EPA-460/3-76-022
July 1976
NITROGEN OXIDE CONTROL
WITH DELAYED-MIXING,
STRATIFIED-CHARGE
ENGINE CONCEPT
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
Ann Arbor, Michigan 48105
-------
This report is issued by the Environmental Protection
Agency to report technical data of interest to a
limited number of readers. Copies are available free
of charge to Federal employees, current contractors
and grantees, and nonprofit organizations - in limited
quantities - from the Library Services Office (MD-35),
Research Triangle Park, North Carolina 27711; or, for a
fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protec-
tion Agency by the Department of Mechanical Engineering,
University of Wisconsin, Madison, Wisconsin 53706, in
fulfillment of Grant No. R803858-01-0. The contents
of this report are reproduced herein as received from
the University of Wisconsin. The opinions, findings,
and conclusions expressed are those of the author and
not necessarily those of the Environmental Protection
Agency. Mention of company or product names is not to
be considered as an endorsement by the Environmental
Protection Agency.
Publication Ho. EPA-460/3-76-022
ii
-------
EPA-460/3-76-022
NITROGEN OXIDE CONTROL
WITH DELAYED-MIXING,
STRATIFIED-CHARGE
ENGINE CONCEPT
by
L.W. Evt-rs. P.S. Myers, anil O.A. I >ehara
Department of Mechanical Engineering
I'niversity of Wisconsin
Madison, Wisconsin 53706
Grant No. R803858-01-0
EPA Project Officer: John J. McFadden
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
Ann Arbor, Michigan 48105
July 1976
-------
ABSTRACT
The purpose of this study is to explore methods of
controlling the nitrogen oxide emissions from internal
combustion engines. From computer calculations, the
delayed mixing stratified charge engine concept was
selected. In the delayed mixing stratified charge en-
gine concept, combustion is initiated and completed in
a fuel-rich region, then air is mixed into those rich
products. A study of existing engines shows that some
operational stratified charge engines limit nitrogen
oxide emissions in a manner similar to the delayed
mixing concept. A single cylinder engine was modified
to include an air injection valve. When air was injected
after rich combustion, the nitrogen oxide emissions were
lower, the hydrocarbon emissions were lower, the carbon
monoxide emissions were about the same and the
efficiencies were lower than for homogeneous operation
at the same overall fuel-air ratio.
11
-------
PREFACE
The material presented in this report is essentially
the Ph.D. Dissertation of L. W. Evers. Much of the
support for the study was provided by the U. S.
Environmental Protection Agency. The theoretical study
given in Chapter II was presented as a Society of Auto-
motive Engineers (SAE) paper (741172) at the International
Stratified Charge Engine Conference in Troy, Michigan
(1974). Other results are to be presented in a SAE paper
in February 1977.
The thesis was written in the conventional British
units. Appendix I has been added to the report which
includes the conversion to SI Units.
111
-------
TABLE OF CONTENTS
Page
PREFACE iii
TABLE OF CONTENTS iv
LIST OF FIGURES ix
LIST OF TABLES xiv
NOMENCLATURE xv
CHAPTER I INTRODUCTION 1
SUMMARY 4
CONCLUSION 11
CHAPTER II A THEORETICAL STUDY OF NITRIC OXIDE
EMISSIONS FROM INTERNAL COMBUSTION
ENGINES 13
A. INTRODUCTION 13
B. NITRIC OXIDE KINETICS 14
C. COMPUTER MODELS 17
1. One System Model 18
2. Two System Model 20
D. POTENTIAL ENGINE CONFIGURATIONS 22
1. Homogeneous Charge-Conventional SI Engine 22
2. Stratified Charge Engine 32
E. CONCLUSIONS 51
CHAPTER III A DISCUSSION OF NITROGEN OXIDE EMISSIONS
FROM VARIOUS TYPES OF INTERNAL COMBUS-
TION ENGINES 53
A. INTRODUCTION 53
B. COMBUSTION CONCEPTS OF STRATIFIED CHARGE
ENGINES 55
IV
-------
Page
C. DEFINITIONS OF SOME TYPES OF INTERNAL COMBUS-
TION ENGINES 58
D. HOMOGENEOUS CHARGE ENGINES 61
E. THREE-VALVE STRATIFIED CHARGE ENGINES 63
F. RAPID MIXING STRATIFIED CHARGE ENGINE 68
1. Instantaneous Mixing Stratified Charge
Engine Concept 68
2. Texaco Contrc-lled-Combustion System (TCCS) 69
G. LIMITED MIXING STRATIFIED CHARGE ENGINES 70
1. Delayed Mixing Stratified Charge Engine
Concept 70
2. Ford Programmed Combustion (PROCO) Engine 70
3. Newhall Engine 76
H. COMPRESSION IGNITION STRATIFIED CHARGE ENGINES 78
1. Open Chamber Diesel Engines 83
2. Divided Chamber Diesel Engines 85
CHAPTER IV THE TEST FACILITY AND INSTRUMENTATION 89
A. INTRODUCTION 89
B. ENGINE 90
C. DYNAMOMETER 93
D. FUEL SYSTEM 94
E. CARBURETION AIR SYSTEM 96
F. INJECTION AIR SYSTEM 97
1. High Pressure Air Compressor System 97
-------
Page
2. High Pressure Air Flow Measurement System 98
3. Air Injection Valve "
G. PRESSURE TRANSDUCER SYSTEM 99
H. OSCILLOSCOPE 10°
I. MULTIPOINT TEMPERATURE RECORDER 100
J. EMISSIONS CART 101
1. Nitric Oxide Measurement 101
2. Carbon Monoxide Measurement 1°2
3. Hydrocarbon Measurement 1°3
CHAPTER V THE EXPERIMENTAL PROGRAM AND TEST RESULTS 105
A. INTRODUCTION 105
B. ENGINE INPUT VARIABLES 107
1. Operational Variables 107
2. Design Variables 109
3. Equivalence Ratios as Alternate Operational
Variables 109
b. Specified Operational Variables 110
C. A DISCUSSION OF THE THEORETICAL TRENDS EXPECTED
IN THE EXPERIMENTAL RESULTS 112
1, Nitrogen Oxide Emissions 112
2. Carbon Monoxide Emissions 113
3. Hydrocarbon Emissions 115
4. Engine Power 117
5. Efficiency 119
vi
-------
Page
D. IARGE DIAMETER NOZZLE TEST RESULTS AND DISCUSSION 121
1. Check-Out 121
2. Test Conditions for Large Diameter Nozzle 122
3. Summary of Results for Large Diameter Nozzle 123
4. Nitrogen Oxide Emissions for Large Diameter
Nozzle 124
5. Carbon Monoxide Emissions for Large Diameter
Nozzle 127
6. Hydrocarbon Emissions for Large Dia. Nozzle 127
?. Engine Efficiency for Large Diameter
Nozzle 130
8. Discussion of Results for Large Diameter
Nozzle 133
E. SMALL DIAMETER NOZZLE TEST RESULTS AND DISCUSSION 133
1. Test Conditions for Small Diameter Nozzle 133
2. Summary of Results for Small Diameter Nozzle 134
3. Nitrogen Oxide Emissions for Small Diameter
Nozzle 136
b. Carbon Monoxide Emissions for Small Diameter
Nozzle 144
5. Hydrocarbon Emissions for Small Diameter
Nozzle 146
6. Engine Efficiency for Small Diameter Nozzle 152
VI1
-------
Page
7. Engine Power for Small Diameter Nozzle 159
8. Start of Air Injection at 128° BTDC 161
9. The Effect of Air Injection on Combustion 164
10. Discussion of Results for Small Diameter
Nozzle 167
REFERENCES 168
APPENDICES 1?5
A. METHOD USED TO CALCULATE NITRIC OXIDE 175
B. ONE SYSTEM COMPUTER MODEL 177
C. TWO SYSTEM COMPUTER MODEL 197
D. DEFINITIONS OF THE PERFORMANCE PARAMETERS
USED IN CHAPTER II 221
E. AIR INJECTION FLOW RATE MEASUREMENT 222
F. MECHANICAL DESIGN OF AIR INJECTION VALVE 227
G. PERFORMANCE OF AIR INJECTION VALVE 23!
H. DATA REDUCTION COMPUTER PROGRAM 240
I. CONVERSIONS TO SI UNITS 241
VI11
-------
LIST OF FIGURES
Page
Figure 1 Two System Model of the Stratified Charge
Engine Conf igurations 21
Figure 2 The Combustion Function for the One-
System and Two-System Models 24
Figure 3 The Expansion Function for the One-System
Model (Weibe) and the Two-System Model
(Crank and Piston) 25
Figure 4 Indicated' specific nitric oxide (ISNO)
versus indicated mean effective pressure
(IMEP) for homogeneous charge engine con-
figuration 29
Figure 5 Indicated enthalpy efficiency (IEE)
versus indicated mean effective pressure
(IMEP) for homogeneous charge engine
configuration 30
Figure 6 The Influence of the Times of Heat Release
for a Homogeneous Charge Engine Operating
with F = 1.0 and EGR = .2 and the Delayed
and Instantaneous Mixing Stratified Charge
Engine Configuration Operating with
Fexhaust = 1-0 and EGR = .2 34
Figure 7 A Comparison of the Heat Release for In-
stantaneous Mixing, Delayed Mixing and
Homogeneous Combustion Corresponding to
Figure 6. 35
Figure 8 The Mass Fraction Burned and the Product
Mixture Equivalence Ratio (for Two of the
Nine Conditions) Corresponding to Figure
9 and Figure 10 37
Figure 9 Indicated Specific Nitric Oxide for In-
stantaneous Mixing Stratified Charge En-
gine Configuration 38
Figure 10 Indicated Specific Nitric Oxide for De-
layed Mixing Stratified Charge Engine
Configuration 3?
IX
-------
Page
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Figure 16
Figure 1?
Figure 18
Figure 19
Figure 20
Figure 21
Figure 22
A Comparison Between Delayed Mixing, In-
stantaneous Mixing and Homogeneous Com-
bustion by Constant Equivalence Ratio 42
A Comparison Between Delayed Mixing, In-
stantaneous Mixing and Homogeneous Com-
bustion by Constant EGR 43
A Comparison of the Indicated Enthalpy
Efficiency of Delayed Mixing, Instanta-
neous Mixing and Homogeneous Combustion 46
The Engine Combustion Configuration Meeting
an Indicated Specific Nitric Oxide Level of
1.0 and 0.1 g/I hp-hr 47
The Engine Combustion Configuration Meeting
an indicated Specific Nitric Oxide Level
of 5.0 and 2.0 g/Ihp-hr 48
Simplified Combustion Process Schematics 56
Exhaust Composition versus Air-Fuel Ratio
for 15° BTDC Spark Timing From Huls, T. A.
(1966) 62
Comparisons of Exhaust "Emissions of CVCC
Engine with Conventional Engine From
Tasuku Date et al. (197^) 66
Comparisons of Exhaust Emission at Constant
Indicated Specific Fuel Consumption From
Yasuo Sakai (197^). 67
Single Cylinder Evaluations HC plus NOX
Emissions From Mitchell et al. (1972) 71
Comparison of Premixed and Stratified
NO Concentrations as a Function of
Equivalence Ratio at 0% EGR. From
Lavoie and Blumberg (1973) 74
Nitrogen Oxide Concentration versus Over-
all Equivalence Ratio for Wide Open
Throttle and 65$ Prechamber Volume, From
Ingham (1976) 77
x
-------
Page
Figure 23 Gaseous Composition in the Center of the
Fuel Spray as a Function of Crank Angle
From Nightingale, D. R. (1975) 80
Figure 24- Nitric Oxide Concentration Versus Equiva-
lence Ratio for Samples Taken From Within
a Direct Injection Diesel After Combus-
tion 82
Figure 25 NOX Concentration, Naturally Aspirated
Direct-Injection and Prechamber Engines
From Pischinger, R. (19?6) 84
Figure 26 Comparisons Between Nitric Oxide Emis-
sions From Models and Measured Results 86
Figure 2? Test Facility Schematic Diagram 91
Figure 28 Nitrogen Oxide Emissions versus Overall
Equivalence Ratio for Various Air Injec-
tion Timings With a Large Diameter
Nozzle 125
Figure 29 Carbon Monoxide Emissions versus Overall
Equivalence Ratio for Various Air Injec-
tion Timings With a Large Diameter Nozzle 12&
Figure 30 Hydrocarbon Emissions versus Overall
Equivalence Ratio for Various Air In-
jection Timings With a Large Diameter
Nozzle I29
Figure 31 Measured and Corrected Indicated Enthalpy
Efficiency versus Overall Equivalence Ratio
for Various Air Injection Timings With a
Large Diameter Nozzle 131
Figure 32 Nitrogen Oxide Emissions for Various
Homogeneous Operating Conditions 137
Figure 33 Nitrogen Oxide Emissions versus Overall
Equivalence Ratio for Selected Air Injec-
tion Operating Conditions With A Small
Diameter Nozzle 139
xi
-------
Page
Figure 3^- Nitrogen Oxide Emissions versus Start of
Air Injection for Overall Stoichiometric
Equivalence Ratio With a Small Diameter
Nozzle 142
Figure 35 Carbon Monoxide Emissions versus Overall
Equivalence Ratio for Homogeneous Opera-
tion I45
Figure 36 Carbon Monoxide Emissions versus Overall
Equivalence Ratio for Air Injection
Operation With a Small Diameter Nozzle 147
Figure 37 Carbon Monoxide Emissions Versus Start
of Air Injection for Overall Stoichio-
metric Equivalence Ratio With a Small
Diameter Nozzle 148
Figure 38 Hydrocarbon Emissions versus Overall
Equivalence Ratio for Homogeneous Opera-
tion 149
Figure 39 Hydrocarbon Emissions versus Overall
Equivalence Ratio for Air Injection Opera-
tion with a Small Diameter Nozzle 151
Figure ^0 Hydrocarbon Emissions versus Start of
Air Injection for Overall Stoichiometric
Equivalence Ratio With a Small Diameter
Nozzle 153
Figure 4-1 Measured Indicated Enthalpy Efficiency
versus Overall Equivalence Ratio for
Homogeneous Operation 154
Figure k2 Measured Indicated Enthalpy Efficiency
versus Overall Equivalence Ratio for Air
Injection Operation With a Small Diameter
Nozzle 155
Figure 4-3 Corrected Indicated Enthalpy Efficiency
versus Overall Equivalence Ratio for Air
Injection Operation With a Small Diameter
Nozzle 157
Figure *44 Measured Indicated Enthalpy Efficiency
versus Start of Air Injection for Stoichio-
metric Equivalence Ratio With a Small Dia-
meter Nozzle 158
xii
-------
Figure 45
Figure 46
Figure 47
Figure 48
Figure El
Indicated Mean Effective Pressure
versus Overall Equivalence Ratio for
Homogeneous Operation
Page
160
Measured Indicated Enthalpy Efficiency
Versus Start of Air Injection for Over-
all Stoichiometric Equivalence Ratio
With a Small Diameter Nozzle 162
Measured Indicated Mean Effective Pres-
sure Versus Start of Air Injection for
Overall Stoichiometric Equivalence
Ratio With a Small Diameter Nozzle
163
Cylinder Pressure Versus Crank Angle
for Homogeneous Operation, Air In-
jection During Combustion and Air In-
jection Before Combustion
Comparison of Measured and Calculated
Flow Restriction Pressures for the
Same Flow Rate
Figure Fl A Sketch of the Air Injection Valve
166
226
228
-------
LIST OF TABLES
Page
TABLE I Values of the Parameters for the Calcu-
lations 28
TABLE II Summary of Results for Air Injection
Operation Extrapolated to a Stoichio-
metric Fuel-Air Mixture 141
xiv
-------
NOMENCLATURE
a shape factor "a" of the Weibe combustion function
c r
a shape factor "a" of the Weibe expansion function
6
A air or air with recirculated exhaust gases
ATDC after top dead center
BTDC before top dead center
CFM cubic feet per minute
CFR Cooperative Fuel Research Engine
CO carbon monoxide
D delayed mixing stratified charge engine
EGR recirculated exhaust gas
F equivalence ratio, fuel to air ratio divided
by the stoichiometric fuel to air ratio
FE equivalence ratio of the combustion products
at the time of exhaust or the overall
equivalence ratio
Fp equivalence ratio products
H homogeneous engine
HC hydrocarbon
I instantaneous mixing stratified charge engine
IEE indicated enthalpy efficiency - percent
IMEP indicated mean effective pressure - psi
ISNO indicated specific nitric oxide - g/ihp-hr
L:A-P delayed mixing stratified charge engine, see
Chapter III, C
xv
-------
L;P-A
MBT
mc
me
NO
NOY
J\.
7
Pl
P2
PE
Pi
ppm
PROCO
Rl
R2
R;P-A
RPM
t
TiR
IOCS
v(t)
vcl
one member of the limited mixing stratified charge
engine group, see Chapter III, C
maximum advance for best torque
shape factor "m" of the Weibe combustion function
shape factor "m" of the Weibe expansion function
nitric oxide
nitrogen oxide, all of the oxides of nitrogen
combined mixed products of system 1 and system 2
product system 1
product system 2
combined products at exhaust conditions
initial pressure of the reactants
part per million
Ford Programmed Combustion Engine
reactant system 1
reactant system 2
instantaneous mixing stratified charge engine,
see Chapter III, C
revolutions per minute
time
time of combustion
time of expansion
initial temperature of reactants
Texaco Controlled Combustion System Engine
volume function
clearance volume of combustion chamber
xvi
-------
vd(t) displaced volume at time "t"
VT(t) total volume of combustion chamber at time "t"
Q overall equivalence ratio
<|>c carbureted equivalence ratio
9si start of air injection in crank angle degrees
0gp spark- timing in crank angle degrees
XVIX
-------
CHAPTER I
INTRODUCTION
The oxides of nitrogen emissions from vehicle internal
combustion engines are one of the causes of photochemical-
smog. Smog is produced by the interaction of nitrogen
oxide, hydrocarbons, air and sunlight. In an attempt to
prevent this form of air pollution, the Federal Government
has set standards for the emissions from motor vehicles.
Nitrogen oxide emissions of conventional spark ignition
engines are difficult to deal with because they tend to
be large at high efficiency and high power operating
conditions.
Our general approach to the problem was to study
engine processes, doing theoretical calculations to
determine if the nitrogen oxide emissions could be re-
duced while maintaining satisfactory engine performance
The most promising concept was then selected for experi-
mental evaluation.
A computer program approximating conventional homo-
geneous engines was used to provide a better understand-
ing of nitric oxide emissions. The computer program was
also used to determine how unusual operating conditions
influence nitric oxide emissions. A second computer pro-
gram was used to study nitric oxide emissions of stratified
-------
charge engines. From these studies it appeared that the
delayed mixing stratified charge engine concept had the
most potential for limiting nitric oxide emissions and
maintaining engine performance.
The delayed mixing stratified charge engine concept
consists of-having all of the fuel concentrated in a
rich region where combustion is initiated. After com-
bustion of the rich mixture is completed, the lean region
consisting of air or air with recirculated exhaust gas
is gradually and thoroughly mixed into the rich products
of combustion.
When the nitrogen oxide emissions of various types
of internal combustion engines were examined, it was
apparent that the delayed mixing concept was part of a
unique group of engines. This group of engines uses
similar principles of operation and has similar trends
in the nitrogen oxide emissions. The nitrogen oxide
emissions of this group tend to be lower than other en-
gines for operation with near stoichiometric fuel-air
ratios. The other group of engines is usually operated
lean to limit nitrogen oxide emissions.
The experimental program is intended to simulate
the combustion processes of the delayed mixing strati-
fied charge engine concept. The experimental engine was
not intended to represent a practical engine configuration
-------
"but was, instead, a means for testing the delayed mixing
concept, A single cylinder (CFR) engine was modified to
include an air injection valve which could inject air at
the desired time in the engine cycle. The experimental
engine was operated at a variety of conditions which
provided the results necessary to evaluate the delayed
mixing concept.
-------
SUMMARY
The theoretical studies of engine' power, fuel
economy and emissions shown in Chapter II indicate that
the delayed mixing stratified charge engine concept is
capable of limiting the nitric oxide emissions while
having reasonable efficiency and specific power. The
delayed mixing process consists of rich combustion fol-
lowed by air being mixed into the rich products. During
the initial phase of rich combustion little nitric oxide
is formed due to the lack of oxygen in the charge. When
the air is mixed into the rich products of combustion,
two factors can be used to limit nitric oxide. First,
if the final overall fuel-air ratio is kept near stoichio-
metric, the nitric oxide will be low because the mixture
is rich for most of its kinetic history. However, if
the overall fuel-air ratio is made lean, the product
mixture will have both the high temperatures and oxygen
required for nitric oxide formation. With delayed mixing
and a lean overall fuel-air mixture, the nitric oxide
emissions can be very high. The second factor which
can be used to limit nitric oxide formation is the time
at which expansion cooling stops the nitric oxide
kinetics. If the end of the active nitric oxide kinetic
period were to occur before the air is added to the rich
products, the final nitric oxide concentration would be
-------
that associated with rich combustion. Likewise, as the
end of the active nitric oxide kinetic period is extended
into the mixing process, the final nitric oxide concen-
tration will "be more affected by the mixing process.
It is shown in Chapter III that some real engines
have low nitrogen oxide emissions when operated near
stoichiometric fuel-air ratios. Those engines and the
delayed mixing concept have similar combustion processes.
This group has been defined as the limited mixing strat-
fied charge engine group because the products of combus-
tion are by some means temporarily prevented from mixing
with the rest of the charge. The mixing process which
follows combustion is characterized by the products
mixing into the air, or by the air mixing into the pro-
ducts, or by some intermixing process. Engines are
judged to be part of the limited mixing group by examin-
ing the influence of fuel-air ratio on the nitrogen oxide
emissions and by examining the combustion processes. The
Ford Programmed Combustion (PROCO) Engine, the Newhall
Engine and the divided chamber diesel engine are included
in the limited mixing stratified charge engine group.
The delayed mixing stratified charge engine concept
was simulated by a single cylinder (CFR) engine with pro-
visions for air injection. A rich charge is drawn into
the combustion chamber through the normal inlet valve.
-------
The equivalence ratio of this charge is called the car-
bureted equivalence ratio. At the desired time in the
cycle the air is injected which reduces the equivalence
ratio to the value called the overall equivalence ratio.
Details of the experimental engine, the entire test
facility and instrumentation are presented in Chapter IV.
The experimental engine was operated with air in-
jection occurring after combustion to simulate the de-
layed mixing processes. The experimental results are
presented in Chapter V. A comparison can be made of the
emissions for homogeneous operation and for delayed mixing
operation at the same overall equivalence ratios. The de-
layed mixing operation is much lower in nitrogen oxide
emissions, lower in hydrocarbon emissions and about the
same in carbon monoxide emissions. From the hydrocarbon
and carbon monoxide emissions, it can be concluded that
the combustion process is completed. The magnitude of
the nitrogen oxide emissions from the delayed mixing
operation is similar to that of homogeneous operation
if its overall equivalence ratio were equal to the car-
bureted equivalence ratio of the delayed mixing engine.
Apparently, little additional nitrogen oxide is formed
due to air injection, because the completion of the com-
bustion occurs late in the expansion when the temperatures
are relatively low. Results with a large diameter or
with a small diameter air injection nozzle were similar.
-------
The indicated enthalpy efficiency of delayed mixing
operation is very low compared to that of homogeneous
operation at the same overall equivalence ratio. Since
the completion of combustion occurs late in the expan-
sion little additional work can be obtained. Attempts
were made to inject the air during the combustion pro-
cess in order to extract more work and find the trade-
off between nitrogen oxide emissions and efficiency.
These experiments resulted in the discovery that air in-
jection during combustion would disrupt the combustion
process. In a sense the air injection would tend to
blow out the flame. As a result the hydrocarbon emis-
sions increased, the carbon monoxide emissions increased,
the efficiency remained low and the nitrogen oxide in-
creased. Other tests were run with air injection be-
fore combustion. In these tests, all the emissions were
similar to those of homogeneous operation at the same
overall equivalence ratio. The efficiency approached
that of homogeneous operation at the same overall
equivalence ratio. Efficiency was less than with homo-
geneous operation because of the greater heat transfer
rates due to motion of the charge caused by air injection.
The motion of the charge also resulted in very stable and
rapid combustion.
Another factor which reduces the efficiency farther
is the work required to compress the injected air. In
-------
the test facility the air is compressed separately and
injected as desired. The measured power and correspond-
ing measured efficiency do not include the work required
to compress the injected air. Estimates of the air in-
jection compression power are made and subtracted from
the measured power to obtain the corrected power and
corresponding corrected efficiency. Generally the mea-
sured efficiency is slightly greater than homogeneous
operation at the carbureted equivalence ratio and the
corrected efficiency is lower.
In order to evaluate the potential of the delayed
mixing concept, it is necessary to know if the low
efficiency is an inherent result of the combustion pro-
cess or if it is a result of the means used to simulate
the combustion process. In the Newhall engine, the PROCO
engine and the divided chamber diesels the combustion
process, which is similar to the delayed mixing concept,
is achieved through the use of fuel injection. Thus, it
would seem possible to devise a delayed mixing engine
that would use fuel injection instead of air injection
to obtain stratification. The resulting efficiency would
not include an external compressor. Much of the efficiency
of the experimental engine can be attributed to the air in-
jection which is the means selected to simulate the delayed
mixing process.
8
-------
A certain amount of inefficiency is due to the com-
bustion process itself. The fact that the combustion is
delayed will mean that it will occur later in the expan-
sion and consequently be less efficient. The Newhall
engine, El-Messir (1973). has a lower efficiency than a
corresponding homogeneous engine. The divided chamber
diesel, Obert (1968), has a lower indicated enthalpy
efficiency than a corresponding open chamber diesel.
The calculation of both the efficiency and the IMEP
depend upon the power of the engine. For the same
fuel flow rate, the efficiency and the IMEP will be pro-
portional to each other. A comparison can be made between
operation with air injection after combustion and homo-
geneous operation at the same mass flow rate of fuel
and the same total mass of air. The low efficiency
associated with air injection after combustion results
in correspondingly low IMEP.
The measured IMEP of the experimental engine simu-
lating delayed mixing was found to be somewhat greater
than that of homogeneous operation. The reason for
the greater IMEP is that the carbureted air flow rate
was maintained constant with air injection adding to
the total mass of the charge. In a sense the air injec-
tion is supercharging the engine.
-------
In summary, the delayed mixing stratified charge en-
gine concept is an effective method of controlling the
nitrogen oxide emissions. The carbon monoxide emissions
are equivalent to those of homogeneous engines and the
hydrocarbons are lower. The disadvantage of the delayed
mixing concept is its low efficiency inherent in the
combustion process.
10
-------
CONCLUSIONS
1. The computer models of homogeneous and stratified
charge engines indicated that the delayed mixing
stratified charge engine concept could be used to
limit nitric oxide emissions, while having reason-
able specific power and efficiency.
2. In general engines can be divided into two groups
representing markedly different responses of
nitrogen oxide emissions to changes in the overall
fuel-air ratio. The first group is characterized
by producing the maximum nitrogen oxide (either ppm
or g/kw-hr) near stoichiometric fuel-air ratios or
slightly lean. The second group, which includes
the delayed mixing concept, is characterized by
producing the maximum nitrogen oxide emissions at
leaner fuel-air ratios and being less sensitive to
changes in the fuel-air ratio.
3. Experimental simulation of the delayed mixing strati-
fied charge engine concept has shown that nitrogen
oxides are much lower than for a corresponding homo-
geneous engine. The emissions of hydrocarbons are
lower than for a corresponding homogeneous engine, while
the carbon monoxide emissions are about the same.
Efficiency is much lower than for a similar homogen-
eous engine.
11
-------
4. The difference "between the predictions of reasonable
efficiency and the low efficiency found with the
simulated delayed mixing engine is due to the late
time in the expansion when the combustion process
is completed and due to external work required to
compress the injection air.
5« Air injection resulted in increased motion of the
charge which increased heat transfer rates.
6. Air injection during combustion disturbed the com-
bustion process which resulted in greater hydro-
carbon emissions, greater carbon monoxide emissions,
and lower efficiency. '
7. Air injection before combustion resulted in emissions
similar to those for a corresponding homogeneous
operation and efficiencies slightly below the
corresponding homogeneous operation. The combus-
tion process was found to be very rapid with re-
duced cycle to cycle peak pressure variations.
8. The major disadvantage of the delayed mixing
stratified charge engine concept is its inherent
low efficiency resulting from the extended combus-
tion process. Attempts at reducing the combustion
period by injecting air earlier resulted in disturbing
the rich combustion process without increasing the
efficiency.
12
-------
CHAPTER II
A THEORETICAL STUDY OF NITRIC OXIDE EMISSIONS
FROM INTERNAL COMBUSTION ENGINES
A. INTRODUCTION '
It is the objective of this study to establish a con-
cept for an internal combustion engine that would limit
the nitric oxide emissions with minimal sacrifice in
specific power and efficiency. Initially the work was
aimed at understanding the formation of nitric oxide. Of
the various computer models which were developed, the most
significant are the homogeneous and stratified charge en-
gine models. A paper which describes these engine models
entitled "A Search for a Low Nitric Oxide Engine" was pre-
sented at the Central States Section of the Combustion
Institute in March, 19?4, (unpublished) and at the Inter-
national Stratified Charge Engine Conference in October,
1974, SAE 7^1172, Evers, Myers and Uyehara (197*0. Excerpts
from the paper are presented to describe the nitric oxide
emissions from homogeneous and stratified charge engines.
