spray penetration was observed with increased boost pressure. Midway through the injection duration period, a liquid spray penetration length of 1.32 cm was observed for +60% boost compared to 1.74 cm for the baseline condition, as shown on Fig. 45. The vapor penetration is also reduced and the low spray pen-

At high boost pressure the combination of lean mixture together with the overall low temperature in the combustion chamber could result in an increase in the soot emissions. Other possibility could be the location of soot 20

etration inhibits air entrainment and mixing. Therefore, this tends to increases soot formation in the case of high boost pressure. However, net soot is the result of a competition between the soot formation and oxidation processes in the cylinder. The abundance of oxygen available in the case of high boost pressures also enhances the soot oxidation process and finally results in reduction of net soot in the cylinder.

Average Cylinder Temperature, K

1500 Baseline 20% Boost 40% Boost 60% Boost

1400 1300 1200 1100

Comparison of the locations of soot formation regions in the combustion chamber was found not to show any considerable difference between the +40% and +60% boost pressure cases. Therefore, the location of soot formation is not a controlling parameter for increased soot in the case of +60% boost pressure.

1000 900 800

Figure 46 shows the average cylinder temperature and instantaneous fuel mass in the bowl for different boost pressure cases. The hatched area represents the crank angle range corresponding to the peak soot formation zone. From this figure it is observed that the peak soot formation corresponds to a constant amount of fuel in the bowl. In other words, the peak soot formation takes place when a certain constant amount of fuel is burned. The decreasing slope in the soot-boost pressure curve at very high boost pressures can be attributed to the combined effect of low temperatures and low equivalence ratios that inhibits soot oxidation.

0


(a) Average Cylinder Temperature

Fuel Mass in the Bowl, g

0.030

Baseline 20% Boost 40% Boost 60% Boost

0.025 0.020 0.015 0.010 0.005 0.000

-0.005 -10

0

10

20

30

40

Crank Angle, degree ATDC (b) Fuel mass in the piston bowl region Figure 46. Effect of boost pressure on average cylinder temperature and instantaneous fuel mass in the bowl.

(a) Baseline at 11.5° ATDC (l= 0.314e-1 h=0.283)

SUMMARY AND CONCLUSIONS A study of boost pressure effects was conducted using a single-cylinder version of a current heavy-duty diesel engine (Caterpillar 3406). A six-mode steady-state simulation of the federal transient test procedure (FTP) was used for the baseline evaluation. The fuel system used in the baseline study was an electronically controlled unit injector (EUI) capable of injection pressure up to 190 MPa. For the boost pressure study, the EUI and a common rail system capable of multiple injections were used.

(b) +60% boost at 6.5° ATDC (l= 0.371e-1 h=0.334) Figure 45. Liquid and vapor spray penetration midway through the injection period

BASELINE EMISSIONS AND PERFORMANCE STUDY – The results lead to the following conclusions: NOx Emissions – The low cycle NOx emissions using the EUI can be attributed to the rising shape of the rate-ofinjection. Especially at mode 4, the shaped injection introduced so little fuel during the ignition delay that the premixed burn was essentially eliminated, thus reducing NOx emissions significantly. 21

Particulate Emissions – The EUI injection also provides benefits in particulate emissions over other systems previously investigated [23]. It is assumed that this is due to the EUI’s rising injection rate. At low loads, when gas densities in the cylinder are low, the injection pressure is low and over-penetration is eliminated. However, at high loads, when gas densities are higher, the injection is longer and high injection pressures are attained, thus providing greater penetration and enhanced air utilization, leading to low particulate emissions.

