EPA-460/3-76-016
August 1976
CHARACTERIZATION
AND RESEARCH INVESTIGATION
OF METHANOL
AND METHYL FUELS
IN AUTOMOBILE ENGINES
FIRST YEAR REPORT
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 .
-------
EPA-460/3-76-016
CHARACTERIZATION
AND RESEARCH INVESTIGATION
OF METHANOL AND METHYL FUELS
IN AUTOMOBILE ENGINES
FIRST YEAR REPORT
by
R.K. Pefley, A.E. Royce, L.H. Browning,
M.C. McCormick, and M.A. Sweeney
University of Santa Clara
Department of Mechanical Engineering
Santa Clara, California 95053
Grant No. R8035480-01
EPA Project Officer: Robert J. Garbe
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
August 1976
-------
11
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 Protection Agency by the
University of Santa Clara, Department of Mechanical Engineering, Santa
Clara, California 95053, in fulfillment of Grant No. R8035480-01.
The contents of this report are reproduced herein as received from the
University of Santa Clara. 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 No. EPA-460/3-76-016
-------
INDEX
Page
LIST OF FIGURES ii
I. INTRODUCTION 1
II. PROGRAM OBJECTIVES 1
III. SUMMARY OF ACTIVITIES RELATING TO THE FIRST YEAR OF THE GRANT... 1
IV.1 ENGINE PERFORMANCE AND EMISSIONS CHARACTERIZATION - METHANOL
VERSUS INDOLENE 4
IV.1.1 Introduction and Summary 4
IV.1.1.1 Indolene Base Line 5
IV.1.1.2 Methanol Base Line 6
IV.1.2 Comparative Experimental Results 7
IV.1.2.1 Experimental System Description and Engine Modifications 9
IV.1.2.2 Carburetor Modification for Methanol Operation 11
IV. 1.2,3 Experimental Procedure 11
IV.1.3 Engine Performance - Indolene Versus Methanol 12
IV.1.3.1 Lean Misfire Limit and MBT Spark Advance for a Multicylinder
Engine 13
IV. 1.3.2 Comparative Power at 2000 RPM 17
IV.1.3.3 Comparative Thermal Efficiency at 2000 RPM 19
IV.1.4 Engine Exhaust Emissions 21
IV.1.4.1 Cylinder-to-Cylinder Equivalence Ratio Variations 23
IV. 1.4.2 C02, CO and 02 27
IV. 1.4.3 C02, CO, 0- Comparative Results 27
IV.1.4.4 UHC, Aldehydes, and NO.,-Indolene 27
IV.1.4.5 Comparative NOX, UBF, and Aldehydes 42
IV.1.5 Conclusions ana Projected Activities 46
IV. 2 ALTERNATIVE FUEL INDUCTION SYSTEMS 48
IV.2.1 The WHB Shock Wave Induction System 48
IV.2.2 Electronically Controlled Fuel Injection System 49
IV. 3 COLD STARTING 52
IV.4 INFLUENCES RESULTING FROM VARIATIONS IN COMPOSITION OF
METHANOL BASED FUELS 53
IV.4.1 Volatility of Fuels 53
IV.4.2 Types of Fuels Tested 53
IV. 5 ENGINE WEAR 59
IV.5.1 Pinto Engine Wear Results 59
IV.6 THERMOCHEMICAL ENGINE PROCESS MODELING 67
V. REFERENCES 72
APPENDIX 74
-------
11
LIST OF FIGURES
Figure
1 Test Engine - Dynamometer Configuration 10
2 MET Spark Timing Effective Lean Limit for Two Load Conditions - 14
Methanol Vs. Indolene
3 MET Spark Timing Showing True Lean Limit and Cylinder 4> 15
Variation, Wide Open Throttle - Methanol Vs. Indolene
4 MET Spark Timing Showing True Lean Limit and Cylinder <)> 16
Variation for Part Load Condition - Methanol Vs. Indolene
5 Brake Horsepower Vs. Equivalence Ratio - Methanol Vs. Indolene 18
6 Comparative Brake Thermal Efficiency - Methanol Vs. Indolene 20
7 Exhaust Emissions Sampling System 22
8 Cylinder-to-Cylinder Equivalence Ratio Variations - Indolene 24
9 Cylinder-to-Cylinder Equivalence Ratio Variations - Methanol 25
10 Methanol Vs. Indolene - Cylinder-to-Cylinder Equivalence Ratio 26
Variation
11 C02 Vs. Equivalence Ratio - Indolene Data 28
12 CO Vs. Equivalence Ratio - Indolene Data 29
13 02 Vs. Equivalence Ratio - Indolene Data 30
14 C0_ Vs. Equivalence Ratio - Individual Cylinder Data - Methanol 31
Vs. Indolene
15 CO Vs. Equivalence Ratio - Individual Cylinder Data - Methanol 32
Vs. Indolene
16 0- Vs. Equivalence Ratio - Individual Cylinder Data - Methanol 33
vs. Indolene
17 Unburned Fuel Vs. Equivalence Ratio - Composite Cylinder Data 35