A large number of combustion and expansion configura-
tions and processes were studied where exhaust NO concen-
X.
trations might vary by several orders of magnitude. Be-
cause of the number of calculations needed short computa-
tional times were judged more important than a high degree
13
-------
of precision. It was further felt that after identifi-
cation of low NO configurations and processes which
Jv
also had reasonable efficiency and specific power, more
detailed and more precise simulations could be run if
desired. Consequently, a simplified computer program
was developed.
In developing the simplified computer program it
was deliberately kept flexible and not identified with
a particular engine configuration. This resulted in
the arbitrary specification of a number of input items.
However, in general, when specifying input items a
specific engine "configuration" is typically in mind.
B. NITRIC OXIDE KINETICS
The kinetics of nitric oxide have been used by many
investigators in predicting the nitric oxide concentra-
tion in the products of combustion. The method requires
the selection of appropriate elementary chemical re-
actions, the corresponding reaction rates, and a group
of simplifying assumptions.
Newhall and Shahed (1971) showed that the two elemen-
tary chemical equations A and B (called the Zeldovich
mechanism) satisfactorily predicted the nitric oxide
concentration of fuel-lean mixtures but somewhat under-
estimated the concentration of fuel-rich mixtures.
14
-------
N2 + 0 £ NO + N (A)
02 + N £ NO + 0 (B)
The tests were performed in a closed constant volume cy-
lindrical combustion chamber. In order to perform the
kinetic calculations, Newhall assumed that all of the
species in the combustion products except nitric oxide
were at equilibrium concentrations.
The Zeldovich mechanism is used for all the kinetic
calculations presented in this chapter. The equations are
easily handled in computer calculations. For more infor-
mation about the method used to calculate nitric oxide
concentrations see Appendix A. Additional equations can
be added to the Zeldovich mechanism in hopes of improving
the accuracy of kinetic calculations. Since we are look-
ing for large changes in nitric oxide levels, the Zeldovich
mechanism is sufficient.
Lavoie et al. (1970) used an expanded set of elemen-
tary chemical reactions which added equations C through F
to the Zeldovich mechanism.
N + OH £ NO + H (C)
H + N20 £ N2 + OH (D)
0 + N20 - N£ + 02 (E)
0 + N£0 £ NO + NO (F)
They also showed that the N and NgO concentrations exist
nearer a steady state value than an equilibrium value.
15
-------
Thus they assumed that N and N20 exist at the steady state
concentrations and all the other species, except NO, exist
at equilibrium concentrations. A computer model based on
this kinetics mechanism was found to be in agreement with
internal combustion engine test results. The comparison
was made for- fuel-lean mixtures, No comparisons were
offered for fuel-rich mixtures. Lavoie pointed out that
equations D, E and F are only important for lean mixtures
at low temperatures.
A thorough review of nitric oxide kinetics done by De
Soete (1975) concluded that equations C and D had little
effect for equivalence ratios less than 1.5 and temperatures
greater than 3240°R (1800°K). DeSoto's choice of the most
acceptable set of kinetics equations are the Zeldovich
mechanism (equations A and B) plus equations E and F.
Fenimore (1971) showed that for fuel-rich hydrocar-
bon flames the Zeldovich mechanism did not predict the
nitric oxide formed in the flame. He suggested that
nitric oxide formed in the flame was due to a different
kinetic mechanism such as equations G and H. The nitric
oxide generated by that mechanism is called "prompt NO".
CH + N2 £ HCN + N (G)
G2 + N2 * 2CN (H)
The assumption of equilibrium concentration of atomic
oxygen in the flame was investigated by Iverach et al.
16
-------
(1973)• They measured the nitric oxide concentration in
a rich premixed flame and "back calculated the Zeldovich
mechanism to obtain the atomic oxygen concentration re-
quired to produce that quantity of nitric oxide. Through
other measurements they were able to determine the actual
atomic oxygen present in the flame. A comparison shows
that the atomic oxygen concentration required by the
Zeldovich mechanism is orders of magnitude greater than
the actual concentration for hydrocarbon flames and about
equal for CO/H2 flames. They concluded that prompt NO
was formed by a different kinetic mechanism as suggested
by Fenimore. Their choice for the mechanism is equation
G and I.
C + N2 ? CN + N (I)
By only using the Zeldovich mechanism in our calcu-
lations we expect to somewhat underestimate the nitric
oxide* concentrations for fuel-rich combustion.
C. COMPUTER MODELS
Since the nitric oxide concentration is not markedly
affected by the intake, compression,or exhaust processes,
they were not included in the computer program in the
interest of reducing the computer time. Calculations of
specific power and efficiency required estimation of the
work of these processes. The intake and exhaust work was
17
-------
assumed to "be zero. An isentropic compression to the state
specified at the beginning of combustion was used to esti-
mate the work of compression. It was further decided not
to include heat transfer in the computer programs. Ne-
glecting the heat transfer shortens the computer time while
resulting in greater pressures, temperatures and an over-
estimate of the nitric oxide concentration. In essence,
the approach was to treat the gases as being in one or
more closed adiabatic containers of uniform pressure with
specified initial conditions and an arbitrarily specified
expansion and combustion process.
1. One System Model
The one system model is intended to describe homo-
geneous combustion similar to a conventional homogeneous
charge spark ignition engine, but in a general manner.
To describe the regions within the one-system model we
considered two approaches. In both approaches the re-
actants were assumed to be one homogeneous region. The
mixed products approach assumes the products to be one
homogeneous region with constant properties throughout,
including pressure and temperature. As new products of
combustion are formed they are instantaneously mixed with
existing products which can change the average properties
of the product region. The unmixed products approach
assumes the product region to be made up of a number of
18
-------
small subregions each with its own individual properties.
Each subregion could have a different temperature but they
all have the same pressure. As products of combustion are
formed they are grouped into the small subregions which
maintain their own identity and do not mix with the other
products.
An advantage of the unmixed products approach is that
temperature gradients and their influence on nitric oxide
formation can be shown in the product region. Blumberg
and Kummer (1971) compared the nitric oxide formed by
the mixed and unmixed products approach. They found the
ratio of nitric oxide predicted by the mixed products
approach to nitric oxide predicted by the unmixed pro-
ducts approach to vary by only ± 1?$ over a range of
equivalence ratios from 0.7 to 1.3-
The mixed products approach was selected for these
calculations because it was sufficiently accurate and re-
quired less computer time. The computer time for the
mixed products approach is shorter since it is only neces-
sary to keep track of two regions, products and reactants.
With the unmixed products approach it is necessary to keep
track of the reactants plus the many small subregions of
products.
The thermodynamic equations and computer programs used
to describe the one-system model are given in Appendix B.
19
-------
2. Two System Model
The two system model is used to represent nonhomo-
geneous combustion as occurs in stratified charge engines.
The program consiste of two one-system models each capable
of handling two regions, reactants and products. As is
shown in Figure 1, initially both system 1 and system 2
are homogeneous reactants.
In the general case, reactants in both systems would
be converted to products but not necessarily at the same
rate. After some interval to be specified (called the
time of combination) the two systems are combined to form
a single system having three regions as shown in Figure
1. The reactants remaining in system 1 or 2 at the time
of combination are maintained intact but the products
from both system 1 and 2 are immediately mixed. See
Appendix C for details of the two system model and computer
program.
All calculations reported herein used the special
cases of either delayed or instantaneous mixing models
as shown in Figure 1. In both the delayed and in-
stantaneous mixing models the reactants in system 2(A2)
are either air or a homogeneous mixture of air plus ex-
haust gas recirculation (EGR) of exhaust composition.
For delayed mixing the combination of system 1 and system
2 occurred immediately but Ag did not start to mix with
20
-------
General Case:
-System 1
-System 2
time of combination
E -
v(t)
Reactant System 1
Reactant System 2
Product System 1
Product System 2
Combined mixed pro-
ducts of System 1
and 2
Combined Products
at Exhaust Condi-
tions
- Volume function
(crank and piston)
Instantaneous Mixing Case:
- Air or air
plus EGR
Delayed Mixing Case:
Figure 1 Two System Model of the Stratified
Charge Engine Configurations
21
-------
products P, until system 1 was entirely burnt. Then A2 is
mixed with P-, at a specified rate until complete mixing
occurs and only products having the exhaust air-fuel ratio
exist in the system.
In instantaneous mixing, combination occurs immediate-
ly; R, is converted to P, at a specified rate; and P-, is
instantaneously mixed with Ap. To illustrate, if the rich
products of system 1 are mixed with A2 (let A2 be air),
the composition of [A2 + SAP,] (see Figure 1) will change
continuously from that of air to that of exhaust products
as R, is burnt. This is in contrast with delayed mixing
where, if A2 is air and R, is a rich mixture, the com-
position of [P, + £AA21 will gradually change from rich
products to products having exhaust composition.
D. POTENTIAL ENGINE CONFIGURATION
1. Homogeneous Charge-Conventional SI Engine
As indicated previously, when specifying initial and
other conditions necessary to run the program,one tends
to have a specific engine configuration in mind. The
first configuration which was studied primarily for re-
ference purposes was basically the conventional spark
ignition engine with a wide open throttle. In order to
use the program for this configuration, the program re-
quires that the pressure, temperature and composition at
the start of combustion, the rate or extent of combustion
22
-------
and a function "to describe the system volume "be specified.
The Weibe (1956) function was selected to represent
both the extent of combustion and the system volume.
Figure 2 illustrates use of the Weibe function to describe
the extent of combustion. Three variables are required to
define a specific combustion process. For the calculations
presented the time of combustion (t ) was varied while m
C \-»
equaled two and a equaled five. Also shown in Figure 2
C
is a cosine burning law which sometimes is used to re-
present combustion, Heywood, et al. (1973)-
The expansion function also uses the Weibe function
to describe the displacement volume (see Figure 3)• The
total volume of the system is equal to the displacement
volume plus the clearance volume. A clearance volume was
selected which is equivalent to a compression ratio of
nine. Since the expansion function shows an increasing
volume from the clearance volume to the maximum volume
this is analogous to a crank and piston at top dead center
and at bottom dead center. Both the time of expansion (t )
"
and m_ were varied while ao was held constant at five.
c e
The initial condition of the one-system model can be
defined by its pressure, temperature, fuel-to-air equiva-
lence ratio, fuel composition and the amount of recircula-
ted exhaust gas. The fuel was taken to be GgH-,^. It was
assumed that the fuel-to-air equivalence ratio for the
23
-------
in
0)
CO
o
to
o
•»—
o
II
[5
•s
i
to
C JD
£ E
x o
UJO
Wiebe Function :
Extent of Combustion (a) =
l.-e-ac(J-)(mc+l]
V'c/
.4 .6
Time(t)
.8
Time of Combustion (tc)
—Time of Combustion
1.0
Figure 2 The Combustion Function for the
One-System and Two-System Models
24
-------
Total VolJvTit)j = Clearance Vol. (Vc))
•f Displaced Vol. [Vd(t)]
VT(t) VC| , Vd(t)
o
ID
CD
U
52
o.
00
Where:
VCI
Vd(te) Compression
Ratio
Vdlt)
0
0
.2
.4 .6
Time(t)
.8
Figure 3
Time of Expansion (te)
— Time of Expansion —-]
The Expansion Function for the One-
System Model (Weibe) and the Two-
System Model (Crank and Piston)
25
-------
recirculated exhaust gas is the same as that of the reac-
tants.
Three parameters were selected to characterize the
performance of an engine. These are indicated specific
nitric oxide (ISNO) which relates the pollution level to
engine power,, indicated enthalpy efficiency (IEE) which
measures the utilization of the fuel, and indicated mean
effective pressure (IMEP) which is related to specific
power output and thus" relative engine size. These terms
are defined in Appendix D which also shows how the compres-
sion work was calculated. See Appendix I for SI units.
One set of calculations represented a study of seven
independent variables. Four of these variables were the
initial conditions, two of these variables were t and m
e e
from the expansion function and one variable was t from
\*r
the combustion function. Each of seven variables was
assigned a high and a low level. All possible combination
of these variables were calculated (2' = 128 runs).
From the seven variable-two level calculations, it is
possible to establish that the four major variables in-
fluencing nitric oxide levels are the fuel-to-air equiva-
lence ratio, the initial temperature of the reactants, the
exhaust gas recirculation, and the time of expansion.
Consequently, another parametric set of calculations was
1. As discussed by Lauck et al. (1962), the indicated
enthalpy efficiency is numerically equal to thermal
efficiency which is more commonly but incorrectly used,
26
-------
performed with these four variables. The values of the
variables used in these calculations are shown in Table I.
A total of 135 calculations were made to represent all
possible combinations.
The results of the calculations are presented in
Figures ^ and 5« In each curve shown in Figure 4,indicated
specific nitric oxide is plotted versus indicated mean
effective pressure for the fifteen combinations of equiva-
lence ratio and fraction of exhaust gas recirculation.
Each of the nine curves represent one of the nine com-
binations of the three values of time of expansion and
the three values of initial temperature of the reactants.
The first observation is that the nine indicated
specific nitric oxide curves are very similar in shape to
each other. Also, not unexpectedly, only at high equiva-
lence ratios (rich mixtures) or low equivalence ratios
(lean mixtures) does the indicated specific nitric oxide
reach desirably low levels. There does not exist some
fortuitous combination of variables which excludes the
influence of equivalence ratio. Exhaust gas recircula-
tion in combination with either rich or lean mixtures
contributes greatly to the reduction of nitric oxide.
The addition of exhaust gas recirculation or cooler
initial temperatures results in lower product temperatures
and consequently lower nitric oxide concentrations.
27
-------
TABLE I
Values of the Parameters for the Calculations
Show in Figures 4 and 5
Parameters
Equivalence Ratio (F)
Initial Temperature of
the Reactants (T-R)
Exhaust Gas Recirculation
(EGR)
Initial Pressure (P.)
Time of Expansion (t )-sec
Expansion Shape (me)
(ae)
Time of Combustion
(tc)-sec
Combustion Shape (mr)
Values^
.6, .8, 1.0, 1.2, 1.4
800.°R, 1200.°R, 1400.°R
(444.°K), (667.°K), (778°K)
.0, .1, .2
250. psi (172400. N/m2)
.005, .025, .05
(^6000 rpm)(^1200 rpm)(^ 600
rpm)
1.
5.
.05
2.
5.
28
-------
INITIAL TEMPERATURE OF THE REACTANTS ( T,R)
I400°R I200°R 800°R
—. 10
ft
£s
8g..
*>
m
q
Ol
001 »o ' «io ' 120 ' i*
»O HO WO 200
IMEP-pti
*-. 10
E
a.
g
(O
1
X
UJ
UL
o
UJ
ll
w o .1
9> 2
S »
m
CVJ
Q «
to 100 110 MO ieo leo 200
IMEP-pti
*0 DO 120 MO KO *O ZOO KO 200 Z2O (40
IMEP-p«i IMEP-ptl
80 IOO 120 I4O KO t*O 2OO
IMEP-p»i
•0200i202402S02BOMOI20
IMEP-pti
E
Q.
O .
O Jr
O I
«> ^
So,
10 s
10 -
o
Q
80 IOO 120 140 ISO 180 200 °°' 80 IOO 120 140 160 180 200 »0 200 220 240 2*0 280 JOO KO
IMEP-psi
IMEP-psi
IMEP-psi
Figure 4.
Indicated specific nitric oxide
versus indicated mean effective pressure
(i:iEP) for homogeneous charge engine
configuration .
29
-------
u>
o
60
50
LU
40
30-
20
INITIAL TEMPERATURE OF THE REACTANTS (TIR)
1400° R
I
l , 1
I , I
8O 100
120 140 160
IMEP-psi
180 200
60
50
UJ
UJ
40
30
20
I200°R
PI o
itil uii 3,
1 'F.I.2.'
/
7
/ /
1
80 IOO 120 140 160
IMEP-psi
ISO 200
20
800°R
180 200 220 240 260 280 300 320
IMEP-psi
Figure 5. Indicated enthalpy efficiency (IBE) versus
indicated mean effective pressure (IMLP)
for homogeneous charge engine configuration.
-------
Peak specific nitric oxide levels occur just to the
lean side of stoichiometric except for high engine speeds.
The peak nitric oxide levels occur in the lean region be-
cause of the greater oxygen concentration and high asso-
ciated temperatures. As the mixtures become leaner, the
flame temperatures decrease substantially and so do the
nitric oxide rate constants and the amount of nitric oxide
formed. However, note the change in the effect of EGR
(at F = 1.0, for example) from the lower right curves
hand to the upper left hand curves in Figure b.
The indicated mean effective pressure is changed by
those factors which influence the specific volume of the
charge or the amount of fuel. Reduced temperature, higher
fuel air ratios in lean region, or less exhaust gas recir-
culation all result in a greater indicated mean effective
pressure. The magnitude of the IMEP is greater than
conventional spark ignition engines because the initial
conditions resulted in a larger mass of charge. Calcu-
lations with a smaller charge have the same trends.
In the computer model, a decrease in the time of
expansion results in lower nitric oxide concentrations
because less time is available for nitric oxide formation.
In the model we have assumed no heat transfer. In an
actual engine, the amount of heat transfer also depends on
the time of expansion (or cycle time) and a decrease in the
31
-------
time of expansion (decreased cycle time) would cause de-
creased heat transfer which will tend to increase the level
of nitric oxide concentration. Actual nitric oxide concen-
trations would reflect a balance of these effects. Since
any engine coupled to the wheels by conventional trans-
missions would operate over a range of expansion times,
it would be difficult to take advantage of possible bene-
fits of rapid expansion to limit nitric oxide formation.
The indicated enthalpy efficiency curves, Figure 5*
have well known shapes. Changes in indicated mean effec-
tive pressure stretch the curves. The indicated enthalpy
efficiency is seen to be mainly a function of equivalence
ratio with a small change due to exhaust gas recircula-
tion. At high equivalence ratios the efficiency drops
due to unburnt fuel. As the mixture becomes leaner than
stoichiometric the efficiency increases because of the
lower temperatures and the accompanying reduction in
specific heats and dissociation.
2. Stratified Charge Engine
Generally, in a stratified charge engine having the
objective of reducing NO 'emissions, combustion takes
.A.
place in a rich region and the rich products and air are
combined in a prescribed way to give a stoichiometric (or
leaner) exhaust mixture. The two cases - delayed and in-
stantaneous mixing - described under the two-system model
represent two extremes of mixing and were used to study
32
-------
stratified charge engine performance. Crank and piston ex-
pansion was used for the two-system model because of the
minor effect of changes in the expansion function had on NO.
The instantaneous mixing model is somewhat similar to
an engine having a prechamber in which the rich reactants
are burned. The products leave the prechamber and are
rapidly mixed with the air in the main chamber. In the
instantaneous mixing model the mixing would be instantaneous,
the pressure in both chambers would be the same and the
volumes of both the prechamber and the main chamber would
change. In the delayed mixing model the air would be in
the prechamber. After combustion, air would gradually be
forced out of the prechamber and into the products by
changes in the volumes of the main chamber and prechamber.
2
Let us define the time of heat release to include
both the energy release due to rich combustion and the
energy release when additional air is supplied to the pro-
ducts of rich combustion. Note that heat release does not
include energy release due to changing chemical equilibrium.
The first computations that were run compared the effects
of changing the time of heat release on delayed and in-
stantaneous mixing (see Figure 6), when the exhaust pro-
ducts are stoichiometric and the overall exhaust gas re-
circulation is twenty percent. The relative shapes of the
2. The term heat release is used in a descriptive sense,
and refers to the analogy between combustion and heat
addition processes.
33
-------
100
10
T I
Q.
.I
C/)
.01
n
QoL/3
.00
80 100
120 140 160
IMEP-psi
180 200
Figure 6
The Influence of the Times of Heat Release
for a Homogeneous Charge Engine Operating with
F * 1.0 and EGR = .2 and the Delayed and
Instantaneous Mixing Stratified Charge Engine
Configuration Operating with Fexhaugt =1.0
and EGRoverall = .2
34
-------
Delayed Mixing
D, D:
D
c
o
o ^
-L. ±-
co nn
to"-1
o
^
1.0
.8
.6
.4
.2
0
1 C_
X*^ ^*** *^ ***
xx^<,4
co GQ o
O '^
.002 .004 .006 .008
Time-seconds
.010
Instantaneous Mixing and
Homogeneous Combustion
H2i2 Hsls
.002 .004 .006 .008
Time-seconds
.010
Figure 7 A Comparison of the Heat Release for
Instantaneous Mixing, Delayed Mixing
and Homogeneous Combustion Corresponding
to Figure 6,
35
-------
heat release curves are shown in Figure ?• It should "be
noted that,as the duration of heat release was increased,
the major amount of combustion for instantaneous mixing
occurred considerably later than that for delayed mixing.
This occurred because,in the delayed case,only the period
of mixing and not the duration of the rich combustion was
changed.
In spite of the relatively delayed combustion for the
instantaneous mixing case, it shows significantly higher
NO for the same IMEP. The reasons for this unexpected
JL.
result is that for instantaneous mixing the product mixture
[&2 + SAP-,] is always in the lean region starting with es-
sentially air passing through the relatively high NO for-
j\.
mation zone of around 0.8 equivalence ratio and finally
reaching a F = 1.0. For delayed mixing the product mixture
[P, + ZAApl is on the rich side starting with F = 1.^ and
ending with F = 1.0.
The next set of computations were run with a constant
time of heat release but the overall EGR and overall fuel-
air ratio were varied. Recirculated exhaust gas was
assigned the composition corresponding to the final over-
all equivalence ratio. Heat release patterns for both
delayed and instantaneous mixing are shown on the top of
Figure 8 and corresponds to Dg and I~ in Figure ?• Also
included in Figure 8, on the bottom, is the variation of
36
-------
c
o
o ^
•*• ^**
tj
$ QQ
o
^~
^
"~O-
O Li-
CE •£
o
(D 3
O "O
c o
— - rT
o
"5 .c
cr-*-
LU M-
^>
1.0
.8
.6
.4
Q
Delayed <,•**"*"'
- Mixing-j^ ^**' ionn DD^/I
s~*~~'/ l^ivjvj r\ i ivi
/ '
- 1 /
» y ^""Instantaneous
// Mixing
^•< 1,1,1,1,1
0 .002 .004 .006 .008 .010
Time-seconds
1.6
1.4
1.2
1.0
.8
.6
.4
o
^Delayed Mixing P, + ZAA2
-~" """•^N|^
\XV
N. ^^ Exhaust F=I.O
X. „. ^. M«— — — — —
^•- \
/ v.
/ x^^ Exhaust F = . 6
/ .^ —•"*"*• ~~~"
/ s^ I20° RPM
/ ' EGRn=0
/ /-^-Instantaneous u
// Mixing A2+2AP,
.002 .004 .006 .008 .010
Time -seconds
Figure 8 The Mass Fraction Burned and the Product
Mixture Equivalence Ratio (for Two of
the Nine Conditions) Corresponding to
Figure 9 and Figure 10
37
-------
100
INSTANTANEOUS MIXING
10
.01
.001
1 I
I I I
1
I i I i I L
80 100 \20 140 160 180 200
IMEP-psi
Figure 9 Indicated Specific Nitric Oxide
for Instantaneous Mixing Stratified
Charge Engine Configuration
38
-------
100
DELAYED MIXING
10
I
O
z:
CO
.01
.001
lilt
80 100
120 140 160
IMEP-psi
180 200
Figure 10
Indicated Specific Nitric Oxide
for Delayed Mixing Stratified
Charge Engine Configuration
39
-------
the product mixture equivalence ratio for delayed and in-
stantaneous mixing at two different exhaust equivalence
ratios (is also called overall equivalence ratio) with
no EGR. The resulting computed indicated specific nitric
oxide for instantaneous mixing is shown in Figure 9 while
ISNO for delayed mixing is shown in Figure 10,
The trends shown in Figure 9 for instantaneous mixing
are obviously different from those for delayed mixing
shown in Figure 10. These differences can readily be ex-
plained by recalling that, in general,the maximum nitric
oxide is formed when the equivalence ratio is around F =
.8. Thus,if the equivalence ratio of the product mixture
is required to pass through equivalence ratios near F = .8,
the nitric oxide levels will be high. If the F = .8 region
can be avoided at least until expansion has cooled the
products, the nitric oxide levels will be low. For ex-
ample (as can be seen in the lower part of Figure 8), with
an exhaust equivalence ratio of 0.6, the product mixture
[A2 + ZAp ] for instantaneous mixing never passes through
a fuel-air ratio having rapid NO formation rates, i.e.,
A.
it never exceeds F = 0.6. By contrast, for delayed mixing
with an exhaust equivalence ratio of 0.6, the product mix-
ture [P, + £AA9] passes through the fuel-air ratio having
-L <—
rapid NO formation, i.e., F = 0.8 region prior to reach-
Jt
ing its final exhaust value of F = .6. Consequently, as
40
-------
shown in Figure 10, exhaust NO concentrations with delayed
X.
mixing increase with leaner exhaust mixtures rather than
decreasing as in the case of instantaneous mixing, the
exact trend depending upon the amount of EGR. This effect
is further accentuated by the rich fuel-air mixture "being
diluted with the recirculated exhaust gases having a lean
equivalence ratio corresponding to the exhaust equivalence
ratio. The resulting products have an equivalence ratio
that is closer to stoichiometric than the rich air-fuel
mixture. It should be clear, however, that regardless of
how rich the fuel-air mixture is made [P-. + ZAA, ] must
pass through the rapid NO formation region, if the ex-
X
haust is to have lean products composition.
In order to obtain a direct comparison between in-
stantaneous mixing and delayed mixing,the data presented
in Figures 9 and 10 are replotted by equivalence ratio
and EGR as shown in Figures 11 and 12,respectively. Also
included in Figures 11 and 12 is a homogeneous combustion
case which corresponds to the same time of combustion and
has a heat release characteristic identical to that of
instantaneous mixing.
Looking first at Figure 11 for the exhaust equivalence
ratio of F = 1.0, delayed mixing has the lowest ISNO and, by
a slight amount, the greatest IMEP. The slightly greater
IMEP is due to the earlier heat release of delayed mixing.
41
-------
100
-Exhaust Equivalence Ratio
/EGR=.I
d
EGR=.2
,'EGRcf.l;
Homogeneous
Delayed Mixing
/ - •»—
instantaneous Mixing
100 130 140 140 160 160 180 200
IMEP-psi
Figure 11
A Comparison Between Delayed Mixing,
Instantaneous Mixing and Homogeneous
Combustion by Constant Equivalence
Ratio
42
-------
10 Or
EGR = .2 _
.01
.001
Homogeneous
Delayed Mixing
Instantaneous Mixing /
;
.6 .8 1.0 V.S .8 1.0 \6 .8
Equivalence Ratio ( F or FE)
i.o
Figure 12
A Comparison Between Delayed Mixing,
Instantaneous Mixing and Homogeneous
Combustion by Constant EGR
43
-------
The fact that the IMEP is slightly greater for the instanta-
neous mixing than for the homogeneous case is due to lower
temperatures of the product mixture for instantaneous mix-
ing and therefore less dissociation although the heat re-
lease curves are identical. At an exhaust equivalence ratio
of F = 0.8, the delayed mixing and instantaneous mixing pre-
dict about the same results since they are both required to
finally arrive at the high nitric oxide formation region.
From an N0x standpoint,instantaneous mixing is "by far the
most desirable at an exhaust equivalence ratio of F = 0.6,
and even homogeneous combustion is better than the de-
layed mixing,since it is not required to pass through the
region of F = 0.8. In summary, at stoichiometric mixtures,
delayed mixing gives lower NO for about the same IMEP than
Jt
instantaneous mixing. However, as the exhaust mixture is
made leaner the situation reverses until, for an exhaust
equivalence ratio of F = 0.6, the instantaneous mixing
gives N0x several orders of magnitude lower than delayed
mixing. This effect is shown more dramatically in Figure
12 where the difference in trends with equivalence ratio
is clearly illustrated. It can also be seen in Figure 12
that all three curves have different shapes at different
values of EGR. For example, with zero EGR the delayed
mixing curve is nearly horizontal while with 0.2 EGR it
has a large negative slope. Note that the shape of the
homogeneous curve is also changed significantly.
44
-------
A comparison of the efficiencies of delayed mixing,
instantaneous mixing,and homogeneous combustion is given
in Figure 13. The instantaneous mixing and homogeneous
combustion cases show about the same variation because
they both have identical heat release functions, but
the efficiency is slightly higher for the instantaneous
mixing because of lower specific heats and less dissocia-
tion. The delayed mixing is somewhat more efficient be-
cause more of its heat release comes earlier in the expan-
sion.