4. Kapich, D.D., “Sequential Hydro-Supercharging System for Turbodiesels” SAE Paper 961744, 1996. 5. Miyaki, M., Fujisawa, H., Masuda, A., and Yamoamoto, Y.,“Development of New Electronically Controlled Fuel Injection System ECD-U2 For Diesel Engines”, SAE Paper 910252, 1991. 6. “Emissions Regulations for New Otto-Cycle and Diesel Heavy-Duty Engines; Gaseous and Particulate Exhaust Test Procedures”, CFR 40 Subpart N, US EPA, June 1997. 7. Amsden, A.A., Butler, T.D., O’Rourke, P.J., and Ramshaw. J.D., “KIVA: A comprehensive Model for 2D and 3D Engine Simulations,” SAE Paper 850554, 1984. 8. Amsden, A.A., O’Rourke, P.J. and Butler, T.D., “KIVA-II: A Computer Program for Chemically Reactive Flows with Sprays,” Los Alamos National Laboratory Report No. LA 11560-MS, 1989. 9. Han, Z., and Reitz, R.D., “Turbulence Modeling of Internal Combustion Engines Using RNG k- Models,” Combustion Science and Technology, 106, pp. 4-6, 267, 1995. 10. Reitz, R.D., “Modeling Atomization Processes in HighPressure Vaporizing Sprays,” Atomization and Spray Technology, 3, 309, 1987. 11. Gonzalez, M.A., Lian, Z.W., and Reitz, R.D., “Modeling Diesel Engine Spray Vaporization and Combustion,” SAE Paper 920579, 1992. 12. Liu, A.B., Mather, D., and Reitz, R.D., “Modeling the Effects of Drop Drag and Breakup on Fuel Sprays,” SAE Paper 930072, 1993. 13. Halstead, M., Kirsh, L., and Quinn, C., “The Autoignition of Hydrocarbon Fuels at High Temperatures and Pressures Fitting of a Mathematical Model,” Comb. Flame, 30, pp. 4560, 1977. 14. Theobald, M.A., and Cheng, W.K., “A Numerical Study of Diesel Ignition,” ASME Paper 87-FE-2, 1987. 15. Kong, S.C., and Reitz, R.D., “Multidimensional Modeling of Diesel Ignition and Combustion using a Multistep Kinetics Model,” Paper 93-ICE-22, ASME Trans., Journal of Engineering for Gas Turbines, and Power, 115, No. 4, pp. 781789, 1993. 16. Schapertons, H. and Lee, W., Multidimensional Modelling of Knocking combustion in SI Engines, SAE Paper 850502, 1985. 17. Abraham, J., Bracco, F.V., and Reitz, R.D., “Comparisons of Computed and Measured Premixed Charge Engine Combustion,” Comb. Flame, 60, pp. 309-322, 1985. 18. Bowman, C.T., “Kinetics of Pollutant Formation and Destruction in Combustion,” Prog. Energy Comb. Sci., 1, pp. 33-45, 1975. 19. Hiroyasu, H., and Kadota, T., “Models for Combustion and Formation of Nitric Oxide and Soot in DI Diesel Engines,” SAE Paper 760129, 1976. 20. Nagle, J., and Strickland-Constable, R.F., “Oxidation of Carbon between 1000-2000 C,” Proc. of the Fifth Carbon Conf., 1, pp. 154, 1962. 21. Senecal, P.K., Xin, J. and Reitz, R.D., “Predictions of Residual Gas Fraction in IC Engines," SAE Paper 962052, 1996. 22. Zhu, Y. and Reitz, R.D., “A 1-D gas dynamics code for subsonic and supersonic flows applied to predict EGR levels in a heavy-duty diesel engine”, To appear in International Journal of Vehicle Design, 1999. 23. Montgomery, D.T. and Reitz, R.D., “Six-Mode Cycle Evaluation of the Effect of EGR and Multiple Injections on Particulate and NOx Emissions from a D.I. Diesel Engine”, SAE Paper 960316, 1996.

BOOST PRESSURE STUDY – Optimal boost pressure at each mode was found to lead to significant improvements in BSFC and particulate emissions. Particulate emissions decreased significantly with increased intake boost pressure due to the increased available air for soot oxidation at elevated intake pressures (while holding brake specific NOx constant). However, the study also showed that there exist engine operating conditions at high boost (low equivalence ratio) beyond which the benefit of boost pressure on particulate emissions is not found. This is because at very high boost pressures, the liquid spray and vapor penetration lengths are reduced. This explanation was confirmed by using computer modeling. Similar trends were found when the present singlecylinder test results were corrected for turbocharger efficiencies of the six-cylinder engine configuration. When the observed particulate and BSFC reductions seen at modes 3, 4, and 5 were incorporated into the sixmode cycle, particulates are reduced from 0.107 g/kW-hr (0.079 g/bhp-hr) to 0.089 g/kW-hr (0.066 g/bhp-hr) and BSFC is reduced from 231 g/kW-hr (173 g/bhp-hr) to 212 g/kW-hr (158 g/bhp-hr). Clearly, these results indicate that enhanced boosting strategies will be important in meeting future heavy-duty diesel engine emissions and fuel consumption goals.

ACKNOWLEDGMENTS Funding from Caterpillar, TACOM/BKM and DOE/Sandia supported this work. The Army Research Office provided additional facilities and support. The authors also thank Matt Thiel and Paul Pachenski for their diligent assistance during the data collection and all the other individuals at the UW-Engine Research Center and Caterpillar who were involved in making this work possible.

REFERENCES 1. Browning, L.H., “Technologies and Costs for On-Road Heavy-Duty Engines Meeting 2004 Emissions Standards”, SAE Paper 973256, 1997. 2. Capobianco, M. and Gambarotta, A., “Variable Geometry and Wastegated Automotive Turbochargers: Measurements and Comparison of Turbine Performance”, Transactions of the ASME, Journal of Engineering for Gas Turbines and Power, Vol. 114, July 1992. 3. Shao, J., Ogura, M., Liu, Y., Yoshino, M., “Performance of a New Axially Movable Vane Turbocharger”, SAE Paper 961747, 1996.

22