18 Aldehydes Vs. Equivalence Ratio - Composite Cylinder Data 36
19 Oxides of Nitrogen Vs. Equivalence Ratio - Composite Cylinder 37
Data
20 Oxides of Nitrogen Vs. Equivalence Ratio - Individual Cylinder 38
Data
21 Oxides of Nitrogen Vs. Equivalence Ratio - Speed Dependency - 39
Individual Cylinder Data
22 Oxides of Nitrogen Vs. Equivalence Ratio - Load Dependency - 40
Individual Cylinder Data
23 Effects of Spark Timing on NOX and IHP 41
24 Effects of Spark Advance on the Oxides of Nitrogen - 43
Individual Cylinder Data
25 Oxides of Nitrogen Vs. Equivalence Ratio - Individual 44
Cylinder Data
-------
Ill
Figure Page
26 Unbumed Fuel Vs. Equivalence Ratio - Methanol Vs. Indolene - 45
Composite Cylinder Data
27 Aldehydes as a Function of Equivalence Ratio - Methanol 47
Vs. Indolene
28 Electronic Fuel Injection System 51
29 ASTW Distillation 56
30 Wear of Pinto Engine No. 1 During Break-in 64
31 Wear of Pinto Engine No. 1 65
32 Wear of Pinto Engine No. 2 66
33 Factors Being Considered in Combustion Modeling 68
Al Maldistribution fi Performance Trends for Indolene - 77
Repetitive Test Points
A2 Trends in Brake Thermal Efficiency - Part and Full Load 78
Comparisons at Various Speeds
A3 Brake Horsepower and Maldistribution - Indolene 79
A4 Brake Horsepower and Maldistribution - Methanol 80
AS Cylinder-to-Cylinder Equivalence Ratio Variations - Indolene 81
A6 C02 Vs. Equivalence Ratio - Composite Cylinder Data 82
A7 CO Vs. Equivalence Ratio - Composite Cylinder Data 83
A8 0_ Vs. Equivalence Ratio - Composite Cylinder Data 84
-------
I. INTRODUCTION
Although this is the first year of grant activities, this report marks
the eighth year of research at the University of Santa Clara relating to
methanol based fuel* as a reducer of pollution and as an extender and
supplement of energy supplies. Over that time period the authors have
experienced a cautious growth in optimism that methanol based fuels are the
most attractive alternate fuels from a user viewpoint which can be produced
from coal, remote natural gas, agricultural and municipal wastes. This
attractiveness is not to be found in any one major advantage. Rather, it is
a composite of the advantages to be found in combustion behavior, emissions,
handling safety, ecology factors, and ease of power plant conversion as
weighed against the cost of fuel production. The past year's progress
reported herein adds to this favorable evidence. While this work relates
primarily to the automotive possibilities, the broader potential advantages
of the fuel should be kept in mind.
II. PROGRAM OBJECTIVES
The program objectives as proposed for the two year grant period are:
(1) To characterize multicylinder, automotive engine performance and
emissions when operating on methanol and methyl fuel by comparison
with gasoline.
(2) To identify and assess alternatives for improvement of engine performance
and reduction of emissions.
(3) To select the best alternates and modify the engine fuel induction
system accordingly.
(4) To evaluate the modifications in terms of performance and emissions and
interpret the results.
III. SUMMARY OF ACTIVITIES RELATING TO THE FIRST YEAR OF THE GRANT
During the past year, literature searches, experimental investigations,
analytical modeling studies and publications have all been parts of the work
program. These efforts have provided some answers to the grant's investigative
objectives and have added focus to the ensuing work. The highlights of the
activities and evidence that have been obtained are:
*Methanol based fuel includes methanol, dissociated methanol (a gaseous
fuel--CO + 2H«) and methyl fuel which contains higher alcohols resulting from
high volume, low cost production and other additives for cold start up of
automotive engines or gas turbines.