Figures 1^4- and 15 can be viewed as the potential en-
gine combustion configurations for an arbitrarily speci-
fied maximum permissible value of ISNO. The logical
choices would be those configurations having the highest
IMEP and efficiency, i.e., the ones nearest the upper
right hand corner. Associated with each point is a group
of symbols which specify delayed mixing (D), instantaneous
mixing (I), or homogeneous combustion (H) engine configura-
tions. The equivalence ratio (F) for homogeneous combus-
tion or the average overall exhaust equivalence ratio
(Fg) for stratified charge engine configurations is in-
cluded. Finally the amount of recirculated exhaust gas
(EGR) for homogeneous combustion or the average overall
amount of recirculated exhaust gas (EG-RQ) for stratified
charge engines is also shown.
45
-------
50
LU
LJ
40
30
20
F=I.O
EGR = 2
Homogeneous
Delayed Mixing
Instantaneous Mixing
80 100
140 160
IMEP-psi
180 200
Figure 13
A Comparison of the Indicated Enthalpy
Efficiency of Delayed Mixing, Instanta-
neous Mixing and Homogeneous Combustion
46
-------
LU
30
ISNO=!.Og/lhp-hr
n F = 6 'ihE=-'
_, EGR0=.!8*EGRo=0\ D,FE=.8
*+Dr~ x
H,F=.G
EGR=0 A • - * r^r- ~
4 EGR^=I3 • D,FE=I.O
H,F=.8/ °' EGR0=.r
EGR=.I8
35
EGR=.09H H,F = I.3
EGR=0
-i 1 1 I i I
80 100 120 140 160 180 200
IMEP-psi
ISNO =0.lg/!hp-hr
45
40
.^o
i
UJ
EGR--.I2 EGR0=.03 •D.F^I.O
I -Instantaneous Mixing EGRo=-l 9.
D-Delayed Mixing
H-Homogeneous _u_ ,.
H H,r =1.2
EGR=.I7
30
EGR=.05
25
80 100 120 140 160 180 200
IMEP-psi
Figure Ik The Engine Combustion Configuration
Meeting an Indicated Specific Nitric
Oxide Level of 1.0 and 0.1 g/l hp-hr
47
-------
ISNO=5X>.g/lhp-hr
I,FE=.8
EGRa-03
80 100 120 140 160 180 200
IMEP-psi
50
45
ISNO = 2.0g/lhp-hr
UJ
35
30
T \ 1 1 1 \ I | I
D,FE=.6 UV.73
EGR0=.li 9 EGR(!0\ D,FE=.8
H,F=.8 EGR = 0 ^
EGR=.I5
I,FF=.8
EGR0=-06
H,F=I.2
EGR=.03
80 100 120 140 160 180 200
IMEP-psi
Figure 15
The Engine Combustion Configuration
Meeting an indicated Specific Nitric
Oxide Level of 5-0 and 2.0 g/Ihp-hr
48
-------
At the very low value of ISNO = 0.1, only a few of the
combustion configurations can be used, as seen in the lower
part of Figure 1^-. The delayed mixing looks attractive be-
cause of its reasonably high IMEP and efficiency. Both
lean homogeneous combustion and lean instantaneous mixing
have slightly greater efficiencies but substantially lower
IMEP's. In contrast, the rich homogeneous combustion for
the H, F = 1.JJ-, EGR = j05 combustion configuration has a
slightly larger IMEP but a considerably lower efficiency.
These results emphasize the basic limitation of the homo-
geneous charge engine with regards to the formation of
nitric oxide. When operated rich,the efficiency is low and
when operated lean, the IMEP is low.
As the limit of ISNO is raised to 1.0, 2.0 and 5.0 as
shown in the top of Figure 1^- and in Figure 15 the choice
of configurations increases. The stoichiometric delayed
mixing stratified charge engine configuration still
appears to offer the best compromise of efficiency and
IMEP in comparison with homogeneous combustion configura-
tion. At these levels the very lean delayed mixing case
(D, FE = .6) appears to be a poor choice because of
its very low IMEP. Note that points representing an
equivalence ratio of 0.8 cluster in the center of the
figures. For this 0.8 value, delayed mixing is slightly
better than instantaneous mixing which in turn is slightly
better than homogeneous combustion.
49
-------
Homogeneous combustion and instantaneous mixing cases
with an overall stoichiometric mixture become permissible
at ISNO =2.0 and 5-0. Their efficiency is about the same
as the delayed mixing stoichiometric case but their IMEP
is less.
All of the above comparisons were made at conditions
analogous to wide open throttle. Comparisons at lighter
loads depend upon the type of load control used, i.e.,
throttling versus overall fuel-air equivalence ratio con-
trol. Note that as the overall equivalence ratio is
changed, the ratio of the volume of the fuel-air mixture
to the volume of air must change. When the equivalence
ratio is reduced for the instantaneous mixing case the
ISNO is reduced along with the IMEP as shown in Figure 9.
If the instantaneous mixing stratified charge engine were
to be controlled by changing the overall equivalence ratio,
it would have the advantage of reduced emissions at re-
duced powers. In contrast, the delayed mixing stratified
charge engine controlled by changing equivalence ratio
would have small changes in ISNO with no exhaust gas re-
circulation (See Figure 10). However when EGR = .2 the
ISNO would increase as power (equivalence ratio) decreased.
At low ISNO levels the delayed mixing engine configuration
would have to operate near stoichiometric with consider-
able EGR. Consequently, throttling would be the only
50
-------
feasible method of controlling engine power under these
conditions.
In summary, the delayed mixing stratified charge en-
gine configuration operated with an overall mixture near
stoichiometric is the "best compromise of efficiency and
IMEP at moderate to low levels of ISNO. However, the
method of load control may depend upon the level of ISNO
permitted.
E. CONCLUSIONS
1. The conventional homogeneous charge engine, "by operating
either very fuel-rich or very fuel-lean, can limit the
nitric oxide emissions.
2. Exhaust gas recirculation is an effective way of re-
ducing the nitric oxide emissions from either a homogen-
eous or stratified charge engine.
3. In the delayed mixing stratified charge concept (rich
combustion followed by the addition of air to the rich
products), extending the period during which the air is
mixed into the rich products greatly reduces the nitric
oxide level and also reduces the power.
^. For the delayed mixing stratified charge concept, the
nitric oxide levels are low when the exhaust gas recir-
culation is high and the exhaust equivalence ratio is
near stoichiometric.
51
-------
5. For the instantaneous mixing stratified charge concept
(products °f rich combustion as they are formed are in-
stantaneously mixed with the air and product mixture), the
nitric oxide levels are low when the exhaust gas recircu-
lation is high and the exhaust equivalence ratio is fuel-
lean .
6. If an arbitrary level of nitric oxide were specified,
the various engine configurations would compare as follows
Configuration Exhaust Eq. Ratio Efficiency IMEP
Homogeneous Charge Rich Low High
Homogeneous Charge . Lean High Low
Instantaneous Mixing Lean High Low
Delayed Mixing Stoich. High High
7. The delayed mixing stratified charge engine concept
has considerable promise because of its reasonably high
efficiency and IMEP associated with low nitric oxide
levels.
52
-------
CHAPTER III
A DISCUSSION OF NITROGEN OXIDE EMISSIONS FROM
VARIOUS TYPES OF INTERNAL COMBUSTION ENGINES
A. INTRODUCTION
Various types of internal combustion engines will tie
discussed and compared on the "basis of their nitrogen
oxide emissions versus overall fuel-air ratio. It will
be shown that, in general* engines can "be divided into two
groups with markedly different responses of nitrogen oxide
emissions with changes in the overall fuel-air ratio.
The first group is characterized by producing the
most nitrogen oxide emissions (either ppm or g/hp-hr) near
stoichiometric fuel-air ratios or slightly fuel lean.
This group consists of homogeneous charge engines, rapid
mixing stratified charge engines, three-valve stratified
charge engines and open chamber diesels. In the first
group the fuel is distributed throughout the air before
combustion or the products of rich combustion are rapid-
ly mixed with the lean ratios.
The second group is characterized by producing the
most nitrogen oxide emissions (either ppm or g/hp-hr) at
leaner fuel-air ratios and being less sensitive to changes
in the fuel-air ratio. This group consists of the limited
mixing stratified charge engine and the divided chamber
53
-------
diesels. In the second group of engines mixing is tempo-
rarily limited between the rich products- of combustion and
the lean region.
The two different groups can be explained on the
basis of the fundamental concepts of nitric oxide chemical
kinetics. Nitrogen oxides are formed in the products of com-
bustion because these are formed at a sufficiently high temper-
ature and have oxygen available. The nitric oxide chemical
kinetics are slowed down by cooling associated
with the expansion process of an engine. Thus nitric
oxide active chemical kinetic period starts with combustion
and ends during expansion. The availability of oxygen is
primarily a function of the local fuel-air ratio (or
equivalence ratio). The temperature of a local region of
products is a function of the flame temperature, the amount
of compression or expansion, the heat transfer, mixing with
other regions, and chemical reaction due to the mixing of
regions with different chemical composition. The fuel-
air ratio is important to the temperature of the products
because it influences the flame temperature and the chemi-
cal reaction resulting from mixing. It is not surprising
that the fuel-air ratio is a significant parameter since
it affects both the available oxygen and the product temp-
erature . The pressure also affects the nitric oxide forma-
tion rate .to a small extent and since the pressure effect
is small it will not be included in the discussion.
54
-------
The final concentration of nitric oxide is primarily
a result of the temperature and equivalence ratio history
of the products of combustion.
B. COMBUSTION CONCEPTS OF STRATIFIED CHARGE ENGINES
In many stratified charge engines a fuel injection
system is used to form a rich combustion region. The lean
region is initially composed of residual product from the
previous cycle plus the fresh charge of air or air with
EGR. The difference between the various types of these
stratified charge engines is in the processes used to com-
plete the combustion when the rich products are mixed with
the lean region. Conceptually these engines can be divided
into two categories, those which rapidly mix the rich pro-
duct of combustion with the lean region and those which
limit the mixing of the rich products with the lean region.
The difference between simplified examples of the
rapid mixing and limited mixing stratified charge engines
is illustrated in Figure 16. The generalized combustion
schematics shown in Figure 16 assume that the combustion
chamber can be divided into a few regions of uniform
properties. The percentage of the mass in the various
regions is shown as a function of the percentage of the
time of heat release.
Figure l6a represents rapid mixing of the products of
rich combustion (P-, ) . As soon as a small element is
55
-------
100
co
cd
o
CO
co
Rich Region
Rl
Lean Region
1(
)0
Time of Heat Release - percent
Figure l6a. Simplified Rapid Mixing Combustion Process
(Instantaneous Mixing. Chapter II)(R|P-A)
100
i
4)
a?
cd
£
o
«0
to
Time of Heat Release - percent
100
Figure l6b. Simplified Limited Mixing Combustion Process
(Delayed Mixing, Chapter II)(L;A-P)
100
rt
o
CO
to
I
Time of Heat Release - percent
Figure l6c, Simplified Limited Mixing Combustion Process
(LiP-A)
Figure 16 Simplified Combustion Process Schematics
56
-------
burnt it is thoroughly mixed throughout the lean product
mixture to give at any instant Ap + SAP, . Initially the
product mixture consists of residual products plus air
or air with EGR (Ap) . This process is symbolized "by
R;P-A for the rapid mixing of the rich products into the
lean product mixture (or lean region). The combustion
processes shown in Figure l6a are similar to the theoret-
ical instantaneous mixing concept discussed in Chapter II
and represent a simplification of the rapid mixing group
of stratified charge engines.
The combustion schematics presented in Figure l6b
and Figure l6c represent two simplified versions of the
limited mixing stratified charge engine group. Both ver-
sions start with a period of rich combustion with no mix-
ing. If the lean region (A2) is mixed into the rich pro-
ducts (P., + £AAp) the process is symbolized by L',A-P.
This combustion process is similar to the delayed mixing
concept discussed in Chapter II. The delayed mixing
concept represents one simplified version of the limited
mixing stratified charge engine group. If the rich pro-
ducts (P,) are mixed into the lean region (A2 + ZAP..) the
process is symbolized by L;P-A. This version of the
limited mixing stratified charge engines was not dis-
cussed in Chapter II and cannot be simulated by the com-
puter program.
57
-------
The combustion schematics are presented as an aid to
describing the more complicated combustion of real engines-
The Texaco Controlled-Combustion System (TCCS) Engine is
an operational version of a rapid mixing stratified charge
(R;P-A) engine as are some open chamber diesels. The Ford
Programmed Combustion (PROCO) Engine, the Newhall Engine
and some divided chamber diesels are operational examples
of limited mixing stratified charge engines (L;P-A).
It should be pointed out that the three-valve strati-
fied charge engine which normally uses carburetion to
distribute the fuel throughout the charge does not fit
into the categories of rapid mixing or limited mixing
stratified charge engines. Instead it is similar to a
homogeneous engine with the small prechamber functioning
as a method of charge ignition.
C. DEFINITIONS OF SOME TYPES OF INTERNAL COMBUSTION
ENGINES
Listed below are definitions of some of the various
types of internal combustion engines as used in this
paper.
Internal Combustion Enginet An internal combustion
engine is an engine in which the products of combustion
are used directly as the working fluid, Obert (1968).
Homogeneous Charge Engine; A homogeneous charge en-
gine is an internal combustion engine in which the design
58
-------
is intended to distribute all of the constitutents of the
charge uniformly throughout the combustion chamber.
Stratified Charge Engine; A stratified charge en-
gine is an internal combustion engine in which the design
is intended to distribute the constituents of the charge
nonuniformly throughout the combustion chamber. For ex-
ample, the fuel can be arranged into fuel-lean and fuel-
rich regions; the fuel-lean region can be air.
Diesel Engine; A diesel engine is a stratified
charge engine which uses the high temperature of the com-
pressed air to ignite the injected fuel.
Open Chamber Engine: An open chamber engine is an
internal combustion engine which has a single combustion
chamber.
Divided Chamber Engine; A divided chamber engine is
an internal combustion engine in which the clearance
volume is divided into two (or more) connected chambers.
The chamber in which combustion is initiated is called
the precombustion chamber or prechamber; the other chamber
in which expansion occurs is called the main chamber.
Three-Valve Stratified Charge Engine; A three-
valve stratified charge engine is a divided chamber spark
ignition stratified charge engine which uses a third valve
to admit a fuel-rich mixture near the spark plug in the
small prechamber and uses the normal inlet valve to admit
a fuel-lean mixture into the larger main chamber.
59
-------
Rapid Mixing Stratified Charge Engine Group: The
rapid mixing stratified charge engines are a group of
stratified charge engines in which all of the fuel passes
through the rich region and the products formed during
the combustion initiated in the rich region are rapidly
mixed with the lean region.
Instantaneous Mixing Stratified Charge Engine (R;P-A)
An instantaneous mixing stratified charge engine is a
theoretical version of the rapid mixing stratified charge
engine group, in which the products formed during rich-
region combustion are instantaneously mixed throughout
the lean region.
Limited Mixing Stratified Charge Engine Group; The
limited mixing stratified charge engines are a group of
stratified charge engines in which combustion is started
in the fuel-rich region, then by some means the mixing
is temporarily limited between the rich products and the
lean region. The lean region can be mixed into the rich
products, the rich products can be mixed into the lean
region or the rich products and the lean region can be
mixed into each other.
Delayed Mixing Stratified Charge Engine (L-.A-P) ; A
delayed mixing stratified charge engine is a theoretical
version of a limited mixing stratified charge engine
group. Combustion is initiated and completed in the rich
60
-------
region and then the lean region is mixed into the rich pro-
ducts .
D. HOMOGENEOUS ENGINES
Most automobiles and many light and medium duty trucks
are powered by homogeneous charge spark ignition engines
which have open combustion chambers. Considerable re-
search has centered on reducing NO emissions from these
Jt
engines. Although many factors influence N0x emissions
fuel-air ratio is the most significant. Huls (1966) mea-
sured the emission from a homogeneous charge engine. A
typical curve from his thesis is reproduced in Figure 17
and shows the variation of emissions with air-fuel ratio.
The maximum NO emissions occur at a slightly lean fuel-
X.
air mixture. At these conditions the two important
factors in nitric oxide formation, high temperature and
availability of oxygen, are in optimum proportions to
produce maximum NO emissions. When richer fuel-air
A.
mixtures are used the lack of oxygen limits nitric oxide
formation. When leaner fuel-air mixtures are used the
lower temperatures limit nitric oxide formation.
The NO emissions of homogeneous engines can be re-
jC
duced by operating either very rich or very lean. Rich
operation results in higher levels of hydrocarbons and
carbon monoxide plus lower efficiencies. For these rea-
sons very rich operation is not desirable. A lean burning
61
-------
16
12
B 10
-------
homogeneous engine has the advantage of lower hydrocarbons,
lower carbon monoxide and higher efficiency but has the
disadvantage of lower specific power, lower flame speed
and difficulty with ignition. When attempts are made to
operate even leaner a point is reached called the lean
limit where the engine will start to misfire.
Recent developments in spark timing control and car-
buretion described by John (1975) and by Adams (19?6)
have allowed homogeneous engines to operate at lower fuel-
air ratios. The "lean-burning engines" have lower ni-
trogen oxide emissions, lower carbon monoxide emissions,
lower hydrocarbon emissions (except at very lean opera-
tion) and higher efficiency due to more complete combus-
tion. The disadvantage is their lower specific power.
E. THREE-VALVE STRATIFIED CHARGE ENGINES
The three-valve stratified charge engine is the most
investigated form of the divided chamber spark ignition
stratified charge engine. It has a long history with
the initial patent being issued to H. R. Ricardo in 1918.
A thorough presentation of evolvement, analysis and pro-
gression of three-valve stratified charge engines is pre-
sented by Turkish (19?4).
A fuel-rich charge is drawn into a small prechamber
through the third valve and a lean charge is drawn into
the main combustion chamber through the normal inlet valve.
63
-------
Combustion is initiated by the spark plug located in the
prechamber. As pressure builds up in the prechamber it
forces high temperature rich products of combustion out
through an orifice into the main combustion chamber. The
hot products initiate combustion of the lean mixture in
the main chamber. Nitrogen oxide produced in the pre-
chamber is low because of its rich charge and its small
percentage of the total charge. The low temperature
associated with lean combustion of the main chamber charge
limits the main chambers nitrogen oxide contribution.
The three-valve engine could be regarded as a ver-
sion of a homogeneous engine since most of the charge is
in the main chamber and the charge in the main chamber
is roughly homogeneous.
The Honda CVCC engine, a three-valve stratified
charge engine, clearly demonstrated the potential for
the low NO emissions by meeting the original 1975 United
•A.
States and 1975 Japanese emissions standards , Tasuku
Date et al. (19?^). Honda's approach has been to opti-
mize experimentally and analytically the prechamber
volume, the orifice size and the fuel distribution as is
described by Yasuo Shiauo et al. (197^)-
The essential aspect of performance of all three-
valve stratified charge engines, that accounts for the
low NO emissions, is the ability to operate at very
A.
64
-------
lean overall fuel-air mixtures in the main chamber. When
a three-valve engine is compared to a similar convention-
al homogeneous engine as was done by Tasuku Date et al.
(197*0 of Honda, by Davis et al. (197*0 of General Motors,
by Purins (197*0 of Ford and by Yasuo Sakai (1974) of
Nissan the results are essentially the same. Davis states,
"At lean overall mixtures, both the Jet Ignition Strati-
fied Charge Engine and the conventional spark ignition
engines have similar HC, CO, and NO emission character-
J\.
istics when overall air-fuel ratio is changed." Tasuku
Date, Yasuo Sakai and Purins presented curves which showed
a direct comparison between their type of three-valve en-
gines and a similar homogeneous engine. These results of
Tusuku Date and Yasuo Sakai are reproduced in Figures 18
and 19- The lowest nitrogen oxide emissions correspond
to leanest operation not normally obtainable by conven-
tional homogeneous engines.
From Figures 18 and 19 it is apparent that the peak
nitrogen oxide emissions occur at approximately the same
overall air-fuel ratio for both the homogeneous and three-
valve engines. The nitrogen oxide peaks coincide because
most of the charge is in the main chamber and is somewhat
homogeneous. The small amount of charge in the prechamber
mixes rapidly with the main chamber charge and represents
a convenient method of ignition .
65
-------
IDLING
••*•.
70 r-
( ISFC « 410 g/PS,-h )
SO mll»/h
(ISFC«220
13 U IS 16 17 18
AIR-FUEL RATIO ( A/F)
13 U IS 16 17 18 19 20 21
AIR -FUEL RATIO (A/F)
Figure 18
Comparison of Exhaust Emission of GVCC
Engine with Conventional Engine
From Tasuku Date et al. (197^)
-------
e
3 0
20
O 10
\\
\\
y \
\^^
1200 rpm ^30%
I.S.F. C. = 250 s/jPSh
____-_^ •**
L
I
6
s
3
2
1
-
V
\
\
-
-
1
\\
X
- — -
JT
o
2
i
:
1
-
/
_L
A
f
/
/
i
TORCH IGNITED ENGINE
CONVENTIONAL ENGINE
\
\
\
v
Xs^-
•w
1 i
13 u 15 16 17 18 19 20
OVERALL AIR-FUEL RATIO
Figure 19
Comparisons of Exhaust
Emissions at Constant
Indicated Specific Fuel
Consumption
From Yasuo Sakai (1971*)
67
-------
F. RAPID MIXING STRATIFIED CHARGE ENGINES
1. Instantaneous Mixing Stratified Charge Engine
Concept
The instantaneous mixing stratified charge engine con-
cept described in Chapter II is a simplified theoretical
version of the rapid mixing stratified charge engine group.
It has the most nitric oxide emissions near stoichiometric
fuel-air ratios as seen in Figure 12. The shape of the
nitric oxide curve is similar to the homogeneous curve.
The combustion process is similar to that described in
Figure l6a, R;P-A.
The lean region is initially air or air with EGR (A2)
and it is at compression temperature. As combustion occurs
the rich products which are formed are instantaneously
mixed throughout the product region. The result of com-
bustion is that the product mixture fuel-ratio and tem-
perature increase toward those of the corresponding homo-
geneous case. The nitric oxide formed by the instantaneous
mixing case will, as usual, depend on the time available, the
product mixture temperature history and the product mix-
ture fuel-air ratio history.
When the final mixture is stoichiometric the in-
stantaneous mixing engine has nearly the same nitric oxide
emissions as the homogeneous engine. Under these condi-
tions some of the temperature history is sufficiently
68
-------
high to increase the nitric oxide formation rate and to
allow the instantaneous mixing engine to form about the
same amount of nitric oxide as the homogeneous engine.
This process is assisted by the fact that the equivalence
ratio of the product mixture passes through the rapid
nitric oxide forming region, which is slightly fuel-lean.
For the other conditions shown in Figure 12 the in-
stantaneous mixing engine produces less nitric oxide than
the homogeneous engine. Under these conditions the tem-
perature and the equivalence ratio of the instantaneous
engine is always less than that of the homogeneous engine
as is the resulting nitric oxide formation rate. The
lower temperatures and leaner product mixture does not
allow the instantaneous engine's nitric oxide concentra-
tion to reach the level of the homogeneous engine before
the cooling associated with expansion stops the nitric
oxide kinetics.
2. Texaco Controlled-Combustion System (TCCS)
Texaco Inc. has developed an open combustion chamber
stratified charge engine which uses the Texaco Controlled-
Combustion System (TCCS). In this engine air is caused
to swirl in the combustion chamber. Fuel is injected in-
to the swirling air and immediately ignited. A flame
front is established that burns the fuel as fast as it
is injected. The rich products of combustion are rapidly
69
-------
cooled "by the excess air. The NO emissions as reported by
X
Mitchell et al. (1972) are presented in Figure 20. His
curves show the variation of NO emissions with changes
X
in the engine IMEP. Although fuel-air ratios are not
plotted directly the lower IMEP corresponds to leaner
overall fuel-air ratios. Mitchell states, "Fuel combus-
tion under rich mixture conditions limits formation of
oxides of nitrogen in the flame zone. A quick quench of
the burned gases by the cold excess air precludes further
NO formation in these post-flame gases." This descrip-
Jl.
tion along with the generally increasing nitrogen oxide
concentration with IMEP implies that the TCCS engine
could be considered a rapid mixing stratified charge en-
gine. The rapid mixing stratified charge engine would be
expected to have increasing nitric oxide emissions with
increasing IMEP as does its theoretical counterpart the
instantaneous mixing engine. See Figure 9 for the cal-
culated results of the instantaneous engine.
G. LIMITED MIXING STRATIFIED CHARGE ENGINES
1. Delayed Mixing Stratified Charge Engine Concept
The delayed mixing concept of stratified charge en-
gines is a theoretical version of the limited mixing
stratified charge engine group which corresponds to the
combustion process described in Figure l6b, L;A-P.
70
-------
*,
o oc
Zi! 5
rife
, MM
O
0
i
30
K
^ 20
M
O
r10
o
* ft
1
p
^£
0
*— ~"
••»— '
^i •••
_^
•»<»*t«^
— BASE ENGINE
WITH MODIFICATIONS
WITH MODIFICATIONS
a CATALYST
y
w.
h^^MM^W
-—-
•—— ,
BEGF
, EGR
•'
IMEP-PSI
120
Figure 20
Single Cylinder Evaluations
HC plus NO Emissions
From Mitchell et al. (1972)
71
-------
All of the fuel is in the rich region which "burns first.
During the rich combustion little nitric oxide is formed
because of the lack of oxygen even though the temperature
of the rich products is quite high. When the rich com-
bustion is finished, the lean region (air plus EGR) is
added incrementally to the rich products. The incremen-
tal addition of the air continues the combustion process
and keeps the product mixture temperature high. If the
overall air-fuel ratio is lean the excess air will cause
a reduction in the temperature of the product mixture.
The amount of nitrogen oxide which is formed with an
overall stoichiometric mixture is less for the delayed
mixing engine than the corresponding homogeneous engine
burning a stoichiometric mixture as shown in Figure 12.
The difference is primarily caused by the equivalence ratio
of the delayed mixing engine being richer than that of the
homogeneous engine. The product temperature for both en-
gines would be similar. Expansion stops the nitric oxide
kinetics before the nitric oxide concentration of the de-
layed mixing engine reached that of the homogeneous engine.
When the overall mixture of a delayed mixing engine
is very lean the nitric oxide emissions from the de-
layed mixing engines are much larger than the instantaneous
engine using a correspondingly lean mixture. The delayed
mixing product mixture starts at a high temperature and
72
-------
is provided with abundance of oxygen which results in the
high nitric oxide concentration. The expansion cools the
product mixture and stops the nitric oxide kinetics before
the reverse reactions can reduce the concentration. In
contrast the homogeneous mixture is always at a low tem-
perature which limits the nitric oxide formation.
If the delay period were to "be extended to the end of
the active nitric oxide kinetic period the nitric oxide
emissions would correspond to that of a homogeneous en-
gine operating at the rich equivalence ratio. In general,
the delay period is not that long and some nitric oxide
is formed due to mixing of the lean region with the rich
products.
2. Ford Programmed Combustion (PROCO) Engine
Ford Motor Company has developed a type of open
chamber stratified charge engine which is called the
Ford Programmed Combustion (PROCO) Engine. Stratifica-
tion is established by causing the air in the combustion
chamber to have a high swirl and to spray the fuel into
the central region of the swirl. A spark plug located
near the fuel spray cone ignites the rich region. Lavoie
and Blumberg (1973) of Ford compared the nitric oxide
emissions from the PROCO engine to that of a similar
homogeneous engine. Their results are shown in Figure
21. They found that near overall stoichiometric fuel-air
73
-------
I04 -
I05
NO
PPM
10 2
— 6 1500 RPM. 70 IMEP
Theory
i— Experiment
10 I I
0.6 0.7 0.8 0.9 I 0
Figure 21
Comparison of Preraixed and
Stratified NO Concentrations
As a Function of Equivalence
Ratio at Q% EGR. From Lavoie
and Blumberg (1973)
74
-------
mixtures,the PROCO engine had less nitric oxide emissions
"but that for very lean mixtures, the PROCO engine had
greater nitric oxide emissions than the homogeneous en-
gine. Blumberg (1973) of Ford has been able to predict
the same trends in nitric oxide emissions by use of a
computer model of the PROCO engine. The theoretical re-
sults are also shown in Figure 21. One essential feature
of the model is that the stratification of the charge
remains throughout combustion and expansion until the
charge temperature is sufficiently low to prevent changes
in the nitric oxide concentration. In the model the
richest element, in the center, burns first and combus-
tion progresses to the leanest element. When the over-
all equivalence ratio is made leaner each element like-
wise becomes leaner. Since the first elements to burn
adds the most to the nitric oxide concentration and the
first elements are made leaner (still rich but nearer
stoichiometric), they tend to increase the nitric oxide
concentration. The last elements to burn which are lean
and become leaner, tend to reduce the nitric oxide con-
centration. The net result is that the nitric oxide con-
centration is nearly independent of fuel-air ratio as
seen in Figure 21.
An interesting comparison can be made between the
delayed mixing engine with zero EGR shown in Figure 12
75
-------
and the PROCO engine also with zero EGR shown in Figure 21.
Both show very little change in nitric oxide concentra-
tion with equivalence ratio. The swirl stratification
model and the delayed mixing model "both have a delay in
the mixing of the rich product region. In the Ford swirl
stratification model, the delay extends until the nitric
oxide kinetics have been stopped by expansion. In the
delayed mixing model the delay lasts only part way through
the active nitric oxide kinetic period. In the actual
PROCO engine some mixing is likely to occur between the
various elements during the active nitric oxide kinetic .