-------
(1) By the end of the grant year,two identical (2.3 litre, 4-cylinder, 1975
model year Ford) engines have been completely instrumented and have
become fully operational. One engine is designated as the test engine.
The second engine is a back-up to the test engine and is used for
adaptation and check-out of modified fuel preparation systems.
(2) Using equivalence ratio* as the primary basis for comparison, our engine
test program is indicating that methanol is superior to gasoline in
thermal efficiency, operational range of equivalence ratio, and emissions
levels with the exception of aldehydes. This evidence corroborates our
published findings which resulted from our earlier experimental work and
literature search.
(3) The OEM carburetor, intake manifold, fuel pump combination for the test
engine show such wide cylinder-to-cylinder and time variant equivalence
ratios as to emphasize the need for the development of alternate mixture
preparation hardware. This evidence accentuates the importance of the
portion of our work relating to the investigation of alternate fuel
preparation systems.
(4) Several fuel preparation systems have been identified as offering
advantages over the OEM venturi carburetor and intake manifold. Two of
these systems have been built under purchase agreements and will be
delivered in the next sixty days. One is a pressure wave carburetion
system which uses the strength of the rarefaction pressure wave generated
by the intake valve opening to meter and nebulize the fuel for each
intake charge. This system has fewer moving parts and fewer total parts
for a four-cylinder engine when compared with the OEM system. The other
is a fuel injection system with electronic control to allow management
of equivalence ratio for each cylinder as well as injection timing.
Prototypes of both of these systems have been operated in our experimental
program during this past year.
(5) With the use of ASTM distillation and Reid vapor pressure procedures,
several interesting possibilities have been observed relative to a methanol
based fuel that has cold starting enhancement. The two most interesting
mixtures to date are pure methanol with a small percentage of butane
addition and the methanol-gasoline blend consisting of 16% methanol
arid 4% isobutanol.
*Equivalence ratio is stoichiometric air-fuel divided by actual air-fuel.
-------
(6) Lubricity of engines using methanol fuel and petroleum based lubricating
oil is important from a wear viewpoint. To track the contrast between
methanol and gasoline in this regard, atomic absorbtion spectrophotometry
has been employed to examine the metal accumulation in the oil with
respect to time. Engine test time using methanol as a fuel is too limited
as yet to draw any comparative conclusions. However, the method being
used appears to provide a suitable basis from which to draw comparative
evidence.
(7) Computer modeling of the JT8D gas turbine using chemical kinetic reaction
rate schemes for jet fuel and methanol shows, that without reduction in
thrust level, NOY emissions can be reduced by as much as 80% if methanol
2
is substituted for jet fuel. Experimental investigations by others in
adiabatic burners and automotive gas turbines appear to confirm this
3 4
prediction. ' This computer model is now being adapted to model the
2.3 litre test engine. Should the model predict satisfactory contrasts
in emissions between gasoline and methanol, it will be extremely useful
in making parametric projections toward improved methanol fueled engine
performance.
(8) Two vehicles, a 1972 Plymouth Valiant and a 1970 American Motors Gremlin,
have successfully completed their fourth and fifth year, respectively, of
continuous operation on pure methanol with no major component failures.
The Plymouth, which is in service in the City of Santa Clara, ranks
competively in fuel economy (miles/mm Btu) over its operational lifetime
with vehicles in like service operating on gasoline. Cold starting is the
most serious driver reported difficulty. An additional 1972 Plymouth
Valiant belonging to the City of Santa Clara has been operating on a
methanol-90% gasoline blend for over a year with no modifications and no
operational problems.
-------
IV.1 ENGINE PERFORMANCE AND EMISSIONS CHARACTERIZATION -
METHANOL VERSUS INDOLENE
Results of an experimental parametric investigation of a four-cylinder
spark ignition engine operating on methanol and on indolene are presented in
this section. This phase of the investigation represents the major proportion
of work performed under the first year effort. In order to keep the presenta-
tion and discussion of the results brief, experimental procedure and technique,
along with data acquisition and analysis methods, will be covered in more
detail in the final report. The following sections cover the topical matter
in five steps. Section IV.1.1 is an introduction and a summary of the conclu-
sions based on the experimental evidence. Section IV.1.2 describes the test
procedure and engine modifications. Section IV. 1.3 compares in detail engine
performance while operating on the two fuels. Section IV.1.4 compares and
discusses at length the exhaust emissions resulting from the use of the two
fuels. Interpretations and projected activities are covered in Section IV.1.5.
IV.1.1 Introduction and Summary: There is considerable literature evidence
comparing engine performance and exhaust emissions of methanol versus gasoline
fueled engines. The majority of these studies,however, has concentrated on
steady state, single cylinder engine comparisons. To fill the need for compar-
ative evidence in multicylinder engines, the first year effort in this investi-
gation has been directed towards the acquisition of complete exhaust emissions
and performance data for a conventional four-cylinder engine operating on
methanol and a reference gasoline (indolene). The engine exhaust was analyzed
for principal combustion species, as well as for the pollutant species of CO,
NO, NO.., unburned fuel (UBF) and aldehydes. Engine performance was described
by indicated and brake thermal efficiency (a measure of fuel economy on an
energy basis) and power. Results are presented primarily as functions of fuel-
air equivalence ratio () since this is, in general, the most influential
operating variable affecting specific exhaust emissions (grams pollutant per
indicated horsepower hour - [gms/ihp-hr]) and engine fuel economy (on an
energy basis).
Multicylinder engine operation introduces an important parameter which is
not seen in single cylinder operation. This is the cylinder-to-cylinder
variation in equivalence ratio which is a result of the less than perfect
mixing and distribution of the fuel and air in the intake manifold. The
exhaust emissions are a strong function of equivalence ratio and thus composite
-------
exhaust emissions (exhaust contributions from each of the cylinders combined)
often show what appears to be confused emissions data because of this "mal-
distribution" effect. Through treatment of multicylinder engines as a
combination of single cylinder engines, the confused emissions data becomes
explicable. In addition, this approach leads to the understanding of the
apparent inconsistencies in the setting of MET (minimum spark advance for best
torque) spark timing. Finally, it leads to an understanding of why conventional
multicylinder engines operate over narrower ranges of equivalence ratio and
spark timing when compared to single cylinder engine operation.
Our principal objective in this first year of the experimental program has
been to provide a comparative study: (1) using a stock four-cylinder engine
(with the exception that the stock exhaust gas recirculation [EGR] was not used)
operating at steady state on indolene for a base line, (2) then minimally
modifying the fuel induction system for methanol operation at equivalent engine
conditions. This is a first step in progressive studies aimed at achieving
optimal engine performance and minimum pollutant emissions with methanol as a
fuel in automotive spark ignition multicylinder engines.
IV.1.1.1 Indolene Base Line: The characterization of engine performance and
emissions with indolene consisted of 140 steady state engine conditions, the
data from which provides a basis for the following observations:
(1) Multicylinder engine operation at steady state yielded a range of
equivalence ratio variations (maldistribution) among the cylinders for
the speed and load range observed. The effect was most pronounced at
wide open throttle conditions.
(2) Exhaust emissions were dramatically affected by this maldistribution.
The oxides of nitrogen, owing to their characteristic strong dependency
on equivalence ratio and spark advance, displayed the most complex
behavior. Resolution of the data in terms of the chosen parameters of
this investigation led to an understanding of why NO*, emissions are
difficult to control in multicylinder engines.
Composite carbon monoxide and unburned hydrocarbons showed characteristic.
lean side minimums up to the lean misfire limit; however, at times the
maldistribution phenomena led to significant quantities of CO near the
composite stoichiometric air to fuel ratio. Aldehydes, a toxic and
therefore important exhaust species to be measured, showed composite
exhaust sample values characteristically increasing with the leaning of
the air-fuel mixture.
-------
(3) Engine power, fuel economy, and thermal efficiency were somewhat
degraded by the maldistribution!but general trends closely follow
results obtained for laboratory single cylinder engines. The only
exception to this general rule is that acceptable engine operation was
obtained over a narrower range of spark settings and equivalence ratios.
In summary, while acceptable steady state engine performance was attainable
for this conventional venturi carbureted four-cylinder engine, fundamental
design defects in the intake manifold-carburetor system led to equivalence ratio
variation among the cylinders. This variation in equivalence ratio caused
composite pollutant emissions to be at times higher and at times lower than
those if all four cylinders were operating at the composite equivalence ratio.
Consequently, it is concluded that control over exhaust emissions (especially
NO..) is precluded by the lack of control over the individual cylinder equivalence
ratios for this and the majority of conventional venturi carbureted multicylinder
engines. However, it is anticipated that if control over equivalence ratio is
attained and spark timing is programmed for speed and load conditions, then the
0.4 gms/mi NO..standard might be met without the use of extensive add-on controls
even for gasoline. Steady state data supports this contention; however,
transient engine characterization is needed before it can be truly verified.