•,
period and yet the measured nitric oxide concentration
is also nearly independent of equivalence ratio. The
delay before the rich products start to mix and the limit-
ed amount of mixing that does occur before expansion
cooling stops the nitric oxide kinetics are the main
reasons for the shape of the nitric oxide emissions curve.
3« Newhall Engine
A version of the divided chamber spark ignition
stratified charge engine has been built and tested by
Newhall and El-Messiri (1973). The fuel is injected
into a large prechamber and only air is brought into
the main chamber. The nitrogen oxide emissions shown
in Figure 22 have a maximum at about an overall equiva-
lence ratio of .?. If the rich products leaving the
76
-------
3000
1000
o
300
o
Tt
•H
c
d>
W)
o
100
30,
.6 .8 1.0
Overall Equivalence Ratio
1.2
Figure 22
Nitrogen Oxide Concentration versus
Overall Equivalence Ratio for Wide
Open Throttle and 6$% Prechamber
Volume, From Ingham(1976)
77
-------
prechamber are rapidly mixed with the air in the main
chamber as is done in the instantaneous model the nitric
oxide maximum would occur near stoichiometric, probably
not leaner than an overall equivalence ratio of .8. Sup-
pose all of the nitric oxide were to be formed in the
prechamber and the mixing were delayed until after the
active nitric oxide kinetic period. If the maximum
nitric oxide occurred at a prechamber equivalence ratio
of .85, the overall equivalence ratio would be about .55
(.85 x .65 = .5525). Apparently a limited amount of
mixing and nitric oxide formation does occur in the main
combustion chamber which accounts for the nitric oxide
maximum occurring between the instantaneous mixing con-
cept and the other extreme of all of the nitric oxide
being formed in the prechamber.
The Newhall Engine could be included in the L;P-A
group shown in Figure l6c. The rich products of com-
bustion probably remain in the combustion chamber for a
short period and when they emerge, they mix rapidly with
the lean mixture.
H. COMPRESSION IGNITION STRATIFIED CHARGE ENGINE
Although diesel engines are stratified charge engine
they are not usually grouped with spark ignition strati-
fied charge engines because of the large differences in
78
-------
the combustion processes. Certainly the diesel combustion
process is considerably more complex than many spark
ignition stratified charge engines. However, the chemi-
cal reactions and chemical kinetics of nitric oxide apply
equally as well to diesel engines as they do to spark
ignition stratified charge engines, as they do to any
form of combustion with air. For example, Tuteja (1972)
showed that the Zeldovich mechanism could be used to cal-
culate the nitric oxide concentration in diffusion flames
with reasonable accuracy. Many diesel engine models
based on the nitric oxide chemical reactions and chemi-
cal kinetics have satisfactorily predicted nitric oxide
emissions. Some examples are Gakir (1974), Shahed (1973)
Nightingale (1975) and Khan (1973) .
Most of the nitric oxide is formed in the regions of
hot products of combustion, some is formed in the combus-
tion zone and none is formed in the cool reactants.
Nightingale (1975) of Ricardo and Co. demonstrated this
point by taking samples inside of a diesel combustion
chamber at various times and locations during the combus-
tion process. His results,shown in Figure 23,are for a
point in the center of the fuel spray. As combustion
proceeds,the hydrocarbons drop to a low level and the
C02 increases to a higher level. The CO peak is due to
the richness of the region being sampled. After combustion
79
-------
800--20
600- •«
C
1
200-'5
0-LO
-10 TDC 10 20 30 40 50
400--10
Figure 23
Gaseous Composition in the
Center of the Fuel Spray As
a Function of Crank Angle
From Nightingale, D.R. (1975)
80
-------
is completed, the nitric oxide level continues to increase
until it peaks. This is evidence which shows that much
of the nitric oxide is formed in the products of combustion
as would be expected from the chemical kinetics.
Nightingale also presented results showing the nitric
oxide concentration versus equivalence ratio for all
samples taken later than 5° ATDC. At an angle of 5° ATDC
most of the combustion should be complete. His results
are given in Figure 24a. In general, the results indicate
that samples of products having an equivalence ratio near
.9 will have greater nitric oxide concentrations than
either richer or leaner samples. A similar curve was
obtained by Rhee (19?6) with samples in the lean region
and is presented in Figure 24b. The magnitude of the
nitrogen oxide emissions reported by Nightingale is m&ch
lower because of the engine modification necessary for
photographing the combustion process. The nitric oxide
maximum occurs at about the same equivalence ratio as
the nitric oxide maximum for homogeneous combustion.
The similarity in the results is because both types of
combustion produce most of the nitric oxide in the com-
bustion products by means of the same chemical reactions
and chemical kinetics, even though the combustion pro-
cesses used to form the products are very different.
A study of Pischinger (19?2) compared the nitrogen
oxide emissions from open chamber (direct injection)
81
-------
Figure
Experimental In-Cyl-
inder NO concentration
as a Function of Equiv-
alence Ratio, From
Nightingale (1975)
•2 -i -6 -8 1-0 1-2 14 1-6
Weak Rich
Equivalence ratio
Figure
Nitrogen Oxide Con-
centration Versus
Equivalence Ratio
For Samples Taken
From 15-20°ATDC,
From Rhee (1976)
6000
ft ^000
ft
z
2000
oo
0
c
0
Tt
o
.8
Equivalence Ratio
i.o
Figure 2*4*
Nitric Oxide Concentration Versus
Equivalence Ratio For Samples taken
From Within a Direct Injection
Diesel After Combustion
82
-------
diesels and divided chamber (indirect injection) diesels.
His results are given in Figure 25- The open chamber
diesel has increasing nitrogen oxide levels with increas-
ing BMEP while the divided chamber diesel has a maximum
nitrogen oxide level at about the center of its operating
range. Monaghan et al. (197^) and others have found
similar results in comparing the open chamber and divided
chamber diesels.
1. Open Chamber Diesel Engines
The open chamber (direct injection) diesel has a
nitrogen oxide emissions curve that is similar to the
characteristic shape of the rapid mixing stratified
charge engine group, as shown in Figure 25. The explana-
tion for the increasing nitrogen oxide concentration with
increasing overall fuel-air ratio depends on how the com-
bustion process is visualized. Shahed (1973) and
Nightingale (1975) present combustion models in which
the combustion is assumed to occur at stoichiometric
conditions and no mixing occurs between the products
formed at various times. With their models the increase
in the overall fuel-air ratio results in more products
at stoichiometric conditions and consequently more ni-
trogen oxide. In Khan (1973) combustion model, air is en-
trained in a vaporized fuel jet with fuel and air mixing
occurring within the jet. The rate of combustion is
S3
-------
000
600
I
5
600
OL
CL
x
o
400
200
RATED SPEED
BASIC TIMING
-DEGR, CA BTDC—
18'CA
0 20 40 60 80 XX)
BMEP PSI
Figure 25
NO Concentration, Naturally
Aspirated Direct-injection
and Prechamber Engines
From Pischinger, R. (1976)
84
-------
determined by the mixing process. In this model, the
changes in the equivalence ratio of the products will in-
fluence the nitric oxide concentration. Khan's model
predicts a nitric oxide peak at high overall fuel-air
ratios whereas Shahed's and Nightingale's models do not.
See Figure 26. The existence of a nitrogen oxide maximum
at high fuel-air ratios is shown experimentally "by Night-
ingale (1975), in Figure 26b. The smoke limit usually
prevents diesels from operating at high overall fuel-air
ratios where the nitrogen oxide maximum occurs.
If Khan's model is used to visualize the combustion
processes, then it would be possible to include the open
chamber diesel in the rapid mixing stratified charge
engine group. The nitrogen oxide emission curve from
the open chamber diesel also indicates that it should be
in the rapid mixing group.
2. Divided Chamber Diesel Engines
In the divided chamber diesel engines, the fuel is
injected into the prechamber where combustion begins.
The pressure in the prechamber increases and forces the
product mixture out into the main chamber. Typically
the design of the combustion chambers promote swirl in
the main chamber. As is shown in Figure 25, the nitrogen
oxide concentration has a maximum which does not occur
for the open chamber diesel normal operating range. The
nitrogen oxide maximum can be explained by postulating
85
-------
Figure 26a
Calculated (dashed
line) and Experimental
(solid line) Exhaust
NO versus Engine
Fueling
From Khan et al. (1973)3
2000
1500
1000
500
%
X
s
s
X
/
1
/
/
«
*
X
\
0 AO 50 60 70
FUELING mur/Stroke
Figure 26b
The Effect of Varying
Engine Overall Air/Fuel
Ratio (Engine Load) on
Exhaust NO Concentration
From Nightingale, D. R.
(1975)
' ^ Calculated results C|2 H
— -— Calculated results Gas oil
30 tO 50
Overall air/luel ratio
Figure 26
Comparisons Between Nitric Oxide Emissions
from Models and Measured Results
86
-------
that the product mixture entering the main chamber has
limited mixing with the main chamber air.
The mixture in the prechamber will always be richer
than the overall mixture because only a fraction of the
air is in the prechamber. At high powers (BMEP) the
prechamber contains a rich fuel-air mixture. The rich-
ness of the mixture would account for low nitrogen oxide
concentrations in the prechamber. Other factors could
limit the nitrogen oxide formation in the main chamber.
For example, it is possible that the secondary swirl
caused by the products entering the main chamber from
the prechamber could result in fluid rotation with the
high temperature (low density) products in the center
and the low temperature (high density) air on the sides.
This stratification could account for some of the delay
in the mixing process. Possibly the products mixture
enters the main chamber too late in the nitric oxide's
active kinetic period to result in more nitrogen oxide.
The increase in heat transfer due to the swirl could re-
duce the temperatures and consequently limit nitrogen
oxide concentrations.
Because of the complexity of the combustion and
mixing processes,it is difficult to attribute the shape
of the nitrogen oxide curve to any one factor. However,
it seems reasonable to group the divided chamber engine
87
-------
with the other limited mixing stratified charge engine "be-
cause it has the same general shape of nitrogen oxide
emissions, the prechamber limits mixing,and the swirl
in the main chamber also limits mixing.
88
-------
CHAPTER IV
THE TEST FACILITY AND INSTRUMENTATION
A. INTRODUCTION
In order to study the delayed mixing concept of strati-
fied charge engines, it was first necessary to decide wheth-
er the test facility should simulate a practical engine or
simply test the concept. There are many problems associa-
ted with designing a practical engine such as keeping the
rich charge and the air separated "before mixing, finding
an effective way of mixing the air with the rich products,
and determining the most appropriate values for the main
variables. Because of the problems associated with the
design of a practical engine it was decided not to attempt
the simulation of a practical delayed mixing engine.
The alternative to designing a practical delayed
mixing engine is to experimentally simulate the combus-
tion processes and study the main parameters. This is
the approach taken here. The test facility consists of
a single cylinder CFR engine fitted with an adjustable
carburetor and an adjustable air injection system. The
test facility simulates delayed mixing combustion by ad-
mitting a rich charge to the combustion chamber, burning
the charge and injecting the air when combustion is
essentially complete.
89
-------
A schematic diagram of the test facility is presented
in Figure 27-
B. ENGINE
An ASTM-CFR (Fuel Research Engine, 4~1957) single cy-
linder engine with variable compression ratio was used
for the experimental program. Three radial access ports
plus the spark plug hole were in the head assembly. One
port directly across the cylinder from the spark plug was
used for the air injection nozzle. The two other ports
are located at ^5° either side of the spark plug. One of
these ports houses the pressure transducer. Prior to
testing,the engine was disassembled, cleaned and rebuilt.
When the compression ratio was adjusted, the influence
of the nozzle, inlet and exhaust valve penetration, and
spark plug on the clearance volume was included in the
calculations. The procedure as deduced from the ASTM
Manual (1971)> is to place a small block,.625 inches
(15.9 mm) high, in the combustion chamber positioned
between the valves while the piston is at top dead center.
Next the head is cranked down until it stops. A micrometer
mounted between the variable head assembly and the stationary
base of the engine is adjusted to .362 inches (9.19 mm). The
compression ratio can be calculated by means of the follow-
ing equation.
90
-------
T13)
batteries
T ignition
"=~ switch
stomexer
MNrl
i/^s
coolant
condenser
coolin
*\\ system
tl
g
At
cooling
water
breaker
points
ffi
induction
coil
Capacitor amplifier
carburetor/^
. . I i *•
ventun
main j
needle
float
bowls
exhaust
mixing
fu«l
scale
fuel
tanks
flow control
dynamometer
restriction valve
higR press
air comp.
cooling water
high pressure
air storage tanks
ng water
FACILITY SCHEMATIG DIAGRAM
Figure 2?
-------
r^m-™~«r,.-,T,vm PO+T~ - ^.5*+ 0.3 + micrometer reading
Compression Ratio - 0.3 + micrometer reading
In this equation the 0.3 inches corresponds to a phy-
sical distance of .263 (.625 - -362 = .263) and 4.5 cor-
responds to the engine stroke. The value of .3 is used
in place of .263 to compensate for a slightly larger cy-
linder diameter in the clearance volume. Corrections for
the volume added by the spark plug is equivalent to an
additional .004 inches, the valve penetration reduces the
volume equivalent to .0253 inches and the air injection
nozzle increases the volume in accordance with the size
of nozzle used. The correction for the 5/16 inch nozzle
is equivalent to an increase of .0156 inches. The com-
pression ratio equation for this nozzle is as follows.
r*™™^^;™ r.o+^ - 4.5 + .294 + micrometer reading
Compression ratio .29^ + micrometer reading
The compression ratio equation for the .0935 inch nozzle
is as follows.
r- «~« -;™ v.o+;~ - 4.5 + .281 + micrometer reading
Compression ratio .281 + micrometer reading
Connected to one end of the engine crank shaft is a
dynamometer and connected to the other end is a sprocket
used to drive the cam-operated air injection valve. The
angular location of the sprocket teeth are sensed by a
*A11 the numbers on this page are in inches, to convert to
millimeter multiply by 25.4.
92
-------
magnetic pick-up for use as timing marks on the oscillo-
scope display. Another magnetic pick-up is used to indi-
cate top dead center; a third triggers the sweep of the
oscilloscope; and a fourth magnetic pick-up senses engine
speed for the emergency shut-down system.
C. DYNAMOMETER
A 50 horsepower (37285 watt) General Electric dynamo-
meter serial number 2127382 is used in conjunction with the
control system. The dynamometer system keeps the engine
speed constant at the desired level for either motored or
powered operation. A 24.75 inch (6278.7 mm) diameter scale
calibrated in .2 Ib- (.89N) increments is used to measure
a force which can be converted to engine power by the
relationship below.
Force Read on Scale x RPM
Horsepower = ^^-
When the engine is producing power, the scale force is an
indicator of brake horsepower based on this conversion.
When the engine is being motored with the ignition turned
off the scale force indicates the approximate frictional
horsepower. The sum of the powered and motored horsepower
is approximately the indicated horsepower.
Also connected to the dynamometer is the sensor for a
tachometer. The engine RPM is displayed on one of the con-
trol panels.
93
-------
D. FUEL SYSTEM
The fuel system is shown as part of the test facility
schematic presented in Figure 27. Fuel is stored in out-
board engine fuel tanks. The tanks are connected to an
automatic fuel weighing system. In the fuel weighing sys-
tem a time is measured which corresponds to a specified
mass of fuel; in our tests twenty grams. A fuel pump
draws the fuel from the scale and delivers it to the float
bowls. The float bowls maintain a constant level by means
of the float bowl valve. Two float bowls in series are
used to increase the stability of the fuel flow rate to
the engine. Fuel leaves through the bottom of the second
float bowl, passes through the main needle valve which is
a 1/4 inch micrometer valve and enters the venturi of the
CFR carburetor where fuel mixes with the air.
A number of calibration runs were performed on the
fuel system by extracting fuel from between the needle
valve and the carburetor. A small valve was installed in
that line for the purpose of controlling fuel flow rate.
During the test period the fuel was collected in a beaker
while the fuel scale was activated as often as possible.
A comparison is shown below of the time interval corres-
ponding to the flow of twenty grams of fuel as obtained
from the fuel scale and calculated from the accumulated
fuel.
94
-------
Time Interval -seconds
Fuel Scale
Time Interval
Average Std. Dev. Accumulated Fuel Ratio
6^.02 .53 65-18 .9822
65.51 1.26 65.93 .9936
66.^2 .634 66.74 .9952
66.15 .721 66.34 .9971
Although the scale is seen to "be quite accurate one
known source of error can be calculated is due to the
buoyancy associated with the tubing used to draw the fuel
from the fuel beaker located on the scale. Since the
fuel level in the beaker is lowered during the weighing
process the buoyancy force exerted on the tubing and the
beaker is reduced. The reduction in the buoyancy appears
to the scale as removal of fuel. Thus the scale would
indicate more fuel removed for the beaker than was actual-
ly removed or the time interval indicated by the fuel
scale for the removal of twenty gram of fuel is shorter
than the correct value. For our system the error is
equivalent of .0922 grams. The ratio of the time inter-
val indicated by the fuel scale to the time interval cor-
rected for buoyancy is .9954. This is approximately the
ratio of the fuel scale time interval to that deduced
from the accumulated fuel. If the fuel scale were to be
corrected for buoyancy of the fuel removing tube the
95
-------
accuracy would be greater. However, it is sufficiently
accurate without the buoyancy correction.
E. CARBURETED AIR SYSTEM
The air that flows through the carburetor, called
carbureted air, is drawn into a surge tank through a
laminar flow element as indicated in Figure 2?. The surge
tank reduces the air flow fluctuations and establishes a
nearly constant pressure differential across the laminar
flow element. An inclined manometer is used to measure
the pressure differential. The laminar flow element was
calibrated by motoring the engine and measuring the air
entering the laminar flow element with a large bellows
meter. The flow of air through the laminar flow element
can be described by the linear relationship shown below.
Volumetric Flow Rate (CFM)* = 5.964. x Ap(inches of HpO)
A choke is located in the carbureted air line to
restrict the air flow for part load operation. An emer-
gency air shut off valve is also located in the air line
which closes in the event of an over speed accident.
The air enters the venturi section of a CFR car-
buretor and mixes with the fuel. Between the carburetor
and the engine is a section of piping which is heated
with a strip heater. A mercury manometer is used to
measure the inlet fuel-air mixture pressure and various
*See Appendix I for conversion to SI units.
96
-------
temperatures are measured "by thermocouples connected to
the multipoint recorder.
F. INJECTION AIR SYSTEM
The high pressures associated with combustion require
that the injection air be compressed to high pressures. A
high pressure air compressor is used to fill high pres-
sure air storage tanks as shown in Figure 27. During
engine operation the air is drawn from the storage tanks
and passes through a pressure regulator which controls
the entrance pressure to the flow restriction. The flow
restriction is used to measure the air flow rate. Fol-
lowing the flow restriction is a needle valve which is
used to control the flow of air. A surge tank near the
air injection valve damps out flow fluctuation caused by
the periodic opening of the air injection valve. After
leaving the air injection valve the air flows through a
nozzle which controls how the air enters the combustion
chamber. More details about the main components of the
air injection system are given in the following sections.
1. High Pressure Air Compressor System
A four stage Joy Model 15-H air compressor capable
of reaching pressures of 3500 psi (24130 KP ) is used to
3.
supply the high pressure air. Five air storage tanks rated
at 2300 psi (15860 KP 1 are used to store the air. The
3.
'a'
97
-------
tanks provide sufficient capacity for normal operation of a
few hours. Typically the air pressure is reduced to about
1200 psi before entering the engine test cell. The pressure
regulator used is pneumatically controlled.
2. High Pressure Air Flow Measurement System
A major problem associated with the air injection
system is the method of measuring the air flow rate. At the
high pressures used, a laminar flow element would have to be
very small in diameter and very long to provide sufficient
pressure drop for accurate measurement. There is also the
question of high pressure requirements of the pressure
differential measuring device.
The approach used was to include in the line a sub-
stantial flow restriction. The flow restriction consists
of a .054 inch (1.37 mm) diameter tube with three .017 inch
(0.43 mm) diameter wires in it. Within the four foot flow
restriction the flow would be turbulent. For fully developed
turbulent flow through a nearly constant temperature tube a
computer program was developed which correlated the flow
through the flow restriction to the inlet pressure, inlet
temperature and the outlet pressure. Numerous calibration
runs were made under various flow conditions. The measured
flow was compared to the computer program predicted flow
rate. A very slight adjustment was made in a constant in
the friction factor equations in order to establish
98
-------
agreement between the measured and calculated results.
The deviation from the calculated values is usually less
than 2 psi (13.97 KP&). The details of the computer pro-
gram and the calibration of the flow restrictions are
presented in Appendix E.
3. Air Injection Valve
The air injection valve had to be specially designed
because of its special requirements for large flow rates
and short open periods. The essential features of the
valve are that it opens and closes in *4-0 crank angle
degrees and it can let in a mass of air about equal to
the carbureted charge at wide open throttle. A sketch of
the valve is shown in Appendix F. Included in Appendix G
is a computer program which describes the flow of air
through the valve and the dynamic forces resulting from
the cam engaging the cam follower.
G. PRESSURE TRANSDUCER SYSTEM
An AVL type KQD 250C pressure transducer is mounted
in one of the access ports in the head assembly. This
type of pressure transducer is water cooled and has
been found to be quite accurate by Lancaster (1975)•
The pressure transducer was prepared, as recommended by
Lancaster, with a General Electric RTV 560 (synthetic
rubber) coating to reduce the heat transfer. The
99
-------
pressure transducer signal goes to a Kistler charge ampli-
fier model 566, SlNS^-8. The signal from the charge ampli-
fier is displayed on the oscilloscope.
H. OSCILLOSCOPE
A Tecktronix Type 35A- oscilloscope is used to display
the cylinder pressure signal and the timing marks. The
timing marks indicate ten crank angle degree increments
with the top dead center pulse being greater in amplitude.
The oscilloscope has a single "beam. In order to show
both the pressure signal and the timing marks it is neces-
sary to use a special plug-in unit which will either al-
ternately display the two signals or simultaneously dis-
play both signals by alternately displaying small incre-
ment of each (chopped). The oscilloscope was checked-
out and calibrated before it was put in service.
The engine speed can be obtained from the oscillo-
scope display of the timing marks. From the oscilloscope
it is possible to determine the elapsed time between top
dead center pulses. This elapsed time can be converted
into engine RPM. The engine speed was determined by
this method.
I. MULTIPOINT TEMPERATURE RECORDER
A Leeds and Northrup multipoint temperature recorder
serial number E79-52108-1-1 is used to record the various
temperatures. The thermocouples are chrome1-alurael, Type
100
-------
K. The multipoint recorder has the capability of recording
24 different temperatures. Twelve of the temperatures can
be in the range of 20°F to 250°F (266.3°K to 394°K) and the
other twelve can be in the range of 200°F to 2000°F (366.3°K
to ISGe^K). The numbers indicated on the test facility
schematic diagram (Figure 27) correspond to the numbers
printed by the multipoint recorder.
J. EMISSIONS CART SYSTEM
The emissions cart system is composed of two parts
the cart containing the instrumentation and the cart con-
taining the calibration and purge gas cylinders. The in-
strumentation cart is set-up to measure nitrogen oxides
(NO and NO ), carbon monoxide and hydrocarbons. There
J\.
are valves on the instrumentation cart that are used to
select a purge gas (nitrogen), room air, the appropriate
calibration gas or a sample from the engine exhaust gases.
The sample is run through an ice bath and filter before
it goes to the instrumentation. Pumps on the instrumen-
tation cart cause the sample to flow from the exhaust pipe
and through the instrumentation. Details of each of the
instruments is given below.
1. Nitrogen Oxide Measurement
The nitrogen oxide measurement is made by a chemi-
luminescent NO-NO gas analyzer model 10A manufactured
J\.
101
-------
by Thermo Electron Corporation. The instrument has seven
ranges for 0-10 to 0-1000 PPM, its sensitivity is .1 PPM
and can measure either NO or NO .
Ji.
The chemiluminescent nitrogen oxide analyzer uses the
photons generated when nitric oxide and ozone are reacted
to count the number of nitric oxide molecules. A photo-
multiplier tube is optically filtered so that it will
only respond to the nitric oxide photons. The rate of
flow of sample through the reaction cell is maintained
constant so that the output signal will be proportional
to nitric oxide. When the nitrogen oxides (NO ) are to
j\.
be measured the sample is passed through a thermoconverter
which converts the various oxide of nitrogen to nitric
oxide before the sample goes to the reaction chamber.
2. Carbon Monoxide Measurement
The carbon monoxide is measured by a nondispersive
infrared detector (Infrared Analyzer - IRISA) manufactured
by Beckman. The instrument consist of two parts; a unit
containing the sensing elements and a unit containing the
electronics and meter. A nondispersive infrared detector
detects carbon monoxide by comparing the amount of in-
frared radiation absorbed by the sample and a reference
gas. Infrared radiation is passed through a reference
cell and also through a cell containing the sample gas.
After the radiation passes through the cells it goes to
102
-------
two cells containing carbon monoxide. If the sample con-
tains carbon monoxide it will absorb some of the energy
of the infrared radiation that is at the carbon monoxide
absorption frequencies. Thus the infrared radiation
leaving the sample cell will be deficient in energy
associated with the carbon monoxide absorption bands.
Since the detection cells are filled with carbon monoxide
they too would absorb energy in the absorption band of
carbon monoxide. The detection cell associated with the
sample cell will receive less energy in the absorption
bands when more carbon monoxide is in the sample cell.
The detection cell associated with the reference cell
will receive the same amount of energy in the carbon
monoxide absorption bands. The differences in the energy
absorbed by the two detection cells results in a pressure
differential which is measured by the deflection of a
diaphragm between the two cells.
3. Hydrocarbon Measurement
The hydrocarbon measurement is made by a Beckman
Hydrocarbon Analyzer, model 109A. The instrument operates
on the principle of flame-ionization. A hydrogen flame
is established that burns with a negligible number of
ions. The sample gases are introduced into the re-
actants of the flame. If the sample gases have hydro-
carbons present they will form ions when they burn. The
103
-------
ions are collected and result in a signal proportional to
the concentration of hydrocarbons in the sample. It is
necessary to maintain constant flow rates to assure the
accuracy of the measurements.
104
-------
CHAPTER V
THE EXPERIMENTAL PROGRAM AND TEST RESULTS
A. INTRODUCTION
The objective of the experimental program is to de-
termine if the advantages of the delayed mixing concept
shown "by the theoretical study presented in Chapter II
can be realized in an experimental engine. The experi-
mental program is also intended to show the influence of
major input variables on the emissions, efficiency and
power.
A CFR engine was modified to include an air injec-
tion valve which can inject air at various times in the
cycle or not at all. Typically a rich charge would be
drawn into the combustion chamber. The charge is ignited
before top dead center. Air is then injected into the
combustion chamber after the rich mixture has finished
burning. The combination of the rich mixture combustion
followed by air injection is intended to represent the
delayed mixing concept as symbolized by L;A-P. The delay
between the combustion of the rich mixture and mixing is
controlled by the timing of the spark and the air injec-
tion. Since it takes a finite amount of time for the
air to mix and react with the rich products, this repre-
sents a difference from the theoretical concept which
assumes that no time is required for the mixing.
105
-------
The engine input variables were examined to determine
the best way to control and measure them, and to decide
which factors to vary and which to hold constant.
In addition to the measurement of the nitrogen oxide
emissions, the carbon monoxide and hydrocarbon emissions
were measured. The engine power and efficiency were
also determined. A section is devoted to a discussion
of the theoretical trends expected in these results.
The experimental program is divided into two sec-
tions. The first set of tests were performed with a
large diameter air injection nozzle and the second set
of tests were performed with a small diameter air in-
jection nozzle. Within each set of data the major
variables studied were the overall air-fuel ratio, the
timing of the start of air injection and, to a lesser
extent, the carbureted air-fuel ratio. As the experi-
mental program proceeded, problems were recognized and
fixed. For this reason some of the data is more accurate
than others.
In general, the data obtained from each experiment
is punched on computer cards and then processed by com-
puter programs to obtain the results in a more convenient
form. The computer programs used to reduce the data are
given in Appendix H.
106
-------
B. ENGINE INPUT VARIABLES
Listed below are the important engine input variables
which influence the performance of the delayed mixing
stratified charge test engine. These input variables
fall into two categories, those that can be changed while
the engine is in operation and those that require a
mechanical modification to the test facility. The first
group will be called operational variables and the second
group will be called design variables.
1. Operational Variables
1) Fuel Flow Rate: The fuel flow rate is controlled by
the main needle valve of the carburetor and is measured
by the fuel scale.
2) Carbureted Air Flow Rate: The carbureted air flow
rate is controlled by the choke and engine speed. A
laminar flow element and an inclined manometer are used
to measure the flow rate.
3) Injected Air Flow Rate: The injected air flow rate
is controlled with the flow control valve and determined
by the upstream temperature, the upstream pressure and
downstream pressure of the flow restriction.
4-) Start of Air Injection: The start of air injection
is adjusted by selecting the desired relative position
of the drive sprocket mounted on the crank shaft and the
driven sprocket on the cam valve before the drive chain
107
-------
is installed. The position of the crank shaft is measured
on the flywheel and the position of the air injection
valve is measured by a protractor mounted on the driven
sprocket of the air injection valve.