IV.1.1.2 Methanol Base Line; At the conclusion of the indolene fueled engine
characterization, the stock carburetor was rejetted for methanol. The resulting
characterization for similar testing conditions (i.e., same inlet air conditions,
manifold vacuum settings, equivalence ratio range, and MBT spark timing) was
termed as the "base line" for methanol. Results have been compiled for 2000
and 2500 RPM. Based on the comparative evidence at 2000 RPM, the following
observations are presented:
(1) Cylinder-to-cylinder variations in equivalence ratio for steady state
operation were consistently worse than those found in the indolene
characterization. Again, the effects were most pronounced at wide
open throttle settings and to a lesser degree at part throttle settings.
(2) Even with this aggravated operating condition, engine power levels
were similar to those produced by indolene. The thermal efficiency
improved from 4% to 10% for the load conditions and, as expected, the
"true" lean misfire limit was extended. However, the maldistribution
phenomena did lead to an instance where the "effective" lean misfire
limit was not as lean as that attained with the indolene.
-------
(3) Composite exhaust emissions showed trends similar to those for indolene;
however, the absolute values of the constituent species showed some
differences. NOX emissions were reduced by a factor of two. Unburned
hydrocarbons and CO were comparable in value. The only species showing
an increase were the aldehydes, which ranged from two to eight times
higher.
In summary, even though the methanol base line is incomplete, the data
indicates that methanol yields improved engine thermal efficiency although the
maldistribution phenomenon was more severe. Emissions of NOY, the most
A
difficult of the pollutant species to control in present day automotive engines,
were reduced by a factor of two. In replacing the jets of the stock two-barrel
carburetor, it was found that internal flow passage dimensions limited the
methanol delivery rate. This indicates, at least for the carburetor used, that
jet replacement would not be an adequate technique for conversion to methanol
even for steady state engine operation. It is anticipated that the modified
carburetor will lead to driveability problems during transient engine operation,
in addition to the lack of control over cylinder-to-cylinder equivalence ratio
already indicated by the steady state testing.
The results of the collective emissions so far obtained indicated that
precise control over cylinder-to-cylinder variations in equivalence ratio would
make possible a significant improvement in operation of spark ignition auto-
motive engines. And, when operated on methanol, the resulting expanded lean
burning capability could produce higher thermal efficiency and a reduction in
oxides of nitrogen approaching one order of magnitude in comparison to current
emissions levels in gasoline fueled engines.
IV.1.2 Comparative Experimental Results: The experimental program in the
first year was devoted primarily to a parametric investigation of steady state
engine performance and emissions of .two laboratory dynamometer mounted 2.3
litre four-cylinder engines (Ford Pinto).
The parameters chosen for the characterization appear in Table 1 and are
collectively termed the "base line test matrix". These five variables (speed,
equivalence ratio (<{>), load, spark timing, and inlet air temperature) are most
commonly employed in the mapping of engine performance and exhaust emissions
in single cylinder engines. Multicylinder engine operation, although, adds a
new variable which we have already termed "maldistribution". This new variable
results from the inadequate mixing and distribution of the fuel-air mixture in
the carburetor.
-------
TABLE 1
BASE LINE TEST MATRIX
Engine
Speed
(RPM)
Idle
1500
2000
2500
3000
3500
4000
Notes:
m Load m
^ } (Fraction)1 J
0.6^ 1/3
0*1 1** J tii
.7 Z/3
0.8 Full
0.9
1.0
1.1
1.2
Spark
Setting
(°BTDC)
Other '
Temperature
70
V Other M
(1) Equivalence ratio, i.e., stoichiometric * actual A/F.
(2) Fraction load is designated by intake manifold vacuum:
Full = 0" hg vac; 2/3 = 7"; 1/4 = 14".
(3) Minimum spark advance for best torque.
(4) Implies less than a full complement of speed, equivalence
ratio, and load settings will be used for these
conditions.
(5) Methanol only
and intake manifold. It is present in all conventional venturi type carbureted
multicylinder engines and differs only in degree from one carburetor-manifold
system to the next. Maldistribution implies variations in among the cylinders,
The data indicates that engine power and thermal efficiency were not strongly
affected by the range of maldistribution encountered. In other words, when
the data was compared with single cylinder engine data the trends were similar.