5) Spark Timing: The spark timing is controlled "by ro-
tation of the housing containing the points and is mea-
sured by either an induction pick-up displayed on the
oscilloscope or by a timing light.
6) Compression Ratio: The compression ratio is controlled
by moving the engine head assembly up or down with
a crank. A micrometer measures the position of the en-
gine head, from that measurement the compression ratio
can be determined,
7) Engine Speed: The engine speed is controlled by the
dynamometer system and is measured by both a tachometer
and the time between top dead center pulses on the
oscilloscope.
8) Engine Inlet Pipe Pressure: The inlet pipe pressure
is controlled by the choke and is measured by a mano-
meter.
9) Engine Inlet Pipe Temperature: The inlet pipe is
heated with a strip heater and its temperature is mea-
sured by a thermocouple.
10) Injection Air Temperature: The injection air tem-
perature can be controlled by a heat exchanger and is
measured by a thermocouple.
108
-------
2. Design Variables
1) Period of Air Injection: The period of air injection
can be changed by changing the cam design, but was not
changed.
2) Nozzle Geometry: The nozzle diameter can be changed
to change the injection velocity and the nozzle direction
can be changed by drilling the nozzle at an angle to the
nozzle center line.
3) Shrouded Valve: A shrouded valve can be used to in-
duce swirl in the combustion chamber, but was not used.
3. Equivalence Ratios as Alternate Operational
Variables
It is customary to discuss engine performance in
terms of fuel-air (or air-fuel) ratio or equivalence
ratio because the performance is usually a strong func-
tion of these variables. In general, equivalence ratio
will be used. The set of three variables fuel flow
rate, carbureted air flow rate and injection air flow
rate will be replaced by an alternate set of three
variables carbureted equivalence ratio, overall equiv-
alence ratio and carbureted air flow rate. It is
necessary to also include the carbureted air flow rate
in the second set in order to specify the amount of
charge.
109
-------
4. Specified Operational Variables
Since there are ten operational variables it would
be impractical to methodically test all combinations.
If two levels of each variable were tested 1024 (2 =
1024) runs would be required. It is necessary to hold
many of the variables roughly constant and examine those
of greatest interest. Listed below are the variables which
were usually held constant and a justification for each.
Engine Speed: The engine speed was kept at about
800 RPM. This was done because of concern over the
possibility of an air injection valve failure. The dy-
namic forces on the cam follower of the air injection
valve increase with speed. At 800 RPM (13.3 rev/sec) the
valve sounds fine and should have a long life based on
calculation by the air injection valve computer program.
When lower speed operation was tried the coupling between
the dynamometer and the engine failed. It has been re-
placed by a larger coupling and has operated well at 800
RPM (13.3 rev/sec).
Carbureted Air Flow Rate: For nearly all of the
runs the carbureted air flow rate was held constant at a
level equivalent to a wide open choke and throttle at
800 RPM (13.3 rev/sec).
With the engine speed held constant it would be
possible to change the mass flow rate of carbureted air
110
-------
by adjusting the choke, but it would also be necessary to
readjust the fuel flow rate at each point "because the
change in inlet pipe pressure affects fuel flow rate. It
takes some time to establish a constant fuel flow rate
because the fuel scale does not provide an instantaneous
measure of fuel flow rate, only an integrated average
value. Several readings must be taken at each fuel flow
setting to determine if the flow is steady and at the
desired flow rate.
The advantage of a constant maximum carbureted air
flow rate is that once a constant fuel flow rate has
been established the carbureted equivalence ratio will
be constant. It is then possible to run a series of
tests at various overall equivalence ratios by changing
only the rate of air injection.
With a constant carbureted air flow rate it is
possible to calculate the fuel flow rate desired for a
specific carbureted equivalence ratio and the air in-
jection flow rate desired for a specific overall equiv-
alence ratio. A number of tables and curves were made
to assist in reaching a specific operating condition.
Carbureted Equivalence Ratio: In general, the
carbureted equivalence ratio was adjusted to approximate-
ly l.Jf. At this condition only about 65 PPM of nitrogen
oxide are produced by the carbureted mixture. This value
111
-------
of equivalence ratio also corresponds to that used in the
theoretical model.
Engine Inlet Pressure: The engine inlet pressure is
related to the carbureted air flow rate which is held
constant, thus the engine inlet pressure will also be
constant.
Compression Ratio: At compression ratios much in
excess of seven the engine will knock. It was
concluded that the engine should be operated at a com-
pression ratio of seven.
Spark Timing: In the large diameter nozzle tests,
the spark was arbitrarily set. Later it was concluded
that the spark should be set at the maximum advance for
best torque (MBT). The spark timing at the MBT repre-
sents the most power and best efficiency that can be
obtained at that operating condition.
C. A DISCUSSION OF THE THEORETICAL TRENDS EXPECTED
IN THE EXPERIMENTAL RESULTS
1. Nitrogen Oxide Emissions
Both Chapter II and Chapter III have dealt with
the trends in nitrogen oxide emissions. In summary,
the nitrogen oxide emissions for a delayed stratified
charge engine can be expected to increase with leaner
overall equivalence ratios, with leaner carbureted
112
-------
equivalence ratios, with earlier air injection and with
advanced spark timing.
2. Carbon Monoxide Emissions
In engines the most significant factors affecting
the formation of carbon monoxide (CO) are the fuel-air
ratio and the CO chemical kinetics. When a fuel-rich
mixture is burned CO is formed instead of C02 because
of the lack of oxygen. Thus the richer the fuel-rich
mixture the greater the concentration of CO in the ex-
haust .
Carbon monoxide is also found in the exhaust of en-
gines when operated with fuel-lean mixtures. The exhaust
gas concentration is determined by the fuel-air ratio,
the temperatures, the pressures and the chemical kinetic
history. In the combustion chamber the equilibrium con-
centration of CO is high after combustion and decreases
during expansion. The actual CO concentration rapidly
approaches the high equilibrium value while temperatures
are high and attempts to follow the decreasing equili-
brium values during expansion. The chemical kinetics of
CO freezes during the expansion with the resulting CO
concentration lying between that of the equilibrium
value at combustion conditions and the equilibrium value
at exhaust conditions. Newhall (1969) has described this
behavior of CO in great detail.
113
-------
In the experimental engine the amount of CO found in
the exhaust will depend upon the mixing and reaction of
the injected air with the rich products of combustion.
If the injected air does not mix and react with rich pro-
ducts, large carbon monoxide emissions would result simi-
lar to those expected for homogeneous operation at the
carbureted fuel-air ratio. If the injected air does
mix and react with the rich products, small carbon mon-
oxide emissions would result similar to those expected
for homogeneous operation at the overall fuel-air ratio.
The carbon monoxide emissions for air injection
operation before combustion will also depend upon the
mixing process. The more rapid the mixing the closer
the level of carbon monoxide emissions will be to the
homogeneous operation at the overall fuel-air ratio.
With air injection the amount of charge is greater
because the injected air adds to the constant maximum
carbureted air intake. Since the charge is greater the
cylinder pressure will also be greater. The greater cy-
linder pressure will not influence the equilibrium con-
centration of CO but it may change the CO kinetics by
changing the concentration of other species which are
pressure sensitive. It would be necessary to perform
a kinetics computer calculation to determine the pressure
effect on CO concentration.
114
-------
3- Hydrocarbon Emissions
Daniel and Wentworth (1962) showed that the primary
source of hydrocarbon emissions for conventional four-
stroke spark ignition engines was due to the quenching
of the flame as it approached the cool combustion chamber
wall. They also showed that oxidation of the hydrocarbon
emissions could occur in the exhaust pipe if sufficient
oxygen and high enough temperatures are present.
After combustion is completed an envelope of un-
burned fuel-air mixture lies around the combustion chamber
surface. The amount of unburned hydrocarbon contained
within the envelope depends of the volume of the envelope
and the fuel density within the envelope. The volume of
the envelope is changed by changes in the quench dis-
tance, piston motion and air motion. The envelope also
includes any crevices that are not penetrated by the
flame front, such as between piston and the cylinder.
The thickness of the quench layer is a function of
pressure, temperature and fuel-air ratio. Increases in
the temperature and pressure tend to decrease the quench
layer thickness. The effect of fuel-air ratio on the
quench layer thickness is complicated by its affect on
the pressure and temperature. In general, the quench
distance reaches a minimum thickness at a slightly rich
fuel-air ratio. The amount of hydrocarbons in the
115
-------
quench layer will also depend on the fuel-air ratio.
Hydrocarbon emissions are found to decrease as the fuel-
air mixture is made leaner until they become low and rela-
tively constant for lean fuel-air mixtures. But, if the
fuel-air mixture is made sufficiently lean to result in
misfire, the hydrocarbon level will increase due to the
unburned fuel.
When air is injected into the cylinder before com-
bustion, there are many factors which will affect the
hydrocarbon level and it is difficult to conclude how
these factors will sum to influence the total hydrocarbon
emissions of early injection. The increased pressure
forces more fuel-air mixture into the crevices favoring
higher hydrocarbon emissions, but also tends to reduce the
quench layer thickness favoring lower hydrocarbon emis-
sions. The increased density of the quench envelope would
tend to increase the hydrocarbons. When air is injected
during the compression stroke, it is unlikely that it will
be able to penetrate the boundary layers and crevices be-
fore combustion. Consequently, the fuel-air mixture in the
quench envelope will be fuel-rich, which would tend to in-
crease the hydrocarbon emissions. It is difficult to pre-
dict how the total hydrocarbon emissions of the early air
injection operation will compare to homogeneous operation
at the same overall fuel-air ratio.
When air is injected after combustion, the quench
envelope has already been formed under conditions of
116
-------
the rich carbureted charge. The greater fuel-air concen-
tration will tend to increase the hydrocarbon emissions.
In contrast, the injection of air will increase the veloci-
ties in the combustion chamber which will reduce the
boundary layer thickness and tend to reduce the hydro-
carbon emissions. Again it is difficult to determine how
these factors affect the hydrocarbon emissions of the
late injection operation as compared to homogeneous
operation at the same overall fuel-air ratio.
4. Engine Power
In a conventional homogeneous engine the specific
power depends on the engine speed, spark timing, fuel-
air ratio of the charge, amount of charge, compression
ratio and a number of other minor variables. Engine
speed and compression ratio have been held constant for
most of the tests. Spark timing has either been held
constant or adjusted for the maximum advance for best
torque operating condition. For most operating condi-
tions the amount of carbureted charge was also held con-
stant at the maximum value. Thus, the fuel-air ratio is
the most significant variable remaining.
When the homogeneous engine is operated fuel-rich,
power is limited by the amount of air available for com-
bustion. Since the carbureted charge is constant the
amount of power is expected to be nearly constant. At
117
-------
very rich fuel-air ratios, the fuel will .displace some of
the air causing a slight reduction in power. When the
homogeneous engine is operated fuel-lean, the power is
limited by the amount of fuel available for combustion.
As the fuel-air ratio becomes leaner the power will de-
crease .
Engine operation with air injection results in a
larger charge since the carbureted charge is constant
and the injected air will add to the total charge. The
extent of the increase in power associated with the
larger charge will depend of the total amount of charge,
the air injection timing, the carbureted fuel-air and
the overall fuel-air ratios.
The air which is injected into the combustion chamber
must be compressed to the appropriate pressure. The work
required to compress the air will depend on the process
used. If a method of compressing the air in the cylin-
der along with the rich charge and by the combustion
process can be developed,then external compressors will
not be required. The work of compressing the air would
be similar to that of compressing the charge. In the
experimental engine system,the air is compressed separ-
ately to greater than injection pressure and stored in
gas bottles until it is injected. The work used to
drive the compressor is not a reasonable estimate of the
118
-------
work required to compress the air for injection. It has
been assumed that the air compression work is best re-
presented by an isentropic compression for room tempera-
ture and pressure to injection pressure as measured in
the accumulator before the air injection valve.
5. Engine Efficiency
The fuel-air ratio significantly influences efficiency
for both homogeneous operation and air injection opera-
tion. With air injection operation two fuel-air ratios
have to be considered, that of the carbureted charge
and that of the overall charge including the injected air.
The efficiency for homogeneous charge engines is expected
to drop sharply for fuel-rich mixtures because of the
lack of oxygen to complete the combustion process. Like-
wise making a fuel-lean mixture leaner results in better
efficiency because the combustion process goes nearer
to completion.
During air injection operation the efficiency assoc-
iated with the combustion of the rich charge would
correspond to the efficiency of the homogeneous opera-
tion at the same rich fuel-air ratio. The addition of
air to the rich charge will tend to complete the combus-
tion process and increase the efficiency over that of the
carbureted charge alone. When the air enters the combus-
tion chamber early in the expansion it can do useful work
119
-------
through expansion even if it does not re.act. Increasing
the amount of air injected will make the overall mixture
leaner and further increase the efficiency. The efficiency
of the delayed mixing engine will fall somewhere between
the efficiency of the homogeneous operation corresponding
to the carbureted fuel-air ratio and to the overall fuel-
air ratio.
If the air injection occurs late in the expansion
process, little additional work can be extracted because
the amount of expansion which remains is small and in-
sufficient time may remain during the expansion to mix
and react the air with the rich products. On the other
hand if air is injected early, before combustion, it will
have time to form a more nearly homogeneous mixture at
the overall fuel-air ratio. Under this form of opera-
tion the efficiency should approach that of a correspond-
ing homogeneous operation at the same overall fuel-air
ratio. The efficiency of air injection operation is
expected to be between that of homogeneous operation
with the carbureted fuel-air ratio and with the overall
fuel-air ratio.
Spark timing is used to adjust for maximum engine
torque. When the engine is operating at the maximum
torque it is also operating at the maximum efficiency
because the power output is greatest for a constant amount
120
-------
of fuel input. With the spark too far advanced the com-
pression work will "be done against combustion pressure
and with the spark too far retarded the pressure peak
will occur later in the expansion. In either operating
condition the efficiency will "be less than maximum.
Another factor which affects the efficiency of the
engine is the heat transfered to the combustion chamber
walls. The convective heat transfer will be increased
by greater velocities in the combustion chamber. When
air is injected the velocities will be greater because
of the kinetic energy of the air jet. The efficiency
of air injection operation could be lower due to more
heat transfer than corresponding homogeneous operation.
Two measures of efficiency will be used in the pre-
sentation of results. The measured indicated enthalpy
efficiency is based on the brake horsepower of the engine
plus the mounted horsepower. The corrected indicated
enthalpy efficiency is based on the brake horsepower
plus the motored horsepower less an estimated power re-
quired to compress the injected air.
D. LARGE DIAMETER NOZZLE TEST RESULTS AND DISCUSSION
1. Check-Out
Much of the initial operation was intended to check-
out the system and to determine how best to control the
121
-------
the system. Some mechanical problems with the air in-
jection valve were discovered and corrected. The coupling
between the engine and the dynamometer was too small and
had to be replaced. The carburetor float bowl was first
replaced by a larger one and later both float bowls were
placed in series to stabilize the fuel flow rate. It
was necessary to replace the fuel pump, one was pur-
chased which incorporated a pressure regulator. The new
fuel pump greatly improved the stability of the fuel
flow to the carburetor. Improvements were made to the
fuel scale to improve its accuracy. Although the re-
pairs and maintenance were time consuming they did not
represent significant technical problems.
2. Test Conditions for Large Diameter Nozzle
The large diameter nozzle has a diameter of .3125
inches (7.938 mm) and is directed perpendicular to the
center line of the cylinder and into the clearance volume.
A flow restriction consisting of a .0935 inch (2.37 mn)
hole in a disc was placed between the air injection valve
seat and the nozzle. The flow restriction caused an in-
crease in the air injection accumulator pressure to pre-
vent the flow of combustion product into the air injection
valve. The maximum velocity through the flow restriction
is sonic velocity which resulted in the maximum velocity
from the nozzle of about one hundred feet per second.
122
-------
The engine was operated near 800 RPM with the maximum
carbureted air flow rate. The compression ratio was held
constant at seven. The spark timing was held constant at
10° BTDC. Both the injection air temperature and the
carbureted air temperature were uncontrolled. These
temperatures were usually above room temperature due to
higher temperature of the engine.
The magnitude of the other variables which changed
during the testing are shown in the figures.
All testing with the large diameter nozzle was done
with air injection after the start of combustion.
3. Summary of Results for Large Diameter Nozzle
1) The nitrogen oxide emissions for operation with air
injection are much lower than homogeneous engine opera-
tion at the same overall air-fuel ratio.
2) The carbon monoxide emissions for operation with air
injection are essentially the same as homogeneous engine
operation at the same overall air-fuel ratio.
3) The hydrocarbon emissions for operation with air in-
jection are lower than homogeneous engine operation at
the same overall air-fuel ratio.
4) The measured indicated enthalpy efficiency for
operation with air injection is lower than homogeneous
engine operation at the same overall air-fuel ratio.
123
-------
5) The corrected indicated enthalpy efficiency for opera-
tion with air injection is much lower than homogeneous
engine operation at the same overall air-fuel ratio.
4. Nitrogen Oxide Emissions for Large Diameter
Nozzle
A set of NO emission results for the low velocity
nozzle is shown in Figure 28. The homogeneous curve was
obtained by operating the engine without air injection
at various overall equivalence ratios. At very rich
operation the nitric oxide concentrations are very low.
As the overall equivalence ratio becomes leaner the
nitric oxide increases until it peaks at about / = .95.
At this equivalence ratio the engine combustion becomes
unstable which in part accounts for reduced NO at lower
overall equivalence ratios. The lower flame temperatures
associated with fuel lean operation also reduce the NO
X
emissions. An NO peak with slightly fuel-lean mixtures
is typical of homogeneous engine operation.
Also shown in Figure 28 is a set of results with
air injection having a carbureted equivalence ratio (/_)
c
of about 1.4 and at various air injection timings. With
the start of air injection at 42° ATDC (crank angle
degrees^ very little additional NO is formed due to an
injection. Advancing the air injection to start at 22°
ATDC increases the N0x slightly. With the air injection
124
-------
1000 —
I 300
o
SE
-------
starting at 12° ATDC,the amount of NO formed increases
Jv.
markedly. When the air is injected earlier, it is injected
into higher temperature products. Thus, in the process of
mixing and reacting with the rich products,it is more
likely to produce additional NO emissions.
X
For each of the air injection timings, the curves were
connected to the homogeneous operating point corresponding
to their carbureted equivalence ratio. The point on the
homogeneous curve represents the amount of NO which is
Jv.
formed during the period of rich combustion. The in-
jection of air will reduce the overall equivalence ratio
and result in additional NO formation. It is possible
Jt.
to operate at different points along the curve by chang-
ing the amount of air which is injected.
One set of data was obtained at a carbureted equiv-
alence ratio of about 1.6 and with air injection starting
at Jj-20 ATDC. Although the NO formed by the carbureted
J\.
charge is lower than that formed by /c = 1.^ the ex-
haust NO is about the same. The slope of the curve
Jv
is much steeper than the corresponding curve at tfc = 1.4-
with air injection starting at 42° ATDC. The difference
in the shape of the curves can be explained by consider-
ing the differences in the amounts of air which must be
injected. To reach an overall stoichiometric equivalence
ratio, it is necessary to inject k-Q%> more air into the
126
-------
combustion chamber if the carbureted equivalence ratio
is 1.4- but 60$ more air if the carbureted equivalence
ratio is 1.6. Since the total charge will be substantial-
ly greater, the cylinder pressure will be greater. The
relative increase in pressure will compress the products
of combustion and increase their temperature. The
higher temperature will result in more rapid nitric oxide
formation.
5. Carbon Monoxide for Large Diameter Nozzle
Presented in Figure 29 is a plot of the exhaust gas
carbon monoxide versus overall equivalence ratio. The
homogeneous operating points, indicated by squares,
are connected by the solid line. The other points,
representing air injection operation, are not connected
because they are very near the homogeneous curve.
These results show that the injected air has suf-
ficient time to mix and react with the products of the
rich charge. The combustion is completed to the same
degree as homogeneous operation with the same overall
equivalence ratio. Under these air injection operating
conditions, the exhaust emissions of carbon monoxide are
mainly a function of overall equivalence ratio.
6. Hydrocarbon Emissions Large Diameter Emissions
The hydrocarbon emissions are presented in Figure
30. For operation with air injection, the hydrocarbon
level is substantially below that of the homogeneous
127
-------
3-0
c
o
o
fc
0>
p.
o
2
o
c
o
o
0>
•o
•H
X
o
c
o
c
o
€
id
o
2.0
1.0
Homogeneous
Operation
6
si
O 12°ATDC
22°ATDG
A i*2°ATDC
O ^2
I I
1.38
1.38
1.6
.8 l.O 1.2
Overall Equivalence Ratio
1.6
Figure 29
Carbon Monoxide Emissions versus
Overall Equivalence Ratio for
Various Air Injection Timings
With a Large Diameter Nozzle
128
-------
3000
c
o
a)
o
e
ex
c
o
-p
c
0)
o
d
o
o
c
o
43
t-l
rt
o
o
2000
1000
J.
J_
Homogeneous
Operation
si
O 12°ATDC
22°ATDC
A
O
_L
.8 1.0 1.2 1
Overall Equivalence Ratio
1.38
1.38
1.6
1.6
Figure 30
Hydrocarbon Emissions versus Overall
Equivalence Ratio for Various Air
Injection Timings With a Large
Diameter Nozzle
129
-------
operation with the same overall equivalence ratio. The
most likely explanation for the lower hydrocarbons is
that the turbulence and general increase in the motion
of the charge due to air injection reduces the thickness
of the quench layer.
?. Engine Efficiency for Large Diameter Nozzle
Because the injection air is compressed separate-
ly from the engine the measured output power does not
include the work required to compress the injection air.
An estimate of the work required to compress the in-
jected air is made and subtracted from the measured work.
As a result the efficiency can be presented as measured
values or corrected values. In Figure 31 both the mea-
sured and corrected indicated enthalpy efficiencies are
presented.
The homogeneous efficiencies reach a maximum at
an overall equivalence ratio of .95- The efficiency is
lower for leaner operation because the engine experiences
partial misfire. The homogeneous efficiency is lower
at greater equivalence ratios because of the incomplete
combustion of the fuel.
Engine operation with air injection shows a general
increase in measured efficiencies with leaner equivalence
ratios. If no air were injected, measured efficiency
would be equal to that of the homogeneous operation at
130
-------
3o
e
SI
Operation
O 12°ATDG
^7 22°ATDG
A 42°ATDC
O 42°ATDC
1.38
1.38
1.6
-p
o
o
fc
0)
ex
III
w
T3
as
20
M = Measured Indicated
Enthalpy Efficiency
C = Corrected Indicated
Enthalpy Efficiency
.8
Figure 31
1.0 1.2 1.4
Overall Equivalence Ratio
Measured and Corrected Indicated
Enthalpy Efficiency versus Overall
Equivalence Ratio for Various Air
Injection Timings With a Large
Diameter Nozzle
1.6
131
-------
the carbureted equivalence ratio. As the -amount of in-
jected air increases the measured efficiency increases
and the overall equivalence ratio becomes leaner. The
addition of high pressure air during the expansion
stroke will result in more measured work. Also, when the
air mixes with the rich products, it will tend to finish
the combustion process, causing higher pressures and an
increase in measured work. Although the measured
efficiency is increased with air injection, it is less
than homogeneous operation at the same overall equivalence
ratio. The lower measured efficiency is due to combus-
tion being completed late in the expansion stroke. The
later in the expansion stroke the combustion occurs
the smaller will be the amount of expansion work.
When the measured work is corrected by subtracting
the estimated work of compressing the injected air, the
resulting corrected efficiency tends to decrease with
increasing amounts of air injection. Apparently the
increased amount of work produced by the injected air is
less than work required to compress the air. A large
compression ratio is required to bring atmospheric air
to injection pressure and only a small expansion ratio
is available in the engine. If the air did not react in
the combustion chamber and only expanded, the lower
132
-------
expansion ratio would limit the work extracted from the
air to something less than the compression work.
8. Discussion of Results for Large Diameter Nozzle
The low values of carbon monoxide and hydrocarbons
lead to the conclusion that the injected air does mix
and react to complete the combustion process. However,
the measured indicated enthalpy efficiency shows only
slightly more output power as a result of air injection.
It appears that the combustion occurs too late in the
expansion to contribute much to output power. Either
the mixing is too slow or the air is injected too late
in the expansion. It was decided that the mixing should
be increased by changing the nozzle to a smaller dia-
meter to obtain a higher velocity for more rapid
mixing.
The large diameter nozzle was picked initially be-
cause of the lower pressure drop across it. In a
practical engine it would be desirable to have the
minimum pressure drop across the nozzle. The engine is
going to have to provide the work required to inject or
mix the air with the rich products of combustion.
E. SMA.LL DIAMETER NOZZLE TEST RESULTS AND DISCUSSION
1. Test Conditions for Small Diameter Nozzle
The small diameter nozzle has a diameter of .0935
inches (2.3Tmm) and is directed at an angle of 45 degrees
133
-------
from a perpendicular to the center line of the cylinder
in a plane parallel to the pistonb top surface. The dia-
meter was selected to "be equal to that of the flow re-
striction placed between the nozzle and .the air injection
valve seat for the large diameter nozzle tests. In this
way, flow through the valve would be similar for both types
of nozzles.
The engine was operated near 800 RPM with the maximum
carbureted air flow rate. Compression ratio was maintained
at seven. The spark timing was varied in attempts to
operate at the maximum advance for best timing, but these
conditions were not always obtained. The injection air
temperature was uncontrolled and was above room tempera-
ture due to temperature of the engine. The carbureted
fuel-air mixture was heated for some of the test con-
ditions by a strip heater wrapped around the inlet pipe.
For the small diameter nozzle, some of the opera-
tion was with air injection before combustion. These
points were run to show the entire range of air-injection
timings.
The values for important variables which changed
during the tests are shown in the figures or in Table II.
2. Summary of Results for Small Diameter Nozzle
l) The nitrogen oxide emissions for engine operation
with air injection usually increase with leaner overall
equivalence ratios.
134
-------
2) The nitrogen oxide emissions for engine operation
with air injection when compared to homogeneous operation
at the same overall equivalence ratio are much lower with
air injection after combustion, are about the same with
air injection "before and rapidly increase with earlier
air injection during combustion.
3) The carbon monoxide emissions for engine operation
with air injection are usually the same as those of homo-
geneous operation at the same overall equivalence ratio,
but greater for air injection during combustion, very
early in the compression, or very late in the expan-
sion.
4) The hydrocarbon emissions for engine operation with
air injection are usually lower than homogeneous operation
at the same overall equivalence ratio if leaner than 1.15.
but greater for air injection during combustion or
or very early in the compression.
5) The measured indicated enthalpy efficiency for engine
operation with air injection is always less than homo-
geneous operation at the same overall equivalence ratio.
The measured efficiency for air injection operation is
low with air injection after combustion, approaches the
homogeneous operation with air injection before combus-
tion and is generally low with air injection during
combustion.
135
-------
6) The corrected enthalpy efficiency for engine operation
with air injection is always much lower than homogeneous
operation at the same overall equivalence ratio.
?) The measured indicated mean effective pressure for en-
gine operation with air injection is higher with lean
operation and with early air injection. Air injection
after combustion has low IMEP. With air injection during
combustion the measured IMEP is at a minimum.
8) When air is injected before combustion the combustion
process is rapid, the cycle to cycle pressure variations
are small, and the peak pressure is greater than the
corresponding homogeneous operation.
9) When air is injected very early in the compression
stroke (128° BTDC) the combustion process will result
in knock.
10) Air injection can cause up to k-Q% more heat trans-
ferred to the cooling water on the basis of BTU per
pound of charge.
3. Nitrogen Oxide Emissions for Small Diameter
Nozzle
The nitrogen oxide emissions associated with homo-
geneous operation are shown in Figure 32. The data re-
presents a band rather than a single line because of the
variation in the spark timing, same operation was with
136
-------
3000
1000
&
p.
o
25
a 300
T3
•H
§
c
o>
100
30
\ L 1111
.8 1.0 1*2 1.4
Overall Equivalence Ratio
1.6
Figure 32 Nitrogen Oxide Emissions for
Various Homogeneous Operating
Conditions
137
-------
engine inlet pipe heated, the accumulation of small mea-
surement errors, and the small variation in operating
conditions. The maximum nitrogen oxide emissions occur
at a slightly lean equivalence ratio as would be expected.
The engine was not operated much leaner than the maximum
NO because of the start of misfiring.
J\.
In Figure 33 a few typical air injection operating
conditions are shown along with the band of homogeneous
operation as a function of overall equivalence ratio.
This same set of data will be shown for the other results
versus overall equivalence ratio. In general, starting
air-injection earlier in the cycle results in large
nitrogen oxide emissions. With air injection consider-
ably before combustion as represented by the start of
air injection at 38° BTDC and 88° BTDC the nitrogen
oxide emissions correspond to the levels expected from
homogeneous operation at the same overall equivalence
ratio. When air injection is very late in the cycle
(92° ATDC) leaner equivalence ratios cause a reduction
in the nitrogen oxide concentration. The air injection
occurs when the temperatures are so low that little addi-
tional nitrogen oxide is formed and the air reduces the
concentration by dilution.
In order to better show the influence of the timing
of air injection on nitrogen oxide and other results,
138
-------
3000
X ~~ " ^
\
N \ A
A \ 0
/v\ v
si
92°ATDG
22°ATDC
2°ATDG
X
^C
1.39
1.42
1.43
i 10
1000
p.
p.
o
2;
5 300
•H
g
-P
•H
2
100
30
18°BTDC 1.43
38°BTDG 1.40
88°BTDC 1.41
Homogeneous
Operation
L_J « L
I 1 L
J
.8 1.0 1.2 1.4 1.6
Overall Equivalence Ratio
Figure 33
Nitrogen Oxide Emissions versus
Overall Equivalence Ratio for
Selected Air Injection Operating
Conditions With a Small Diameter
Nczzle
139
-------
they will be plotted versus start of air injection. Since
various operating conditions were run at one air injec-
tion timing the points shown will correspond to an over-
all stoichiometric mixture. The magnitude of any speci-
fic point was obtained by plotting the variable versus
overall equivalence ratio, fitting a curve to the data
and noting the intersection of that curve with a stoichio-
metric mixture. A summary of the extrapolated stoichio-
metric results is shown in Table II. Generally the car-
bureted equivalence ratio was about 1.4. A carbureted
equivalence ratio is estimated for each point in Table
II. Also shown in Table II is the spark timing and start
of air injection.
The nitrogen oxide emissions versus the start of
air injection are presented in Figure 3^- Generally NO
J\.
is low for air injection after combustion and high for
air injection before combustion. When the air injection
occurs near top dead center during combustion, the nit-
rogen oxide emissions are very sensitive to changes in
the air injection timing and to changes in the other
variables. The spread in the nitrogen oxide emission
at 2° ATDC and 8° BTDC attest to the sensitivity of this
region.
It was particularly difficult to adjust the spark
for the best torque with air injection near top dead
140
-------
TABLE II
Summary of Results for Air Injection Operation
Extrapolated to a Stoichiometric
Fuel-Air Mixture
Symbol
O
A
O
A
O
O
O
O
<7
O
O
v
O
O
O
O
A
A
A
o
e .
SI
(deg.)
92
22
22
2
2
2
2
-8
-8
-8
-8
-18
-28
-28
-38
-48
-68
-78
-88
-128
6
sp
(deg.)
-26
-30
-30
-20
-20
-30
-40
-10
-10
-20
-22
-30
+10
-6
-8
0
-6
-5
-10
-6
*
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
c
39
42
60
44
38
43
35
36
39
40
42
43
,44
,24
,40
.39
1.42
1.45
1.41
1.42
NO
X
(ppm)
59
100
97
95
115
160
150
630
260
285
185
580
1350
1250
1200
1650
1250
1200
1200
260
CO
W
2.6
.5
.55
>3.0
3.0
.45
.40
1.4
>3.0
1.5
2.4
1.0
.92
1.3
.6
.85
1.3
.5
.5
2.6
HC
(ppm-C)
1150
720
950
1850
1800
780
750
1300
2030
1100
1550
700
1150
1250
640
1220
1000
1350
630
1900
Meas .
IEE
(%
21.
24.
22.
22.
21.
24.
24.
)
2
5
8
5
,5
.5
.2
24.2
21,
23,
22
22
21
26
24
25
25
26
25
22
.6
.4
.3
.7
.1
.0
.8
.8
.0
.5
.1
.3
Meas.
IMEP
(psi)
165
133
132
120
113
130
120
122
111
120
120
123
137
128
129
136
132
142
134
128
141
-------
J\J\J\J
1000
s
p<
p.
t
.§ 300
-H
c
d»
o
•H
z
100
30i
. ' ./*' ' , _, '" x' . ' /'~
- ' Homogeneous Case 4> = 1 .
/ ' - •
O.
/ /
/\/\ & /'"^Cy
mat-
••
9
o
&
"
o
r A 6
•w
^ t 1 1 i 1 1 t i 1 1 1 I 1 1 1 1 1
20 80 , 40 0 40
BTDG TDC ATDC
Start of Air Injection - Crank Angle Degrees
Figure 34
Nitrogen Oxide Emissions versus Start
of Air Injection for Overall Stoichio-
metric Equivalence Ratio With a Small
Diameter Nozzle
142
-------
center. When the air is injected "before combustion the
best timing is about 6° BTDC probably because of the
greater turbulence. When the air is injected after com-
bustion the best spark timing is about 25° BTDC. Also
changes in the spark timing will affect the timing of
the pressure peak in the cylinder, which in turn influences
the injection air pressure. The injection air pressure
takes some time to stabilize. Thus, if one were to ad-
just the spark timing by merely observing the engine
torque it would be easy to select the wrong timing. Ob-
servation of the cylinder pressure was used to determine
the spark timing near top dead center. This method re-
quires judgment as to the most appropriate position of
the pressure peak with respect to top dead center. The
adjustment of spark timing by observing the cylinder
pressure is better than observing the engine torque, but
not exact.
The spread in the nitrogen oxide data with air in-
jection starting at 2° ATDC can be explained mainly by
the spark timing. The two points with the greatest
nitrogen oxide emission have spark timings of 30 BTDC
and 40° BTDC; from the pressure trace the spark is too ad-
vanced. The two other points have spark timings of 20
BTDC; from the pressure trace the spark is too retarded.
The advanced timing causes higher pressures and temperature
143
-------
which result in more nitrogen oxide emissions. The re-
tarded timing causes lower pressures and temperatures
which result in low nitrogen oxide emissions.
The two points with 20° BTBC spark timing with start
of air injection at 2° ATDC differ in their carbureted
equivalence ratio. The low nitrogen oxide point also
corresponds to the richer carbureted equivalence ratio.
Since the richer carbureted equivalence ratio would pro-
duce less nitrogen oxide before air injection it is rea-
sonable to expect a lower final nitrogen oxide emission.
The same type of reasoning can be used to explain
the spread in the nitrogen oxide data points with air
injection at 8° BTDC. The more advanced timing results
in greater NO emissions and richer carbureted equivalence
A.
ratio result in lower NO emissions.
When air is injected before combustion the nitrogen
oxide concentration corresponds to that of homogeneous
operation with the same overall equivalence ratio.
Apparently the injected air mixes quickly to form a some-
what homogeneous mixture at the overall equivalence ratio
before combustion.
*K Carbon Monoxide Emissions for Small Diameter
Nozzle
The carbon monoxide emissions for homogeneous opera-
tion are presented in Figure 35- A curve drawn through
144
-------
3.0
c
0)
o
*-.
0)
Pk
c
o
2
•p
c
0)
o
o
o
Q)
TJ
•H
X
o
c
o
s
c
o
rt
o
2.0 —
l.C —
.8 1.0 1.2
Overall Equivalence Ratio
1.6
Figure 35 Carbon Monoxide Emissions versus
Overall Equivalence Ratio for
Homogeneous Operation
145
-------
the data points describes the homogeneous carbon
monoxide emissions. For rich operation the carbon mon-
oxide levels are high while for lean operation the carbon
monoxide emissions are low.
The carbon monoxide emissions for the typical set
of air injection operating conditions are presented in
Figure 36 along with the dotted line corresponding to
the homogeneous operation. All of the data points fall
near the homogeneous curve except for operation with the
start of air injection at 8° BTDC and 92° ATDC. High
values of carbon monoxide are expected with very late
air injection (92° ATDG) because of the short time avail-
able for combustion and because of the low temperatures
Many points with air injection starting near top
dead center have greater carbon monoxide emissions than
homogeneous operation as seen in Figure 37. It is pos-
sible that air being injected during the period of com-
bustion, as is the case for these points, can quench and
disturb the combustion process. Probably the regions of
incomplete combustion do not mix and as a result appear
in the exhaust.
5. Hydrocarbon Emissions for Small Diameter Nozzle
The hydrocarbon emissions for homogeneous operation
are shown in Figure 38. Two factors account for the
spread in hydrocarbon emissions. Part way through the
146
-------
3.0
-p
c
4)
O
f-l
4>
PH
O
£
c
o>
O
c
O
O
4)
T3
.H
X
O
c
O
£
C
O
,0
2.0
/O
U- Homogeneous
Operation
e
si
O
92°ATDC
22°ATDG
2°ATDG
8°BTDG
18°BTDG
0 38°BTDC
88°BTDC
0
i
.8 1.0 1.2 1.4
Overall Eauivalence Ratio
1.39
1 . 39
1.40
1.41
1.6
Fig-are 36
Carbon Monoxide Emissions versus
Overall Equivalence Ratio for Air
Injection Operation With a
Small Diameter Nozzle
147
-------
3.u
c
0)
o
0)
i 2.0
c
o
-p
g
c
0)
o
c
o
o
0)
•o
•H
X
o 1.0
c
o
as
c
o
JO
(L,
OS
0
/~»
— O
O
o
. 0 °
tJ
,
— ^/
O
o
_
/^>w
V_/
L ^^ Q a
/ v
- L— Homogeneous Case $ = 1
i i i i i i i i t i 1 i i i 1 _j —
120 bO 40 0 40
BTDC TDC ATDC
Start of Air Injection - Crank Angle Degrees
Figure 3?
Carbon Monoxide Emissions Versus Start
of Air Injection for Overall Stoichio-
metric Equivalence Ratio With a Small
Diameter Nozale
148
-------
3000
c
o
rt
o
2000
c
o
'•p
I
c
0)
o
c
o
o
c
o
•s
0)
o
o
1000
D
j I
I I
I I
•8 1.0 1.2 1
Overall Equivalence Ratio
1.6
Figure 38 Hydrocarbon Emissions versus
Overall Equivalence Ratio for
Homogeneous Operation
149
-------
testing program it was found that the wall of the engine
inlet pipe was wetted by fuel and there existed the pos-
sibility of liquid fuel entering the inlet valve. For
this reason the walls of the inlet manifold were heated.
As a result of the heater the hydrocarbon emissions
dropped. Another factor is the improvement of the cali-
bration procedure of the hydrocarbon analyzer. The hydro-
carbon analyzer is sensitive to the flow rate of sample
and calibration gas. These two flow rates should be the
same. A pressure regulator in the instrument is used to
adjust the scale reading to correspond to the hydrocarbon
concentration of calibration gas. It was found that
changes in the calibration gas pressure supplied to the
instrument would change the calibration gas flow rate.
A procedure was devised to measure the calibration gas
flow rate with the carbon monoxide flow meter and to set
the flow rate equal to the sample flow rate.
The hydrocarbon emissions for the typical set of
air injection timings are shown in Figure 39 along with
the dotted lines representing the band of homogeneous
operating results. With the exception of air injection
starting at 8° BTDC the data shows lower hydrocarbon
emissions for air injection operation with overall
equivalence ratios leaner than 1.15. With air injection
starting very late in the expansion (92° ATDC), the
150
-------
3000
c
o
X)
k
rt
o
P.
C
o
d
o>
o
C
o
o
o
£>
t-i
cd
o
I
2000
1000
92°ATDC 1.39
A 22°ATDC 1.42
O 2°ATDC 1.43
8°BTDG 1.39
18°BTDC 1.43
38°BTDC 1.40
88°BTDC 1.41
.8 1.0 1.2 1.4
Overall Equivalence Ratio
1.6
Figure 39
Hydrocarbon Emissions versus
Overall Equivalence Ratio for
Air Injection Operation with a
Small Diameter Nozzle
151
-------
hydrocarbons are nearer the homogeneous operation. The
greater turbulence in the chamber reduces the hydrocarbon
envelope and consequently the hydrocarbon emissions.
Air injection starting near top dead center results
in greater hydrocarbon emissions as shown in Figure ^0.
The disturbance of the combustion process by air injection
would account for the greater hydrocarbons and would be
consistent with the greater carbon monoxide emissions
found for the same injection timing.
6. Engine Efficiency for Small Diameter Nozzle
Shown in Figure 4-1 is the measured indicated en-
thalpy efficiency as a function of the overall equivalence
ratio for the homogeneous operation. The efficiency de-
creases with richer equivalence ratios. The spread in
the data is primarily due to variations in the timing
and the accuracy of reading the dynamometer force. The
low speed of the engine and the cycle to cycle variations
of the engine resulted in a pulsating output torque. The
needle on the dynamometer scale would vibrate requiring
an estimate of its average value.
The measured indicated enthalpy efficiency for the
typical set of air injection data is shown in Figure ^2
with the homogeneous operation shown by dotted lines.
Earlier air injection as well as leaner overall operation
tend to increase the measured indicated enthalpy efficiency.
152
-------
3000
c
o
JO
a
o
6 2000
c
o
cd
-p
c
0)
o
C
o
o
§ 1000
cd
o
o
Homogeneous Case = 1
A
0
'
J 1 L
II I
J I L
J I L
120
80
40
BTDC
0
TDC
40
ATDC
Start of Air Injection - Crank Angle Degrees
Figure
Hydrocarbon Emissions versus Start of
Air Injection for Overall Stoichiometric
Equivalence Ratio with a Small Diaweter
Nozzle
153
-------
30
20
c
0)
o
IH
t>
P,
W
3
to
flS
o
10
.8 1.0 1.2
Overall Equivalence Ratio
1.6
Figure
Measured Indicated Enthalpy Efficiency
versus Overall Equivalence Ratio for
Homogeneous Operation
154
-------
30
20
c
-------
Note that with injection "before combustion,the measured
efficiency is less than the corresponding homogeneous
operation. The difference in efficiencies is probably
due to greater heat transfer caused air injection induced
charge motion. An increase of 40$ in the heat transferred
to the cooling water on the basis of BTU per pound charge
was measured for operation with air injection.
In Figure 4-3 the corrected indicated enthalpy
efficiency is shown as a function of overall equivalence
ratio. With injection before combustion the corrected
efficiency is about equal to the efficiency of the
homogeneous operation with the carbureted equivalence
ratio. The estimated work required to compress the air
is a substantial percentage of the engine output.
The measured indicated enthalpy efficiency is plotted
versus the start of air injection in Figure *»4. The
greater measured efficiencies occur when the air is in-
jected into the combustion chamber before ignition. The
measured efficiency is larger because the injected air
has time to mix with the rich charge before combustion
thus allowing more fuel to burn nearer top dead center.
When air injection occurs well into the expansion stroke
little additional work can be obtained from the comple-
tion of combustion.
With the start of air injection at 2° ATDC the spread
in the efficiency can be attributed to the different spark
156
-------
30
-p 20
c
o
0)
p.
1
w
M
-P
O
£
£ 10
o
o
0
\
\
-
N ^ .^Homogeneous
\ \ Operation
\ \
\ \
N x
X \
"• x. ^
\\
-
v X
yx^/-y^^V ^ ^ Xx
— -^^^/^fc^ V^-
C? ^4^^^^ ^
—L^Qj
— ^ \r
r-jf' ^\
^^/^
" oT
\-S
9si
— O 92°ATDC
A 22°ATDC
O 2°ATDC
S7 8°BTDC
^7 18°BTDG
O 38°BTDC
£S 88°BTDC
i i t i t i i
.8 1.0 1.2 1.4
' N.
^ ^x
^C
1.39
1.42
1.43
1.39
1.43
1.40
1.41
i
1.
Figure 43
Overall Equivalence Ratio
Corrected Indicated Enthalpy Efficiency
versus Overall Equivalence Ratio for
Air Injection Operation With a
Snail Diameter Nozzle
157
-------
30
1: 20
E
0)
1
w
f-r^
TJ
DO
cd
V
a 10
0
1
/~~ " "^ / x'' . / • ••'
/' Homogeneous Case 0 = 1 /'
AA, o 0
' ° O 03 *
^HA °
0 ^0
••^••K
—
i i i ! i i i 1 i i i iii _l — i —
20 80 " 40 0 40
BTDC TDC ATDC
Start of Air Injection - Crank Angle Degrees
Figure
Measured Indicated Enthalpy-Efficiency
versus Start of Air Injection for Stoi-
cniometric Equivalence Ratio With a
Small Diameter Nozzle
158
-------
timing. Spark timing of 30° BTDC and *4-0° BTDC have greater
efficiency than the two points at 20° BTDC. The four
points corresponding to air injection starting at 8 BTDC
have a spread that can not be explained entirely by spark
timing. The highest and lowest efficiency points both
have a spark timing of 10° BTDC. The difference between
the two points is due to the carbureted equivalence
ratio. The higher efficiency point has the leaner car-
bureted equivalence ratio. It is reasonable to expect
that leaner carbureted equivalence ratios would result
in greater efficiencies.
The measured indicated enthalpy efficiencies are
quite low for air injection during the combustion pro-
cess. This result is consistent with the high CO and
hydrocarbons for the same injection timing. If the in-
jection of air disturbs the combustion process and cause
incomplete combustion less energy will be available to
do work.
7. Engine Power for Small Diameter Nozzle
The measured indicated mean effective pressure (IMEP)
will be used as a measure of the specific power of the
engine. Presented in Figure ^5 is a plot of IMEP versus
overall equivalence ratio for homogeneous operation.
The IMEP is nearly constant in the rich region and de-
creases in the lean region with leaner operation. The
159
-------
•H
ro
150
1*4-0
3
M
0)
0)
fc
a,
130
o
a>
«M
«H
w
rt
a>
H
T3
0)
-P
at
o
•H
120
110
100
90
J.
J_
.7" .8 .9 1.0 1.1 1.2 1.3 l
Overall Equivalence Ratio
1.5 1-6
Figure
Indicated Mean Effective Pressure
versus Overall Equivalence Ratio
for Homogeneous Operation
160
-------
homogeneous data represents a band of results mainly be-
cause of the difficulty in setting the MET spark timing"and
the fluctuations in the dynamometer torque reading.
Presented in Figure 46 is the typical set of air
injection data versus the overall equivalence ratio. The
measured IMEP is seen to increase with leaner overall
equivalence ratios. This trend reflects the increasing
amount of injected air. In a sense the injected air
supercharges the engine. The influence of the timing
of air injection is given in Figure 4?. A minimum in
the measured IMEP occurs near top dead center when the
air is being injected during combustion. The air in-
jection apparently disturbs the combustion process causing
incomplete combustion with reduced power.
When the IMEP is corrected by subtracting the esti-
mated work required to compress the injected air the
magnitude of the corrected IMEP is equal to or less than
the corresponding homogeneous operation. Air injection
before combustion results in the largest corrected IMEP.
8. Start of Air Injection at 128° BTDC
The results obtained with air injection starting
at 128° BTDC will be presented separately because they
deviate from the pattern of the other data. By referring
to Table II it is possible to compare the result of the
128° BTDC air injection timing with the result of other
161
-------
150
e
31
140 -
O 92°ATDG 1.39
A 22°ATDC 1.42
O
2°ATDC
130
CD
ft
a,
* 120
01
«J
0)
110
Q 38°BTDG 1.40
Homogeneous
Operation
100
90
I I
i I
.8 1.0 1.2 1.4
Overall Equivalence Ratio
1.6
Figure 46
Measured Indicated Enthalpy Efficiency
Versus Start of Air Injection for Over-
all Stoichiometric Equivalence Ratio
With a Small Diameter Nozzle
162
-------
i!?u
140
130
•H
CO
P<
6 120
55
H- i
0)
CO
2 no
as
100
90
A
o O
A A
o 9
O<> °
S§
, C\
7 7^ — 7 7" / ^2 W ' / /
Homogeneous Case = 1 \/ / /
ill i I i i i f | i i i J —
>0 80 40 0 40
BTDC TDC ATDC
Start of Air Injection - Crank Angle Degrees
Figure
Measured Indicated Mean Effective
Pressure Versus Start of Air In-
jection for Overall Stoichiometric
Equivalence Ratio With a Small
Diameter Nozzle
163
-------
early air injection timings. The 128° BTDC air injection
timing has less nitrogen oxide emissions, greater carbon
monoxide emissions, greater hydrocarbon emissions and
lower efficiency. Knock was also observed for the 128
BTDC air injection timing and not in the other air in-
jection timings.
A second set of data was obtained at the 128° BTDC
air injection timing with the nozzle rotated 90 so that
the air jet is directed down into the cylinder and at an
angle of ^5° to the cylinder center line. For the ro-
tated air jet, the nitrogen oxide, carbon monoxide and
hydrocarbon emissions approach the values of other early
air injection. The pressure trace indicates knock,
which accounts for the low efficiency.
The probable reason for the difference in the re-
sults is that when the air is injected in the plane of
the piston surface a stratification results with air on
top and rich fuel mixture below. By rotating the nozzle
down into the cylinder the stratification is prevented.
The knock for both cases is probably due to the
extra compression of the charge when it is injected
very early because of the large cylinder volume at the
time of air injection.
9. The Effect of Air Injection on Combustion
One interesting discovery was the effect of air
164
-------
injection on the combustion process. The time interval
between the spark and peak pressure was reduced from about
*K) crank angle degrees for homogeneous operation to 15
crank angle degrees with early air injection. As a re-
sult of the more rapid combustion,the MET spark timing
was changed from about 25° BTDC homogeneous operation to
about 6° BTDC. Also the cycle to cycle peak pressure
variations were nearly eliminated. Presented in Figure
W is a set of pressure versus time curves which show typical
homogeneous operation, air injection during combustion,
and air injection before combustion. The marks along
the bottom of the pressure trace represent 10 crank
angle degree increments with the large mark at top dead
center.
The most probable explanation for the rapid and
consistent combustion is that the turbulence resulting
from air injection causes increased flame velocities.
Lancaster et al. (19?6) have shown that the flame velocity
increases with greater turbulence in the combustion cham-
ber of spark ignition homogeneously charged engines. The
kinetic energy of the injected air is large because of
its high velocity. Much of the kinetic energy will be
converted to random turbulent motion. The influence of
the turbulence can be seen when the air is injected
during the combustion process as shown in Figure ^8b. A
165
-------
600
Figure 48a
Homogeneous
Operation
sp
BTDC
Figure 48b
Air Injection
During Comb.
1.10
V5
2°BTDC
6 . =
31
e
sp
20°BTDC
Figure 48c
Air Injection
Before Comb.
*o - 1.J7
0 ~~ X • ™T 0
6^= 38°BTDC
e
31
i
sp
8°BTDC
•H
CQ
ft
I
0)
3
(O
CO
0)
Ok
•H
CO
I
0)
01
CO
0)
CO
p<
I
0)
(D
(^
a,
300
0
600
I i i
TDG ^4-0 80 120 160
300
I i i
120 160
600
300 _
TDC 40
Crank Angle
80 120
- degrees
160
Figure 48
Cylinder Pressure Versus Crank Angle
for Homogenous Operation, Air Injec-
tion During Combustion and Air In-
jection Before Combustion
166
-------
definite increase in the rate of combustion is observed
as a result of air injection.
10. Discussion of Results for Small Diameter Nozzle
It is apparent that air injection after combustion
can reduce the nitrogen oxide and hydrocarbon emissions.
Carbon monoxide emissions are at essentially the same
level for either operation with air injection or homo-
geneously. However, when the timing of air injection is
advanced nearer the combustion period to extract more
work, all the emissions increase and the efficiency de-
creases. It seems very unlikely that the delayed mixing
concept will be able to limit emissions while operating with
homogeneous mixture efficiencies. The measured efficiencies
are low with air injection because of the incomplete com-
bustion when the air is injected during combustion and
because of the increased heat transfer due to charge motion.
When the work required to compress the injected air is
charged against the cycle the efficiency is very low.
Likewise the corrected IMEP will also be low.
The air injection before combustion resulted in very
rapid and consistent combustion. Even with the desirable
combustion, the measured efficiency was less than a cor-
responding homogeneous operation probably because of the
additional heat transfer associated with the air injection
induced charge motion.
167
-------
REFERENCES
Adams, W. E., Marsee, F. J., Olree, R. M. , and Hamiliton,
J. C. (1976), Emissions, Fuel Economy, and Durability of
Lean Burn Systems," SAE Paper 760227, Automotive Engi-
neering Congress and Exposition, Detroit, Michigan, Feb.
23-27, 1976.
ASTM (1971), ASTM Manual for Rating Motor. Diesel, and
Aviation Fuels. American Society for Testing and Materials
1916 Race St., Philadelphia, Pa. 19103, 1971-
Blumberg, P. and Kummer, J. T. (1971), "Prediction of NO
Formation in Spark-Ignited Engines - An Analysis of
Methods of Control," Combustion Science and Technology,
1971, Vol. 4, pp. 73-95.
Blumberg, P. N. (1973), "Nitric Oxide Emissions From
Stratified Charge Engines: Prediction and Control,"
Combustion Science and Technology, 1973, Vol. 8, pp. 5-24.
Borman, G. L. (1964), "Mathematical Simulation of Internal
Combustion Engine Processes and Performance Including
Comparisons with Experimentation," Ph.D. Dissertation,
University of Wisconsin,
Cakir, H. (1974), "Nitric Oxide Formation in Diesel En-
gines," The Institution of Mechanical Engineers, Combus-
tion Group, Vol. 188 46/74, pp. 477-483.
168
-------
Chapman, A. J. and Walker, W. F. (1971) • Introductory Gas
Dynamics. Holt, Rinehart and Winston, Inc., New York,
1971-
Daniel, W. A. and Wentworth, J. T. (1962), "Exhaust Gas
Hydrocarbons - Genesis and Exodus," SAE Paper *J-86B, SAE
National Automobile Week, March, 1962.
Davis, G. C., Krieger, R. G. and Tacaczynski, R. J. ( 197*0 •
"Analysis of the Flow and Combustion Processes of a Three-
Valve Stratified Charge Engine With a Small Prechamber,"
SAE Paper 7*4-1170, International Stratified Charge Con-
ference, Troy, Michigan, Oct. 30-Nov. 1, 197^.
De Soete, G. G. (197*0, "Overall Kinetics of Nitric Oxide
/
Formation in Flames," First Joint Meeting of the Chemistry
and Pulverized Fuel Panels of the International Flame Re-
search Foundation, October 3-^,
El-Messiri, A. I. (1973), "The Divided Combustion Chamber
Concept and Design for Control of SI Engine Exhaust Air
Pollution Emissions," Ph.D. Dissertation, University of
Wisconsin, 1973-
Evers, L. W., Myers, P. S. and Uyehara, 0. A. (197*0 , "A
Search for a Low Nitric Oxide Engine," SAE Paper 7^1172,
International Stratified Charge Engine Conference, Troy,
Michigan, Oct. 30-Nov. 1, 197^-
169
-------
Fenimore, G. P. (1971), "Formation of Nitric Oxide in
Premixed Hydrocarbon Flames," Thirteenth Symposium (Inter-
national) on Combustion, pp. 373-380, The Combustion
Institute.
Heywood, J. B. and Keck, J. C. (1973). "Formation of
Hydrocarbons and Oxides of Nitrogen in Automobile Engines,"
Environmental Science and Technology. Vol. 7, Number 6,
March 1973, pp. 216-223.
Huls, T. A. (1966), "Spark Ignition Engine Operation and
Design," Ph.D. Dissertation, University of Wisconsin,
1966.
Ingham, M. (1976), Unpublished data provided by personal
communication. *
Iverach, D., Kirov, N. Y. and Haynes, B. S. (1973), "The
Formation of Nitric Oxide in Fuel-Rich Flames," Combus-
tion Science and Technology, 1973t Vol. 8, pp. 159-164.
John, J. E. A. (1975), "Lean Burning Engine Concepts -
Emissions and Economy," SAE Paper 750930, Automobile
Engineering and Manufacturing Meeting, Detroit, Michigan,
Oct. 13-17, 1975-
Kays, W. M. (1966), Convective Heat and Mass Transfer,
McGraw-Hill Book Company, New York, 1966.
170
-------
Khan, I. M., Greeves, G. and Wang, C. H. T. (1973).
"Factors Affecting Smoke From Direct Injection Engines
and a Method of Calculation," SAE Paper 730169, Inter-
national Automotive Engineering Congress, Detroit,
Michigan, Jan. 8-12, 1973-
Lancaster, D. R., Krieger, R. B. and Lienesch, J. H.
(1975). "Measurement and Analysis of Engine Pressure
Data," SAE Paper 750026, Automotive Engineering Congress
and Exposition, Detroit, Michigan, Feb. 2*J»28, 1975-
Lancaster, D. R., Krieger, R. B., Sorenson, S. C. and
Hull, W. L. (1976), "Effects of Turbulence on Spark
Ignition Engine Combustion," SAE Paper 760160, Automo-
tive Engineering Congress and Exposition, Detroit,
Michigan, Feb. 23-27, 1976.
Lauck, P., Uyehara, 0. A. and Myers, P. S. (1962), "An
Engineering Evaluation of Energy Conversion Devices,"
SAE Paper 463A, Automotive Engineering Congress, Detroit,
Michigan, Jan. 8-12, 1962.
Lavoie, G. A., Heywood, J. B. and Keck, J. C. (1970),
"Experimental and Theoretical Study of Nitric Oxide For-
mation in Internal Combustion Engines," Combustion
Science and Technology, 1970, Vol. 1, pp. 313-326.
171
-------
Lavoie, G. A. and Blumberg, P. N. (1973), "Measurements
of NO Emissions From a Stratified Charge Engine" Com-
parison of Theory and Experiment," Combustion Science
and Technology, 1973, Vol. 8, pp. 25-37.