Certain of the exhaust emissions (notably NO..) and the determination of spark
timing were greatly affected by these
-------
performance output = f (speed, load, composite , spark timing,
(ihp, Bhp, n. ., r\ , ) and air temperature)
where n and r\ . are brake and indicated thermal efficiency, respectively,
and composite <|» is the average of of the four cylinders, and
composite emissions species = f (speed, load, <(>,, 2, <(>_, 4, spark
(gms/ihp-hr) timing, and inlet air temperature)
where is defined on a fuel-air basis; that is, values less than
unity are fuel lean and values greater than unity are fuel rich. However, due
to the considerable lack of clarity in the literature of defining and
presenting this variable in a consistent manner, "lean" or "rich" conditions
are indicated on each graph where appears as the independent variable.
IV.1.2.1 Experimental System Description and Engine Modifications: Figure 1
is a schematic of the engine-electric dynamometer test configuration and
associated equipment. The engine depicted was designated as the principal test
engine from which all data appearing in this report was taken. A second engine-
dynamometer system provided a back-up in the event that the principal engine
failed. No engine failures occurred in the first year and the second system
served primarily as a "check-out" system for the modified stock carburetor.
In the second year this system will be used in the transient engine characteri-
zation and for initial testing of the alternative fuel-air induction systems.
The principal test engine was instrumented with conventional equipment;
however, a few additional pieces of specialized instrumentation are worth
noting. A fast time response hot wire anemometer flow transducer was used to
measure instantaneous fuel flow rate into the carburetor. In addition, a fast
time response pressure transducer was connected upstream of the float bowl fuel
shutoff valve for observation of instantaneous pressure. Together, they
provided a means for observing the dynamic fuel flow characteristics. Stain-
less steel exhaust sampling probes and thermocouples were placed in each
runner of the exhaust manifold. Their location is indicated in the detail of
Figure 1. Likewise, thermocouples were placed in the intake manifold runners
as pictured. The thermocouples provided temperature measurement one-quarter
inch above the runner floor. A high sensitivity load cell was connected in
the linkage hardware of the dynamometer scale for more precise measurement of
torque during the setting of MET spark advance. Manual control over the
spark timing was provided by a bracket and lever device which rotated the
-------
HEAT
EXCHANGER
BLOWER
VALVE
HINGED
DOOR
FUEL PRESSURE
EXHAUST SAMPLE VALVES
(I PER CYLINDER, I COMPOSITE)
• TO
ry\trfi-M\ ir- I0
r *&/rrr, -'yr ' EMISSIONS
"-Vh ®»*W \ \ BENCH
? 1 /Ok ; \ >T-HEATED LINE
>^A^; ^
J /\\ ®*^ •—INSULATION INTAKE MANIFOLD |
INTAKE MANIFOLD
(BOTTOM VIEW)
T
MANIFOLD
HEATER
COOLING
WATER
P.C.V.
E.O.R.-P.C.V. SLOT
LAMINAR
FLOW
ELEMENT
FUEL
FUEL FLOW
TRANSDUCER
EXHAUST
SYSTEM
CHART
RECORDER
THROTTLE
CONTROL I J
BLOW BY-
CLUTCH
FUEL TANK
AND
SCALE
LOAD (LBF) RPM
4 CYL ENGINE
2300 C.C.
EXHAUST MANIFOLD
(FRONT VIEW)
BOWL
PRESSURE
CHART
RECORDER
AIR-FUEL
CONTROL
PUMP
VACUUM
FIG. I TEST ENGINE-DYNAMOMETER CONFIGURATION
I :
THERMOCOUPLE LOCATIONS
/L I
I. CYL I-HEAD
2. CYL I-PLENUM
3. CYL 2-HEAD
4. CYL 2-PLENUM
5. CYL 3 -HEAD
6. CYL 3 -PLENUM
7. CYL 4-HEAD
8. CYL 4 -PLENUM
9. WATER TEMP -PLENUM
10. WATER TEMP - HEAD
II. OIL TEMP
12. L.F. E. AIR TEMP
13. CYL I EXHAUST
14. CYL 2 EXHAUST
19.CYL 9 EXHAUST
16. CYL 4 EXHAUST
17. EXHAUST EMISSIONS SAMPLE
-------
11
distributor. An external heating source and fluid was substituted for heating
of the intake manifold plenum. (Normally, the stock engine provides engine
coolant for this heating.) The external source permits exploration of plenum
temperature effects on engine operation; however, comparative test data to date
has been taken for a fixed plenum heating temperature of 180 +_ 5°F (consistent
with engine coolant temperature).