Mabie, H. H., Osvirk, F, W. (1958), Mechanics and Dynamics
of Machinery. John Wiley and Sons Inc., New York, N. Y.
Mitchell, E., Alperstein, M., Cobb, J. M. and Faist,
C. H. (1972), "A Stratified Charge Multifuel Military
Engine - A Progress Report," SAE Paper 720051, Automo-
tive Engineering Congress, Detroit, Michigan, Jan. 10-14,
1972.
Monaghan, M. L., French, C. C. J., and Freese, R. G.
(197*0, "A Study of the Diesel as a Light-Duty Power
Plant," EPA 460/3-74-011, U. S. Environmental Agency,
Emission Control Technology Division, Ann Arbor, Michigan,
48105, 1974.
Newhall, H. K. (1969), "Kinetics of Engine-Generated
Nitric Oxides and Carbon Monoxide," Twelfth Internation-
al Combustion Symposium, August, 1969» PP- 603-613.
Newhall, H. K. and Shahed, S. M. (1971), "Kinetics of
Nitric Oxide Formation in High-Pressure Flames," Thir-
teenth Symposium (International) on Combustion, pp. 381-
390, The Combustion Institute.
172
-------
Nightingale, D. R. (1975), "A Fundamental Investigation
into the Problem of NO Formation in Diesel Engines,"
SAE Paper 75084-8, 1975 SAE Off -Highway Vehicle Meeting,
Milwaukee, Wisconsin, September 8-11, 1975-
Obert, E. F. (1968), Internal Combust ion^Enginejs , " Inter
national Textbook Company, Scranton, Pennsylvania, Third
edition, 1968.
Pischinger, R. and Cartellieri, W. (1972), "Combustion
System Parameters and Their Effect Upon Diesel Engine
Exhaust Emissions," SAE Paper 720756, National Combined
Farm, Construction and Industrial Machinery and Power-
plant Meetings, Milwaukee, Wisconsin, Sept. 11-1^, 1972.
Purins, E. A. (1974-) , "Pre-Chamber Stratified Charge
Engine Combustion Studies," SAE Paper 7^1159, Inter-
national Stratified Charge Conference, Troy, Michigan,
Oct. 30-Nov. 1,
Rhee, K. T. (1976), Unpublished data provided by per-
sonal communication.
Shahed, S. M., Chiu, W. S. and Yumlu, V. S. (1973). "A
Preliminary Model for the Formation of Nitric Oxide in
Direct Injection Diesel Engines and Its Application in
Parametric Studies," SAE Paper 730083, International
Automotive Engineering Congress, Detroit, Mich., Jan. 8-
12, 1973-
173
-------
Tasuku Date, Shizuo Yagi, Akira Ishizuya and Isao Fujii
(19?4), "Research and Development of the Honda GVCC
Engine," West Coast Meeting, Anaheim, Calif., August
12-16, 1974.
Turkish, C. M. (197^), "3 - Valve Stratified Charge
Engines: Evolvement, Analysis and Progression," SAE
Paper 7^1163 i International Stratified Charge Engine
Conference, Troy, Michigan, Oct. 30 - Nov. 11,
Tuteja, A. T. (1972), "The Formation of Nitric Oxide
in Diffusion Flames," Ph.D. Dissertation, University
of Wisconsin, 1972.
Weibe, I. (1956), "Halbempiriesche Formel fur die
Verbrennungsgeschwindigkeit," Moskau; Verlag der
Akademie der Wissenschaften der VdSSr.
Yasuo Sakai, Kazuya, Kunii, Saburo Tsutsumi and Yasuhiko
Nakagawa (197*4-), "Combustion Characteristics of the Torch
Ignited Engine," SAE Paper 7*4-1167, International Strati-
fied Charge Engine Conference, Troy, Michigan, Oct. 30-
Nov. 1, 197^.
174
-------
Appendix A
METHOD USED TO CALCULATE NITRIC OXIDE
The rate of formation of nitric oxide is described
by the following set of equations.
Rate Constants:
#2 + 0 -> NO + N ; K = 7 x 1013 exi
NO + N + N2 + 0 ; Klb = 1.55 x 1013
02 + N -» NO + 0 J K2f = 13.3 x 109T
NO + 0 - 02 + 2V 5 £9, = 3.2 x 109T exp(-39,100/7?T)
Rate Equations:
[0] + X?-[02]
^^L»2J L^J "lfc
K2b[NO][0]
or
K2f[02][N]
B = Xl
For the case of constant temperature, pressure and
equivalence ratio, A and B are constant. The solution
to this first order differential with constant coeffi-
cients is given below.
[NO] 2 = f + MO}!- f; «-*'**-*'> ; t = time
*degrees Kelvin
175
-------
This equation was used to predict the nitric oxide forma-
tion by evaluating A and B at an average value of tempera-
ture, pressure and equivalence ratio for a time step.
176
-------
APPENDIX B
ONE SYSTEM COMPUTER MODEL
The following set of equations can be written for
the one system model when it consists of two regions, re
actants and products as seen in Fig. Bl.
Conservation of Energy:
= - PVR
t\ n
Conservation of Mass:
mP
Conservation of Volume:
7=7 + v
VT R P
Ideal Gas Relationships:
PVp =
Also the internal energy, its rate of change with the
temperature, and the gas constant for the reactants can
177
-------
be obtained from equations describing these properties
as functions of pressure, temperature, fuel equivalence
ratio, and exhaust gas recirculation. The internal
energy of the products and the gas constant, their rates
of change with respect to pressure, temperature, and
equivalence ratio are obtained from equations describing
these properties as functions of pressure, temperature,
fuel and equivalence ratio.
The two region equations can be solved for the
following rate equations.
178
-------
Reactants
Temperature (TR~)
Mass (mfi)
Volume (^p)
Exhaust Gas Recirculation (EGR~)
Internal Energy
Gas Constant
Enthalpy (hR~)
Products
Temperature (T )
Mass (m )
P
Volume (Vp)
Internal Energy (w )
Gas Constant (fl )
Enthalpy O )
System
Pressure (p)
Fuel to Air Equivalence Ratio (F)
Volume (7T or y^)
Rate o£ Change o£ Volume (Vy or yy)
Fraction of the system mass which is products (a)
Rate of combustion (m or a)
Figure Al. One System Thermodynamic Model
179
-------
a
o
r> s aT " oT
2y?T
"
V rl f TP
TP IF
where: _ P
15
-------
Main Computer Program: The computer program uses
a modified Euler method to predict the temperature of the
reactants and the products. The pressure at the new
point is determined from the temperatures at the new
point, the gas constants and the total specific volume
at the new point.
Computer Program Input:
NUR: The number of different sets of engine operating
conditions to be calculated.
TITLE: A one line title for each set of operating con-
ditions, this could be the run identification
number.
P: The initial pressure in psia.
PHE: The equivalence ratio of the charge.
TR: The initial temperature of the reactants in
degrees rankine.
EGR: The fraction of the total mass which is re-
circulated exhaust gas.
TOE: The time of expansion in seconds.
XXA: A shape factor of the Wiebe function describing
expansion, see Figure 3.
XXM: A shape factor of the Wiebe function describing
expansion, see Figure 3.
TOC: The time of combustion in seconds.
181
-------
XA:
XM:
Function
DELT(s):
Function
VOL(s):
Function
ALF(s) :
Subroutine
FENERG
Subroutine
EQFLT( ):
Subroutine
NO ( ) :
Subroutine
EQBM:
Subroutine
ENERGY:
The shape factor of the Weibe function de-
scribing combustion, see Figure 2.
A shape factor of the Weibe function de-
scribing combustion, see Figure 2.
This function assigns the magnitude of the
time step depending on the time (s) in the
expansion.
The volume function assigns a system volume
depending upon the time in the expansion.
It uses the Weibe function.
This function describes the mass fraction
burnt by a Weibe function dependent on the
time in the expansion.
The FENERG subroutine is used to calculate
the properties of the reactants.
This subroutine is used to calculate the
equilibrium flame temperature.
The NO subroutine is used to calculate the
nitric oxide formed during a time step.
This subroutine calculates the equilibrium
concentrations of the various species in
the products of combustion.
The ENERGY subroutine determines the pro-
perties of the products of combustion. This
subroutine was developed by G. L. Borman
(1964) and can be found in his thesis.
182
-------
FO^'ATP T(«:.! HE., I "IN V4RIA^LF VOL'JMF C^.?tSTlC\ - .-'AY 11*1973')
nr'F SlO ' TIT|_=(2*)
CG'1M "M/yi'/TCE, XXA> XX :/C
'ATA C & T / 8 . / * r1 \ T / 1 ft . / > L ", T / 0 . / / F AP C / . 5 / 1 3» t V / 1 9 1 5 7 , / t ' >C F R / , / > K L U / I /
C = . 1 ? 5
R t T ' : ( ft ,(* 5 )
-EA-i,'j"R
1 5 • i: F 1 n ( 5 * 3 E A D , P ^ P « E / T « ^ F ", A
,X/.v 12
c r n '^ './/"•< \ " 1 ^
>• 4"' en-»'.'AT(' EQ» I '/At b <~.f Si T !'"'» ' t fr>,2> ' EX-iALST GAS RFCwlMlG-. "AlE=' lo
oo ^ F X P V '~ T I" - " I F = ' ,r E i 0 . . )/( ZZZ*32 ,+ZZZ*3 .7^*28 . 011) 23
CAI I FQFl" T(T?^ ^, TK^P^E, Qi-V/ F£S.» FAPC* EG«> C AT* nAT/:.1AT } 2^
CALL FFN^RG ( P/ Ti?> PUF^Ji^ (,UTK^ RR* UHV* FAS, FAPC * tGR^C AT* ^i\T*ClaT) 25
;G*-"nL('!> 27
«L«0. - ^'
F»0. 31
XN"P=0. 3i
J«l 3^
-------
CD T" 33
nT«DE|_T(t
S-S-OT
ALL»AL
AL«AIF
-------
12
13:
75
OP •(-?*'•>* VOL
P) } } / (
-OuP )
-M* ( 3U
+TP
i1=( { l .-A:
03 T '• 13"
CALL fr'f^
-JP/ F
LT,» 'JUF/ 0«P*LRT* ?RF )
ZP»(
rP«(-P*OTVDL
T p o s T P
TALL
ZP«(
*TP#DPT
/(rOT+TP*tr^ r+Rp j
)/(TVi,L-TF*DRP -ZP*< RP*TP/P-TP*D« P-I:.
nP*(R°*TP/P-TP*['Po .QLip ))/(DUT +TP*DPT +P.P)
en
CALL NH(
*KLO)
IF< IFRR.
yF«wP + (PP>P)*(
i.v'RTTP(6*?5)S/X
F(]RMATC«ri5.5)
IF(S-TGP)60>5o^5n
FTA*?.l5£-6*XMp*P*(
IERR*NfJFR
50
) /i' F
72
7-'*
9-
93
97
102
105
-------
FO*V.AT(2*A3/5E14, 9/5614. 5/2E14. 5) 106
srR!T*(6jft5>ETA 10?
FarUiATC NITRIC OX I OEA T«K * ' / E 15 . 5// / J 103
•^'UH* UP-1 109
1FM 'R> 59* 59* l^c U«-
ST'IO 111
PNr' 112
FU^rrr.M OELT(S) 113
II' f .sin- B(y) 114
5)lj?*lO 117
^)2>3*?0 ll'i
3)3,4/ 30 119
H V. lF(S-1.9r-3U,5*.*0 120
™ 4" IF' T » I -f 1 1 2 '»
•i TsT + 1 12?
4 I»T+1 125
» I«I*1 127
? 1*14-1 12-H
1 TeT+1 ]2?
"ELT-B(I) 13'
'•'ETu-\ 131
t*r- 132
-------
GO r 3o
'/HI .at
00
FFfS-TOci10,20*2-.;
£LF= i .-^yp(-XA*( (S/T".C )**(
SI] T • 30
J =
133
13*
135
13':
137
13-
1 39
U •
U2
143
14'+
145
1 4 s
147
143
149
150
151
-------
1QAT) 153
wFV«.3HV-19l»6, + ( ( ( <1.5737E-l!»*TR-5.433le-l2)*TR-4t50«2E-C8)*TR J54
1 0. 62855-0* > *TR + 3 • *7'>5 t-OZ)*TR + 15. 79 l + l5St 07 155
E-0*) 156
^57
1 5t»
15;,
16
06) J61
162
' 'JTA«( ( (-.l57r "-16* L'AT ) 171
nATA HHE/.2/ 172
CALL FEr -CBt,f P^Tw./PHe^Uw.^rL'TR/Rt jCHV^FA5,FAPC/EGR/CM/HATjiJ T ) l'73
TP»T^430"0. 175
11- CALL F"P^GY( ^,TP,PHr/^-\S/,(t / f- A^C/ XP/ 'xiP/ MT* )UF/ DKf>/ O^T, L kr ) 176
;pB"r> + wp*TP 177
: H « HI - _ ^ P 1 7 ;<
*•' IF f ATSO-)-[HE » 1 '/10/2 ) 17->
7 Tp«Tf^r).|/ ([-.nT + yp + DyT*T P ) IH-")
,n T n • m
1 ' CJ TINJE 132
•FT-j<-j lh 3
-------
CO •'•'•< A'/R! -?C/C ,^ , : >*• *T >* *XH*XG>XN>XH2*Xl>ljXCU,X. U*XL2j 187
"
•'ATA V/l.
X N ". C r- = X N.n
OCAT lql
^S-IAT 193
F.pHe l9;
7 * 197
oo « sK L
^ CALL
fF( T.NC-."-) G!"J Tn 99
A2»X !?*X'.*13.3F.9*TEE*EXP<-7080./ 215
10 CONTINUE 21*
RPT'!5M 217
-------
vo
o
C
c
C
c
c
c
c
c
c
c
c
c
C'IERI/,M OLlKAF.A
.-I r,cnns in/ MAnisnf;
SUnROUTIt-E
DfcvELUPEl* BY
UNIVERSITY H
/f-l.nc/
XlpXl?
Ol-'.E.iSI
[)ATA jF/r/
** SECTIf'N 100 CALCl'lATES TMH CnilSTAflTS 'iSRC
R0»(
K*i^O
t.l«RO*3.73H37
VgRSICM MCV 1972
! T'K f,U^n!ITi;u:
IH(K,GT.C.J>*AN) .VI T 11
I E i- • R « 1
&u TD 7jf
11C
ij2n2.0*R/A|;
11?
115*112,115
EQi-ILIEPlUf-: f 'VjS'Af.'T!: '.TT.E C'-IRVE FITTED (LFAST SQUARTS) IM
E ^OC Tf- 3«cn DE^. K (1440 TP 6«4^ DEG R) FPTll DATA I'l JAi|/.
FM!CAL TA3LFr- SECl'lp EpITlDN ( 197C ^ .
EQBM
EQDM
EQBM
EQBM
EQBM
EQBi'i
EQbH
EQCf,
EQBM
EQliM
EQBM
F.QBI;
EQDM
EQbH
EQBM
EQE^M
rQBf'.
tQBK
EQlif-i
£ Q P!-'
. EQl.I1
EQuM
EQP.K.
EQBM
EQBM
EQBM
THEE
EQBM
EQBM
EQBM
EQBM
EQt.r,
EQDM
1
2
3
4
5
6
7
n
9
10
11
12
13
15
17
2"
21
2?
23
25
26
27
Z*
31
3?,
33
-------
EQBM 35
EQBI'i 3*
J>0.448fj3'?E-?*TAS"») ?/"OP EQBM 3"
5/-.CP EQBi; 40
)/rCP ' " EQBM 4?.
EQBH 44
^C.£.->3 1.23-0. i54C33E-l-TA EQBH 45
EQBM 4*
2.611'U+C ,254^B9*TA EQBM 47
9-0.l6l547E-l*TAS'j) )*".QP EQBM 4*
i - 0 . ! 2 3 6 9 3 E -1 * T A S> ">) ! * r, ^ p c Q B M 5 f>
C EQBM 51
C ** StCTlCM ?<0 D'-CTf-rS AETHER nF, NHT Tfi ^ Ak r A iJEW ESTlMATb EQBM 52
C (-H X^*X6JX« AiNu 'Ul. EQBM 53
C EQBM 54
jF(KLfi-l) ?o^?0'v^l HQBM 5.«i
20f> 1 f; t JF.F.tf.O) CO TM ^0- EQBM 56
j h ( Pi.iI-Pv. I PI; ) :??^/ ^1^^305 EQUH 57
210 IF{ AHS(T/TP;-1.0) .GT.0.2) r,H TO 305 EQBM 5P
1^ {AK$(P/PPK-1.G) ,r,T.C.2J GO Til 305 EQBM 59
GO in uU EQBM 6>>
C EQBH 61
C ** SECTICN 3fO C.\" r.'.K.T AM INITIAL ESTIMATE f?F X4/X6/X8 AHD XU, EQBM 62
C EQBM 63
305 iFfPHI.GT.l.e) C'? TC 310 . EQBM 64
pA.^ = i,C/{R + Ai-K2 + ''''.2"*AH) EQBM 65
r,Q TC 315- EQBM 66
31C pA-sl.i:/(iU + *2 + A" + 0.r;*AM) EQBM 67
315 FU--U2.0»Av*ClO EQBM 6*
EQBM 6
-------
N)
320
QX*1.0
c
c
c
c
FOx«(FUNl*SQCJX*A'l)/,LF.l. iF-2) f.n TD 330
JN isjNu + 1
lFrnp,LF.2'J) OP TT T2"7
SQ ;X«SORT(nX)
IX)
X8.UX
Xll«Rl*PAR
** SECTK.N <*oo CALCI.-I.A-ES TME ELEMENTS OF TK MATRIX OF
EUjATIL'NS,
T 7 M a 0 •
EQBM 70
EQBM 71
EQBM 72
EQBM 73
EQBM 74
EQBM 75
EQBM 7/»
EQBM 77
EQBM 7P
EQBM 7<>
EQhM 80
EQBH 81
EQBH 82
EQBM 8*
EQL.fi 84
EQW'I 87
EQHM 8»l
EQBM 8<»
EQhM 9o
EQBM 91
EQBM 9?
EQbM
EQbK
EQBM 97
~ISEDEQbM 9b
EQftl* 9"
EQBM100
EQBH101
EQBM02
EQBM103
EQpl',104
-------
vo
U)
•»55 CU T Js ,{.;
5 0 X 4 « i ; & T ( X
50,V EQBM135
EQBM136
l.(HT106) EQBK137
-------
vo
C
C
C
C
A<3/1)*0.0
AO/2)»T76-D3*(1.0+T106)
A<3/3)*T78-P3*T108
A ( • 1 .0+T7&+T l'">6+';4* < i . 0+T 106 )
8(4)
** SECTK-N 500 S'lLVET, "HH MATRIX
jSlNG MAXI^uv. PT/nr rPiNT STp,ATEOY.
DO 505 K»l,3
IF(8IG.GE.1.C'E-J'M G" TC 5?^
JBIG«K
DO 510 I-KP1/4
lF(APS{A(I/K))tLr,bI-J ^ TH 510
510 CU .TINuE
IF(BIG.GF.1.0E-1TO G ' TO 512
ItKR*3
&U TO 710
512 iF(lBjG.E3.K) GO Tn ^-20
DU 515 J«K^4
Tfc;-.P.A(K,j)
A(K.» J)«A( I?IO/ J)
515 CU "
EQBM140
'EQBKUl
EQEMH?.
EQBK143
EQBM144
EQBM145
CAI.SSIA'1 ELlMINATinf;
EGBf',132
EQ&M5?
EQB?-:15-J
I.QF,i,l59
£QBMl6i»
[".Qt>!;l6l
EQt.f.162
FQfiMl63
EQBM165
EQB'!l6fi
EQUH167
EQF.M6
EQHIU71
EQF»!-!17?,
-------
T6-.P»B«B(IBIG) EQBf',176
R( lBlG)=T£v'p EQBi:177
!>20 00 525 I*KPl>4
nU 530
A I ! * 0 )*A( I, j )-A('> j )*T"R;l EQbHiai.
530 CU"TINuE EQBM182
R( ! )*8< I )-B(K)*rrK!| EQBM183
525 CJ :T'lN'v)E EQbHl84
505 c°''TlNUE EQBM18.3
IK A3S;(AbS(Sl/X4 ) ,GT.i..-E-5) 'ICK.1 .EQBM199
EQBK200
),CT,l.:-'E-3) MCK-l EQBM201
EQBK20?.
IF(A8S(S3/X8 ).GT.l.nE-5) MCK.l EQBM203
EQ8H204
'ICK*1 EQBM205
IF((X4tGT.OiO),At!D.(X8.CT,0.0).AMO.
-------
IF (INC.IT.25) GO Ttl r.22 EQhH2lo
GO TO 710 EQBM212
622 lNn*IND+l EQHK213
GO TD 455 EQBH214
625 lF(X6.GE,0.fl) GLJ Tfi T? EQBM213
GO TD 71C EQBI',217
C • EQ£!i2n
C ** SECTION 700 CALCl'LA'FS TUf. f1PLrRACTI H! ,5 /i',;D PETUR'lS WITH KRR«OEQk.M21'>
C TO THE CALLING Pm&F \." nr. IF Afl ANSWER "ASNCT rFTAINfO RETURN1". F.Qfif.ZZO
C wlTH IERR»(T^E ERPnP CPDD EQI.M221
C
70? IE«R-0
jF.l EQF>!:224
PHiPK*PHIf
H TPi «T
« PP; =P
SCRT(X4) EQbM22it
EQbh231
X2«C2*5QX8 EQBM232
X3»C3*SOXU EQBN233
EQDM23?
fIQE!,23'>
EQr>'-:237
EQF.I',23"
715 X"7
GO TQ 72?
720 X7.A!iOH*Xl3 EGLK242
725 KfcTUKN PQKl'24:)
7in jFB0 cQt!-i24<«
RfeTJRN EQhf!24S
-------
APPENDIX C
TWO SYSTEM COMPUTER MODEL
The following set of equations can be written for
the two system model having three regions, lean reactants,
rich reactants and combined products as seen in Fig. Cl.
Conservation of Energy:
mLRULR = ~ PVLR ~ hLRmLP
mRRURR = ~ PVRR
mpup = - pVp
Conservation of Mass:
mT =
Conservation of Volume:
VT * VLR + VRR + VP
Ideal Gas Relationships:
PVLR = mLRRLRTLR
= mRRRRRTRR
pV = m R T
P P P P
197
-------
RICH REACTANTS
PRODUCTS
Temperature
Mass
Volume
Internal Energy (u,,)
Gas Constant (Kp)
Enthalpy (hp)
Fuel to Air Equivalence Ratio (F' )
LEAN REACTANTS
Temperature C?LR)
Mass (m or &,)
LK L
Volume (VrD or y )
•"•" LR
Exhaust Gas Recirculation
Fuel to Air Equivalence Ratio of
The Recirculated Exhaust Gas (F,
Internal Energy (HLR)
Enthalpy
Fuel to Air equivalence Ratio of
the Fuel and Air (FT)
SYSTEM
Pressure((p)
Volume (V~ or v~)
j. j. • •
Rate of Change of Volume (V- or v )
• J. JL
Rate of Combustion (m^ or b. )
Fraction of the System which is Product Co, ;
Fraction of the System which is Lean Reactants ($
Fraction of the System which is Rich Reactants ($
Figure Cl
Two System Thermodynamic Model With Three Regions
198
-------
The internal energies and gas constants will be handled
as discussed in the one system model.
The rate equations for the two system model having
three regions are as follows.
199
-------
p =
V
3u
( — ^
1
R
*
F
(Tp Tf~ + Rp)
Rp)
VT '
LR
RR
- a
R T
m *~ »
TP ap +
P 3T
V
*RRTRR
200
-------
TP "
4- T *- + R
* JP 821 P
Where:
-------
Main Computer Program: The two system model consist
of a main program and eight subprograms. The main pro-
gram keeps track of combustion, expansion, and mixing
processes; It also handles the program input and out-
puts. The subprograms are called as needed to perform
their specialized tasks. Listed below are the input
terms as an aid to understanding the program:
Computer Program Inputs:
NUR:
TITLE:
TCM:
GAM:
CL:
THDO:
RPM:
C:
XLR:
The number of different sets of engine
operating conditions to be calculated.
A one line title for each set of calcula-
tions, such as run number.
The time of combination of the two product
systems (seconds).
The initial fraction of the mass in the
lean system.
The time between the initiation of combus-
tion of the two systems (seconds).
Crank angle at which combustion and the
computer program begins (degrees).
The engine speed (revolutions per minute).
Clearance volume divided by the displaced
volume.
The length of the connecting rod divided by
the radius of the crank shaft.
202
-------
TOCL:
TOCR:
XML:
XMR:
XAL:
XAR:
PL:
PR:
FL:
FR:
TLR:
TRR:
EGRL:
EGRR:
PEL:
Time of combustion for the lean system
(seconds).
Time of combustion for the rich system
(seconds).
Shape parameter "m" of the combustion func-
tion for the lean system.
Shape parameter "m" of the combustion func-
tion for the rich system.
Shape parameter "a" of the combustion func-
tion for the lean system.
Shape parameter "a" of the combustion func-
tion for the rich system.
Initial pressure of the lean system (psia).
Initial pressure of the rich system (psia).
Equivalence ratio of the lean system.
Equivalence ratio of the rich system.
Initial temperature of the lean system re-
actants (degrees Rankine).
Initial temperature of the rich system re-
actants (degrees Rankine).
Mass fraction of recirculated exhaust gases
in the lean system.
Mass fraction of recirculated exhaust gases
in the rich system.
The equivalence ratio of the recirculated
exhaust gases in the lean system.
203
-------
PER:
Subroutine
RPT ( ):
Subroutine
(PRDMIX( ):
Subroutine
PRESEQ ( ):
Subroutine
EQFLT ( ):
The equivalence ratio of the recirculated
exhaust gases in the rich system.
The RPT subroutine is used to calculate the
reactant to products process for a time
step. This subroutine is capable of handling
two regions of reactants and one region of
products. The output of RPT is the new
properties at the end of the time step.
By setting the fraction of one of the re-
actant regions to zero the subroutine will
handle two regions, one product region and
one reactant region.
The PRDMIX subroutine is used at the time
of combination when the two product regions
are combined into a single one. This sub-
routine determines the properties of the
combined product region.
After the products are combined the pres-
sures in the two reactant regions and the
combined product region are probable dif-
ferent. PRESEQ subroutine adjust the
volumes of each region until they all have
the same pressures. The regions expand or
contract isentropically.
EQFLT is used to calculate the equilibrium
flame temperature.
204
-------
Subroutine
NO ( ):
Subroutine
PRD ( ):
Subroutine
RENERG C )
Function
BELT ( ):
Subroutine
EQBM:
Subroutine
ENERGY:
The NO subroutine is used to calculate the
nitric oxide formed during a time step.
The PRD subroutine calculates the change
in properties for a time step when com-
bustion is completed and only products re-
main.
This subroutine calculates the properties
of the reactants.
This function assigns the magnitude of the
time step depending on the time in the ex-
pansion.
The EQBM subroutine calculates the equili-
brium concentration of the various species
in the products of combustion. EQBM is
shown in Appendix B. This subroutine was
developed by Cherian Olikara.
The ENERGY subroutine determines the pro-
perties of the products of combustion.
This subroutine was developed by G. L.
Borman (1964) and can be found in his
thesis.
The computer program follows.
205
-------
DIMENSION TIHE(24)
DATA CAT/8./ jHAT/i 6t //OAT/0. /*FAPC/»3//QM/l9l37.//NnFR/0//K|.0/ 1/
45 FORMATC STRATIFIED CHARGE ENGINE SIMULATION - JAM E2jl97V/)
WRITE (6/45)
ftEAD,NUR
150 READ(5,35)TITLE
*RtT-E(6,35>TlT|.E
35 PQHMAT(24A3)
READ, TQM, GAM, Cl
READ, TOCl, XML, XAL
15 pa«MAT(« .RPM«I,FB.O, ' TI'ME WHEN MIXING ,
l« CUMBUSTION LAG*',R12.5, ' MASS FRACTIDN CF LEAN«',F3.3)
25 FORMAT{ i LEAN-EQ''IV F.ATI H« ' , FS , 2, ' EXHAUST GAS fiEClR*1 ,P5.3/
p EXHAUST G^S EQl'IV RATJHa ', F5.E,
2i COMBUSTION - TIMr«'/E10.4, ' XML*1 ,FA.l,' XAL-'fF*,!)
85 FOUMATC RICH-EQ'.'IV "ATin« ' ,F5 .2, ' EXHAUST GAS F;EC ZR- ' ,F5.3,
li EXHAU«,T GAS Ent'Iv RATIO*1, F5. 2,
2' CDMBuSTK'N - yIMP* ' , ^10.4, ' XMP.
wKiTE(6,55>
55 FORMAT!' TIME CRANK ANGLE VOLLME NQ-PPM
1 TFMP PROD PnE n*r> TEf'.P REACT PRESSLRE FRACTION BURNED
2
27 ZZ
THFsTHD*. 017453
If-(THE>lb,l7,l6
-------
17 VG.C
GU Tu 18
16 VG «C+.5*COS(TrtE)/2.+XLR*(l.-SQRT
-------
00 TO 400
S«S+OT
017453
,5-COS(THE5/2.+XlR***2
TLPP.UP
GU TO (215
213 ALiSS»AlL$
) IL»2
g CALL RTP31*>275.»3 >0 5 / JJ
275 jj«2
GO TD 202
-------
2Q3 IF J/2.
pRRsPR
TR?P»
GO TO
.^^
EGRRjFp^*TI-.'RR/DTVR/FAS/FAPc*QHV/Fr'WT/CT)
GO TQ 335
-------
311 ALRS.l,
AtRSSai,
RMflR«0.