Air to fuel ratio (or <|>) control was obtained primarily through pressuri-
zation or evacuation of the air space in the float bowl chamber of the
carburetor. A small vacuum-pressure pump provided the necessary range of
pressures. In addition, an intake air blower provided small positive intake
air pressure and thus permitted additional control over the air to fuel ratio.
Combined with appropriate main jet changes for methanol, this control technique
provided the desired <(> range for both fuels.
For all testing the EGR system was disconnected; however, the positive
crankcase ventilation (PCV) valve was connected in such a way that precarburetor
air was passed into the EGR-PCV slot of the carburetor. Blow-by gases were
vented into the laboratory exhaust removal system as were the engine exhaust
gases. Engine coolant temperature was maintained by the stock thermostat, and
a shell and tube heat exchanger provided cooling capability for all engine
load conditions.
IV.1.2.2 Carburetor Modification for Methanol Operation: The stoichiometry
of methanol-air mixtures dictates a fuel flow rate approximately double that
required for gasoline-air mixtures. This is understood a little more clearly
in terms of the lower heating values of the two fuels. Methanol's heating
value (Btu/lbm basis) is roughly half that of gasoline; thus, to produce
comparable power levels roughly twice as much methanol must be burned (assuming
comparable thermal efficiencies). After 130 steady state runs on indolene,
the stock two barrel venturi carburetor was minimally modified to permit this
increased fuel flow. The main and idle jets were replaced but it was found
that internal flow passages would not permit adequate fuel flow rates. After
sealing off the primary emulsion tube, acceptable fuel flow rates were
attainable. This constituted a first-cut conversion, and all the methanol
data presented in the report was taken with this configuration.
IV.1.2.3 Experimental Procedure; The procedure for setting a test condition
consisted of the following:
-------
12
(1) Setting a speed and load condition.
(2) Fixing <|>.
(3) Finally checking the spark timing for MET conditions.
In the early stages of the testing, MET was found by noting changes in
torque reading on the dynamometer scale. Repeatability in making this setting
was not particularly good (primarily in the lean region) and it was found that
the condition could be established more accurately using a high sensitivity
load cell and strip chart recorder. Even with this improvement, the MET spark
timing curve still showed some erratic behavior. As discussed in the following
sections this behavior was attributed to engine operating characteristics
rather than errors in experimental technique.
To establish changes in engine performance resulting from the number of
test hours or "aging" of the engine, a specified test condition was repeated
at regular intervals throughout the steady state testing. This data was used
to establish a correction factor which when applied to the performance data
eliminated the time dependency or engine "aging" effect. In this way, a fair
comparison of indolene and methanol has been made. An illustration of these
time dependent effects appears in the Appendix.
The sequence of testing to date has been:
(1) 130 steady state test conditions on indolene, followed by
(2) 47 steady state test conditions on methanol, followed by
(3) 10 steady state runs on a commercial unleaded gasoline.
The latter 10 runs were repeat points of various indolene test conditions.
Equivalency of power, thermal efficiency, and emissions for indolene and the
commercial unleaded gasoline was established at the end of the indolene base
line thus permitting subsequent comparisons of engine results for the repeat
test condition.
In the discussion sections that follow, the results are interpreted in
light of the maldistribution phenomena. Engine performance appeared to be
less sensitive than engine emissions to this phenomena and therefore we have
chosen to present performance results initially. More detail of the <|> varia-
tions among the cylinders (maldistribution) will be presented in the emissions
section.
IV.1.3 Engine Performance - Indolene versus Methanol: The 140 steady state
test points obtained on indolene provided a comprehensive map of engine
-------
13
performance. Plots of brake power and thermal efficiency across the speed and
load range showed behavior typical of multicylinder spark ignition engines.
(Some of these plots appear in the Appendix.) The general performance trends
also closely follow typical single cylinder engine results such as those
presented by Ebersole and Manning. It should be noted that all the power
results presented are both temperature and manifold vacuum corrected
^reference = 70°F' and Pmanifold = 14'°' 7'°» or °'° in' H* vacuum.and Patmosphere
» 29.92 in Hg). Some of the performance data will be taken at 4000 RPM;
however, the low speed,part load conditions are considered to be of more
importance since it is at these conditions that typical automotive engines
spend the majority of their operating lifetimes.