TRR«0.
IR.3
310 CALL PRD(PR,TRP,FRP,r>TyR,TVRp,,FAS,FAPC/DT>
335 xNoPRR-XNOPR
IF(FRP LT.l £-6) GO TO 342
CALL NQR/TRPP/TRP/FRP/ALPSS,ALRS*OT/CAT/HAT*QAT
lF(IERR,Ng,0) GO TO 50
342
301 rfR!TE
CO TO
3Q3 I»-(S-ABS(
10 304 JJ«3
g G° T0 200
3.Q5 PspR
FP.FRP
TP»TRP
F.O.
CCl'CCCL
vGn=vG
TVDL»GAM*VDLS+(l.«-GA':)*vnRS
VO.TVDL
GO TQ 401
400 C^lL
4Q5 CALL PRESeC
j^EGRR/FEP^
VGn=VG
-------
ALP»ALl,S*GA'"UALRS*U.-r,AM)
+(l.-GA'M* 'Fp.
75 FURMAT(« THE PfmuCTS ARE MIXED AT THIS
OU TO
402 orsDE
VG «C*.5-COS(THE)/2.+XLR*(l.-SQRT(l.-( (SlMfHE ) /XLR.'**2
TVnLL»TVCL
GO TO U42
ALLSS»ALLS
IF-(S-CCL.G(=,TQCL) 11*2
450 GO TO (452/455/455)^ TR
452 ALRSS
IR=2
455 ALoPsALP
ALo«ALLS*GAM*ALRS*(l.-GAM)
FPP*FP
FPAAs (FLP*ALLS*GAH/{l.*FLP*FAS)*FRP*ALRS*(l.-GAH)/Ut*FRP*f:AS))
I/ALP
FP«FPAA/(1.-FA5*FPAA)
/2.
/DT
-------
CAt.1
l,F
420 GO
ALtSS-1,
RMRL«0.
TLftsO.
1L.3
412 GU TO (417/413, 4'17),
ALt>SS«lf
TRR=0.
1R.3
417 lF(lU-3
M 48" 1H(IR-3)'435*481/481
10 481 ILa4
AFP=FP
ALPP-1.0
00 TO 435
410 CALL P*D(P/Tp ,FP/nTvCL/TVnLL>FASjFAPC/PT)
INOpR/KLD)
JF( IERR.N:E.n> GQ TO 50
401 WR!TE
GU TO
95
80 |HTHD"l
50 CONTINUE
CALL ENERCY(P*TP*FP/FAS/!'P/FAPC/RP/DUP/OUT/C')F/CF.P/DRT/DRF)
-------
UR«GAM*ULRB+( 1 ,-" AM > *URRP.
'JB.UP-UR + WF
VB.VPF-VPF
rtR I TE ( 6/ 1 05 > Up/ UP./ Wp , UB
105 FORMATC ENERGY BAI.AMCE - UP«'^E12.5/» UR««/E12,5/
l« WGRK««,E12.5,,» S'JM»'/E12.5)
H >jRlTE(6/106)vPF/yRF/\'B
w 106 FORMATC SPECIFIC /aiUME BALANCE AT fMAL COriClTIdMS- VP«f
•«RlT£(6, IQTNC,^, , fK. „.,., ,
io7 FO«HATC PERFQK.-IANCE PARAMETERS - ESTIMATED COMPRESSION WDRK-'
l*Ei2,5/« ENTHALPY f-Ff-l EI EMCY« ' > E12.5/ "" " "
Z' ISNO«'*Ei2.5//)
FDRMAT(5E14,5>
I»-(NURJ59, 59/150
59 STrjP
END
SUBROUTINE RTP(P ,GA'i /TLR/TRR,TP*FL^FR'FpP'DFP/ ALLSS/ALRSS/RHBL
JF(GAM,LE,1,E"»4) FFL»0,
IF{GAM.GE. .9999) FFR=0.
CALL ENERGY (P^TP^FPP, FAS, UP, FAPC/RP/DUP, OUT, DUF,DRP/DRT,DRF)
-------
CALL RENER0(p/TRR/rR/Er>Kr>/FErs/URR,DUTRR,RPR,QHv/FA5/FApC/FMWT>
ALP«ALl.S5*CAh+ALRSS*(l.-'5AM)
ZURlR*TlR/ )
ZR»RRR*THR/(r*(D'lTRF;4-PpR) 5
l.~AUSS)*GA':*P.U
l.-GA :.)*RP,R
1 t-GAM)* (PkR*TKR-RP*TP)-P*DTVOL
np! -HUD
( l.-GAf:)*(HF.R-HF
10 [V?
GO TO 30
JO D"f>=F' ) /F
IQ TR«R*TRR
TL*R«TLR
t)I».RR*DTFR
i/l L«R*r>Tl.R
[)TPP»DTP
TL?*TLR*PTLR*OT
TP«TP*DTP»DT
-------
TV ,L =
P*( <1.-ALLS)*GAM*RLR*TLR+<1.-ALRS")*{1.-GAM*F.PR*TRR*AIP*RP*TP)/
ITV.--L
CALL ENERGY P.PjDi.^,0!JT,CUF,CKP,ORT,DRF)
ZRsRRR*TPR/
AAai ( TP*DRT +
RLa(l,*ALLS
BR«(l.-ALRS)*( l.-
C*t*NIBL*GAM*(nLK-!lP)*j MBR*( 1 .-GAM ) * ( HRR-HP )
) / ( G-ZL*Bl-ZF.*«P-0*AAj
TPafPP*( DTP + DTPP )*DT/?..
P-( (l.-AlLS)*GAM*RLR*TLR-»-(l.-ALRS)*(l.-GAK)*f;RR*TRR*ALP*RP*TP)/
IT VOL
RETURN
SU9ROUTINE
GAr;#FAS*FA
CALL ENERGY < PL, TLP>Fl.P, FAS jiJLP/FAPC/RLP* DIP, OUT, DUFjORP/ORTjORF)
-------
P/ FAS/Up,P> FAPC'RRP/ t>LP/DUT/
J
BR«ALRS*< i. -CAM) /ALP
vP«Bi.*K[.p*TLP>rL* FL /TP.P/pR/pR /ECRL/FEL
/eo^K/FEP^ALLb^ALf'f,, -,A'V<,TVnL,FAS/FAPC/QHV/ FM'T)
CALL E:iEf.CY(PS/Ti1S/Fn^rAS/l!Pp>^FAnc^P-P / O'JF^C! 'T, COF/ D»;P/ DRT/ OKr )
-------
,-f,A '
TUS-TLR
TK'iSaTkR
NJ D - " S
_i r ~ K -S
TL:.aTLi-.S*( (^^KLRr./rri.*rL" ) )#*( (XKL-1. ) /X.r.U )
TR«STRP.5*( (P*K>-<': /( rP.^RF,^.) )**( (XKR-l. )/XKR)
CALL REMFRCd1 /T!.p,Fi ^G~lj F Ll; ' UR> D< 'TI.R, P I,P, QI!W FAS,
CAl.L PENFRGtt' /Trr^F' ^ W.j FrR/UF-.^^ DMTFR.^PRP.. CMV/PAS, FAPC/
ER.' '.•/«CL*L!LR-»-CR*U" p-!r A
P-PS-10.
30 TL--' = TLRS*( (P*RL»"/ ( r..*P Lr ) }**( (XKL-1. ) /Xf.L ) )
TRp.*TRRS+( (t;***3.Pr> /('•PrtBr1' ) )**( (XKR-l. J/XKR) /
TPs ( P*TVOL-CL*^ L' ' *TL" -CR*R.RR*TRR ) / ( RP*AUP )
CA|,L RENFKG(>'*TLP ^rL
CALL RENERG(P*TKl- /Fr.
CALL E.NEKCV(P*TP*FP/
20 PPP=P
p = p-ER
pPaPPP
-------
GU TO 30
10 CONTINUE
RETURN
NE EtFLTdP/P/TR/PJlE/QUV/FAS/FAPC/ECH/FC/FMWT)
DATA DHE/.2/
TP.TR+3000.
110 CALL EUE*GY
-------
XN,,CCaXNi-PP*»'P/ ( 1 ,F + ' *14
CA( L ECU''
IE' JRP, HRT> [)RF )
TV •L =
i5.r -DUP ))/(DUT +TF*DPT +KPJ
> FAS/ -ip/ FApORP* DyP^ Dl,T-» LUF^ DRP* DRT, DRp
-ZP*(P
))/(DUT
-------
N) 10
RETURN
ENn
I.E Wf-NrROf P, *B, P'lE, EfJR/FR/UR/DllTP. *RP,Ql!V, FAS,
L)H*HFV-RF*Tr
wfel JRN
&N;;
FU -CTIjNi DF-tf (!»
1=1
iFf S.CE.f .9VF-3) 1*1*1
-------
APPENDIX D
DEFINITIONS OF THE DEFORMANCE PARAMETERS
USED IN CHAPTER II
Indicate Specific
Nitric Oxide
grams of NO^g/lb 2544 . 4 (Btu/hr-hr
m
The work required for exhaust and intake is assumed to be
zero. (i _i,~\
L *J
Compression =
Work
x 7y8 — = T2 x F*
CR = compression ratio
*The term F was evaluated slightly differently
for the one system and two system models.
Indicated Mean
Effective Pressure
.-ft ~lb->
Btu
,AA
144
Indicated
Ethalpy
Heat
°Comb.
fw -w •*
v exp Gomp)
Btu
^
J
total rmx
(1-EGR)
BTU
mx
ar
221
-------
APPENDIX E
AIR INJECTION FLOW RATE MEASUREMENT
The flow rate of high pressure air going to the air
injection valve is measured by means of a flow restric-
tion. The flow within the flow restriction is turbulent
and is assumed to be isothermal. The heavy brass walls
of the flow restriction and its very small cross sectional
area will assist in making the flow more isothermal. A
computer program was written which calculates the flow
rate as a function of the inlet pressure, inlet tempera-
ture and the exit pressure. This computer program is
incorporated into the data reductions computer programs.
When the computer program was compared to the cali-
bration data it was found that the friction factor equa-
tion had to be adjusted in order to obtain a fit. Shown
in Figure El is a plot of the actual and calculated exit
pressure of the flow restriction versus the actual
pressure drop across the flow restriction. Only a few
points fall outside of an error of two psi. The smallest
divisions on the pressure gage is one psi.
The subroutine called FRI calculates the flow rate
from the measured inlet temperature, inlet pressure and
outlet pressure. The program uses a trial and error
method to determine the flow necessary to have a calculated
222
-------
outlet pressure equal to the measured outlet pressure.
Many of the equations used were obtained from Chapman
(1971), the friction factor equivalence were obtained
from Kays (1966).
The subroutine FRI follows.
223
-------
SUBROUTINE FRI ( PATM ,TQ« Po»PBO»FR )
DATA GC/32.l739/»R/53,36/»XK/1.399/.VES/12.^E-6/
l»RH/4.066E-4/»XL/3.92/tA/1.112E-5/tZ/.105/»Y/.25/»Q/.2/
TO*TO+459.69
PO=(PO+PATM)*l44,
Pa0= ( PBQ+PA TM ) *14.4»
FRR=0.
PB = PO
60 FRA=FR/A
CON=PO*SQRT(XK*GC/{R*TO)>
XMArFRA/CON
20 FRAA=COM*XMA/«l.-»-(XK-l.)*XMA*XMA/2.}**(
EER=(FRA-FRAA) /FRA
IF(ABS(EER)-1 •E-4)30»30«10
10 XMA^IFRA-FRAAUXMA/FRAA-^XMA
w GO TO 20
10 30 VFA=1,+(XK-1 . )*XMA«XMA/2.
*" TA = TO/Vf A
PA=PO/(VFA**(XK/(X<-1. ) ) )
CA=SQRT(X«#R#TA#GC)
REA=VA#4.*RH*PA/(R»TA*VES)
IF(REA-3.E4)3l*31»32
32 E=Z*t <3.E4)**
-------
8.Q XMB = XMB + (DFB-DFBB 1* L.XMA-XMB1X (DFA-DFRBl
GO TO 40
50 PBB=PB
PB=PA*XMA/XMB
ERR=PB-PBO
IF«ABS(ERR)-.1)55»55»70
70 FRS=FR
100 FR=FR-ER**(FR-FRR)/(PB-PBB)
FRR=FRS
GO TO 60
61 FR=.8*A*CA*PBO /(R#TA)
PRINT»FR
55 CONTINUE
RETURN
END
to
NJ
U1
-------
to
+10
to
Pu
-p
o
CVJ
0,
CL
-10
o
rH
nJ
O
•k
CVJ
-20
- Q
O w
o o
o
o
O
O
o
o
-30
-L
J I L
20
Figure El
60 100
Pressure Drop = P2factual - PI (
200
400
Comparison of Measured and Calculated Flow Restriction
Pressures for the Same Flow Rate
600
-------
APPENDIX F
MECHANICAL DESIGN OF AIR INJECTION VALVE
The air injection valve is a spring loaded and is
activated by a cam engaging the cam follower. A sketch
of the air injection valve is shown in Figure Fl. The
cam follower is mounted in a holder which can slide back
and forth to open and close the valve. The spring is
used to keep the valve closed. Downstream of the valve
seat is a flow restriction which prevents back flow of
high pressure and temperature products of combustion.
The large diameter nozzle is shown in the sketch, it
screws into the engine.
The cam is mounted on a disc that is connected to a
shaft which in turn is mounted on two bearings. The
drive sprocket is keyed to the shaft. The shaft is
perpendicular to the center line of the valve steam.
With the valve closed, the center line of the shaft is
five inches from the center line of the cam follower.
A computerized milling machine was used to make the
cam surface. An eight-power polynomial as described by
Mabie et al. (1958) is used for the motion of the cam.
A computer program is used to calculate the shape of the
cam surface. The computer program for calculating the
cam surface follows.
227
-------
Pressure Air
to
NJ
00
Section A-A
F1°W / K, 1
Restriction Nozzle
Figure Fl A Sketch of the
Air Injection Valve
-------
A second computer program is necessary to convert
the shape of the cam to punched tape instruction for
the milling machine. M. Fugelso very capably made the
punch tape and operated the milling machine.
229
-------
to
CO
o
XL*, u
K ; j * i> t
K = . 6 2 >
isK«t*l
J.J U> i
/< 5' ( '
« ^ ' * • " - * b
A, * 6, 'w/y
0 Ts I *. J /C
7 T
r *Slu( -M ) •' :pr---
n £ I = .i t T A '•' i ' 0 • / 3 , i '* • A'
(JAcoA.i';!'l ,,~ . • , . ; ^l'^
PRi'-il/oB/Kr M y ^ K ^ ^ T ( . P .-A ,.
'
? fr.
.-^R P*C 'IS ! "A
E T •>">"
2
3
4
s
A
7
n
(s
O
If
11
12
13
14
15
16
17
i£
19
zr*
21
22
23
26
27
2?
29
-------
APPENDIX G
PERFORMANCE OF AIR INJECTION VALVE
A computer program was written which predicts the
flow rate through the nozzle for various angular posi-
tions of the cam. The program also calculates the dynam-
ic forces resulting from the cam engaging the cam follower
The flow is assumed to be isentropic. The flow is not
allowed to be supersonic because neither the valve seat
or the air injection nozzle have a converging-diverging
shape.
The flow path through the valve starts with an
isentropic compression to the valve opening area, then
a sudden enlargement to the inside diameter of the valve
seat and followed by an isentropic compression to the
diameter of the nozzle. The subroutine ISENT handles
isentropic compressions, the STEP subroutine handles
sudden steps in the diameter and when the nozzle is near
choked flow it is necessary to reverse the calculations
of the step which is done by PETS.
The air injection valve performance computer pro-
gram follows.
231
-------
DATA GC/32,l739//R/5%36/>XK/l,309//
1Q/3 • 1*159/ , QG/.01T4MZ //'•'/ 1. //SK/40.//FU/ IOC./
READ/NUR
PRINT, DC
RE AD/ pH, TO
NJ AL« -tt
C» 166667
TOBF. '.1+459, 69
PQ.P 1*144,
SQ" IH/(6
V'G«C + ,5-
PE«PC'C*(
PE«PE*14
T=TH/B
lF(T-l,)
12 T=M3SlT-
11 CO-TINUE
1-571, 6053* (T**5j +143. 4132* (T**6)
-------
-i3» 60965*7, *(T**6)+2.5^°5*8*(T**7) )
RRnSi-D*XL
XL.XL/IZ,
XD«XL*D
38 AA
GO TH 37
39 &A
37 PQE»PU/PE
AC«Q*OC*DC/4.
SAB^AA/Ak
w RBC«AB/AC
<*> RAc3RAB*pBc
XM«SQRT((->1. + (PUE**(UK-1.)/XK) ) ) *2 , / ( XK-1
Ip( U-XM)^l*42/42
^l XMtl.O
15 XMA*XM
GO TO 17
16 XMA«XM/RAC
17 £R^=PUE
XMAA«0,
95 IF
22 CALt
GO TD 23
21 CALL 5TEP(XMAjRAR/XMr.jpA&/tAr.,iLA)
23 CALL
91 «BA*1./RAB
93 CALL PETS(XMAjRA3/X'!8,PA3/TAB/LA)
-------
97
92
?6
90
IF(LA-J )90/SO/97
CALL 5TEP(XMA/RA'*-/Xf,.-/PAp;/TA-/LA)
CALL I5EM{XMB,R».C,X':C,P"C,T-C,Lr, )
GQ rn 90
CALL ISEf'TUKBjRr.A/X'lA/Pr.A/T-A/LA)
TA .
-1, )*X'iA«X,"A/?, )**(XK/(XI>1.
'A/Z,
65
00 Tl 95
XMS«XMA
03 T-; 95
76 DRlNT,ERfi,pnE/PUr
75 CO .«PU*SORT(XK+Gf
.* ( Xl-'-l
. ) **(
I ")/TUA
lt-XMA>77/77/73
-------
77 PB-PE
TB«0,
PC-0,
TC«0,
GQ TiJ 79
78 PB«P4/PA'r
PC-P8/P9C
TCaFB/TBC
TB.TB-^59.69
TC.TC-459,69
PR INT, XD/AA,RA3jr.SC/
PRjNTjFRA,FRBjFRC
PRINT, SU
PRlNTjPQjPAjPB/PCjPE
PRlNTjTQjTAjTBjTC
70 STOP
END
79 FCKSFU + SK*XD + 3.*:>uP '*FP !*ACC
PC <•-. "S I N ( A L > *F C* / T ' 15 ( ;,l )
to FC-.»FCK/Cr)5(AL)
h{ aLLL«AL/'*Q
-------
SUP ROUTINE ISENTfXMA/RAB/XMR/PAH/TAB/L)
DATA XK/1,399/
l"l
RAS»<( U,«-(XK-lt)*XM.\*XMA/2.)/((XK+l.}/2t) )**( (XK+1,
U/XMA
&1
CB«XMB/( (i.+txk-i. )*xfn*x:in/r.. >**(
CA«CXMA*P.AB)/( { I ,*(XK-l, J *XMA*XMA/2. )**( (XK+1. ) / ( 2 , *( XK-1 . ) j ) )
51 XM.S«XMA
lF(bRK"fcPRfi 5 1QO/43/KC
XMAA-XMS
GO 1:1 31
CA.(XMA#RAB>/( ( l. + (X'r.-l. )*XMA*X'1A/2. }#x<( (XK-»lt ) / ( 2 1 * < XK-1
30
50 XMS«XMB
XMP«XMB-fcRR*(Xr'B«X'!F" ) / (Ff.P-r.RRR )
XMM'-'XMS
GO T'J 30
-------
to
70 CONTINUE
60
SUBRQUTlf E $TEPCXMA,RAB,X:iP/PAfJjTA8,L)
DATA XK/1.399/
L«l»
62 !F(X",A-lt ) 60, 61,61
«sl X.Mi = l.
L-2
CA»
XMV>*1,
)*
?0
101
102
90
CB«XMA*SCRT(
F.RKK«EKR
6RR=(CA-CB)/CA
IFtABS(EF)-.
EFaEF/2,
GO TO 102
XMRB«XMS
*X'1.A*Xf-',A/2 ,
XK-
^ ) *XI1B
-------
GO TO 30
43 PRlNT
40 FBA«(
PAB"XMB*5QRT(FBA>/(X:iA*pA^)
RETURN
EN'1'
SUf»RCiUTlf-E PETS ( Xf ;A, RAB/XMR, PAD/ TAB
•>ATA XK/1,399/
50
101 EF«"ER«*{XMA»X'MAA}/( f EPR-EP RP. ) *X 1A )
102 lF(AaS(EF)-.l )10^/1P"/00
90 EFnfcF/2.
GO TD 102
*Xi A
00 TO 30
43 PRlNT/ERf
-------
: /(X"A*RAB)
to
U)
vo
-------
APPENDIX H
DATA REDUCTION COMPUTER PROGRAMS
The data obtained from each operating condition (run)
is punched on computer cards to expedite the calculations.
Two computer programs are available one with a short out-
put in which the important results are on one line, the
other longer output provides more detail. One important
advantage of using a computer program to reduce the data
is that, if changes are desired in the calculations, all
of the calculations can easily be repeated.
The long output and then the short output computer
programs follow. Note that the subroutine FRI called
by the programs is included in Appendix E.
240
-------
APPENDIX I
CONVERSIONS TO SI UNITS
Term
Length
Force
Mass
Temperature
Energy
Energy
Power
Pressure
IMEP
ISNO
CFM
RPM
Units Used
In Report
inches
lbf
Ib
m
°R
BTU
hp-hr
hp
psi
II
g/ihp-hr
ft3/min.
rev . /min .
To Obtain
SI Units
x 25.4
x 4.448
T 2.2
T 1.8
x 1055.
x 2684519.
x 745.7
x 6895.
11
x .0003725
x. 0004719
* 60.
SI Units
millimeters (mm)
Newton (N )
Kilogram (kg)
Deg. Kelin (°K)
Joule (J)
Joule or Watt-sec
Watt (W)
Pascal (Pa) (N/m2)
II
kg/J or kg/W-sec.
m /sec.
Hertz (Hz) (rev. /sec.)
241
-------
55
DIMENSION' TITLE<23)
DATA K/53.36//TAM59.6(V/'-J/.fH40° //XK/ 1 .3*<>/> VC/37t 33/
1/FAS/.06628/
dRlTE(6M3)
45 FOsMAT
15'0 *EAD F7 . n, ' J-: HG ^4 A* '/ F7, !/ '
- :C= S F7 .r(/ i nr i-| <•;-;= « / F7, - ,, ' r
C- 1 ^= ' > E
185 FOM-IAT{' RMReljfT^.B., i T13s ' , p7
-------
IF(FRIA)U, 11/12
12 CALL FPI
75 FQHMATP MASS FLPVI R'.TF- FUEL* ' f F9 . "», ' US /S "C C AF-BMRfcTrn AIR*
IjF'i.S1 INJECTED Ap.= ',r\6)
FRTA«FKCA+FKIA
i»lT«FRIA/FRTA
ftCT*FRCA/FRTA
:vRlTE(6,85)RCT,RIT
^5 FORMAT (' FRACTlLl1' IF AIR C Ap-r 'JRETED* '> F7 . 2, ' F^CT
ICTEO««/F7,5)
£Ru=FKF/ (FRTA*FA5 J
w 95 FQ^MATC EQUIVALENCE RATP- ClVF.f-ALL. '/ F7.4/ ' C A" nUr.ETFn- f , F7 .
./ (C*S)
195 FO.MAT(i
? AT'! ) / ( n2 + PAT M ) ) ** < /XK) ) «
XK -l.)
HPlC-HPI-HPC
. <
105 FQxMATC HQRSPOW^K- GROSS BP AKE= ' / F7,4/ ' ^ EST JfjJ AJR Cr
1F7.4,« NET H -
-------
EEJC-HPIOA
'/,RirE(6,U5)EEB/EERC,EI:I/E
,15 FQkMAT(i ENTHALPY EF^ 1C I F'lC Y- Gr.QSS 3RAKE-* ' , F6.4/ ' "E
1,F6.4/' GRCSS r!iIC.\TED-',F6.4, ' NET I' ,C K ATHC = ' / H> . 4
XT fcP«iiPJ*&
125 pOxMATP MFAr- EppECTTyr n^CSSuP.F- G^.'-.SS ETJKEs '/ p7
H iMf-r BKAKE. '/?•», s/ • c-!>1PRrssrv'U'/r7.?/' cerss
2' NET r-n«SF7.?.)
X*E/HPS
BCSNX»XNL'X*E/MP8C
XirsX«XWriX*E/HPlC
'•'R I T E ( 6 , 155 ) PS^UX/ F.C c-f !X/ X I SMX/ X I C SX
55 FQ^MA P SPECIFIC MPX- GRPSS RR AKE= • , 1 12 , P, ' G'l/ ( ' !P-. I )
i.'' r^nS55 ! "^= ' , r 12 . 5, ' NFT I ,J'J= ' , E
. r 12 .
F-. 01*28, 011*13
BCSCOeCG*F/HpBC
XI.sCi!*CU*F/HPI
-,u.,
FQ.MATl' SPECIFIC C.P- C^SS BRAKE = ' , E 1 2 . 5, ' MV (
li NF.T BKAKE.SE12.5,' ^"S^ I N.r,, • , r 12 , ?, ' NPT
'SsKMrv*0.u566'?*(Tl3-Tl4)
(i ;>- ; -
JMi)« '
HTC A'Ss
-------
T;J
ST P
CI b' SID' TITLE(24)
/ VC /37.
l*FAS/.06t-29/
" RI re (o> ^5)
FORMAT (//? s'jiWAKY 'P Rrs'iLTs -F'//J
>/RlTE(6*201)
FQhMATf' RUf ER"* r"C Fi.CA Si-CA SP-(.A rPM MP Ct'P
1 IREG IEEN IMEPO IMEP'I '^X-PPM '^HX-Sr, C^.pFR CP-SC .'1C-PPM
2')
35
«
P E A U / P A T f • 1 / P 2 * P 4 A t v A B * P 4 £
12 CALL
-------
fcRri«FRF/(FRTA*FA?)
£RC«FRF/(FRCA*FAS)
Kp:i«XN*60,/(C*S)
HP1«HPB+HPM
EEJ«HPl*A
EElC»HPIC*A
*31')nn» /114
fe*n
X*b/HPI
20J FU
13/F7.0/)
150
ST.:^
EN i.'
-------
TECHNICAL REPORT DATA
{Please read Inuniclions on the reverse be/ore completing)
i. R£cor< i NO.
EPA-460/3-76 -022
4. TITLE AND SUBTITLE
Nitrogen Oxide Control With the Delayed—
Mixing Stratified-Charge Engine Concept
3. RECIPIENT'S ACCESSION NO.
5, REPORT DATE
July 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHORS)
L.W. Evers, P.S. Myers, O.A. Uyehara
8. PERFORMING ORGANIZATION REPORT NO.
9. PER FO TIMING ORG "KNI Z ATI ON NAME AND ADDRESS
Dept. of Mechanical Engineering
University of Wisconsin-Madison
Madison, Wisconsin 53706
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
R-803858-01-0.
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection AGency
Emission Control Technology Division'
Ann Arbor, Michigan 48105
13. TYPE OF REPORT AND PERIOD COVERED
11-10-75-11-9-76
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTfcS
16. ABCfUACT
The purpose of this study is to explore methods of controlling the
nitrogen oxide emissions from internal combustion engines. From
computer calculations, the delayed mixing stratified charge engine
concept was selected. In the delayed mixing stratified charge en-
gine concept, combustion is initiated and completed in a fuel-rich
region, then air is mixed into those rich products. A study of
existing engines shows that some operational stratified charge en-
gines limit nitrogen oxide emissions in a manner similar to the
delayed mixing concept. A single cylinder engine was modified to
include an air injection valve. When air was injected after rich
combustion, the nitrogen oxide emissions were lower, the hydro-
carbon emissions were lower, the carbon monoxide emissions were
about the same and the efficiencies were lower than for homogeneous
operation at the same overall fuel-air ratio.
17.
KEY WORDS AND DOCUMENT ANALYSIS
CK... 1IPTORS
exhaust emissions, combustion
products, internal combustion
engines, nitrogen'oxides, valves,
burning rate
r;. DI.: r 0
% s r A i r_M
Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS
stratified charge
engines, delayed
mixing, post
oxidation
I'J. SECUFIITY CLASS ('1 His liffmrtj
f led
2O. SECURITY CLASS (Tliis page}
Unclassified
COSATl Field/Group
0702, 1311,
2102, 2107,
2111
21. NO. Oh CAGES
246
22. PRICE
EPA Form • J20-1 (9-73)
------- |