IV.1.3.1. Lean Misfire Limit and MET Spark Advance for a Multicylinder Engine:
The lean burning capability of methanol in engines is well documented in the
literature; however, most of the data is for single cylinder engine operation.
In multicylinder engines, nevertheless, this lean operating capability can be
effectively masked if significant $ variations exist among the cylinders.
Figure 2, a comparative plot of MBT spark advance as a function of the composite
(average) for two load conditions, demonstrates this fact. Note that at the
wide open throttle condition (0 in. Hg manifold vacuum), the indolene shows an
"effective" lean limit which is leaner than the methanol results. .The lean limit
is termed "effective" because one cylinder is operating much leaner than the
composite indicates. Figures 3 and 4 indicate the "true" lean limit for the
two fuels in this four-cylinder engine. In these plots, four individual
cylinder 's appear for each composite <|> that appears in Figure 2. Figure 3
(wide open throttle conditions) shows that methanol's true lean limit is beyond
that of indolene. This is not obvious in Figure 2. These 4> variations among
cylinders are responsible for indicating an apparent (or effective) lean limit
at = 0.82 when, in fact, it is due to the leanest cylinder operating at
<(> = 0.64. Figure 4, the part load comparison corroborates the effective lean
limit results of Figure 2. This results from the fairly good agreement in <|>
among the cylinders for both fuels.
The results indicate that as the cylinder-to-cylinder variation in is
decreased, the effective lean limit approaches the true lean limit. Practically
speaking, this means that in the absence of maldistribution, engines can
operate at leaner 's without engine misfire for any fuel. This condition was
approached at the higher manifold vacuum setting, as indicated in Figure 4.
-------
10
FIG. 2 MBT SPARK TIMING EFFECTIVE LEAN LIMIT FOR TWO LOAD CONDITIONS
METHANOL vs INDOLENE
-------
2000 RPM
INDOLENE
•,9,9, METHANOL
TRUE LEAN LIMIT
MANIFOLD VACUUM - 0" H6
FIG. 3 MBT SPARK TIMING SHOWING TRUE LEAN LIMIT AND CYLINDER 0 VARIATIONS, WIDE
OPEN THROTTLE — METHANOL vs INDOLENE
-------
70
60
50
40
o
QC
2
CO
30
20
K)
2000 RPM
INDOLENE
METHAWOL
- TRUE LEAN. LIMIT
MANIFOLD VACUUM.- 14" HG
0.5 0.6 0.7 0.8
0.9
0
1.0
I.I
1.2
1.3 1.4
FIG. * MBT SPARK TIMING SHOWING TRUE LEAN LIMIT AND CYLINDER 0 VARIATIONS
FOR PART LOAD CONDITION — METHANOL vs INDOLENE
-------
17
Both Figures 3 and 4 clearly show methanol's leaner operating ability in
comparison with indolene.
These MET curves supply one additional piece of information. With
significant maldistribution, none of the four cylinders are operating at their
individual MET spark settings, and this is especially true in the lean region
where spark sensitivity to equivalence ratio is the greatest. The engine's
MBT spark setting is actually a compromise. Maldistribution, where one
cylinder is excessively lean,as scan in Figure 3,causes spark timing to be
advanced to keep that cylinder from misfiring. But, this in turn reduces
power and efficiency in the richer cylinders because the spark is over-advanced
for those cylinders. This results in lower overall power and efficiency for
the four cylinder engine. The over-advanced and retarded spark condition also
limits the range of acceptable spark timing at which the engine will operate.
During the indolene runs, there was one condition in which this spark tolerance
band was only 5 degrees in width (27° to 33° BTDC). This condition was at
2000 RPM-wide open throttle with a composite of 0.80. This lean setting
was an unstable condition for which load had to be adjusted constantly to
maintain engine speed. Data was taken at both spark settings; however, at
greater spark advances severe knocking occurred, and at lower spark advances
misfire occurred.
These non-optimized spark settings also affect hydrocarbon, aldehydes,
and oxides of nitrogen emissions. Special attention is paid to the NO., in
the section on emissions results .
IV.1.3.2 Comparative Power at 2000 RPM: As indicated in Figure 5, even with
the severe maldistribution of the methanol runs, comparable power was
delivered by the engine for like load and equivalence ratio settings. The
leanest operating engine condition occurred with methanol at a manifold vacuum
of 7 in. Hg. The distribution of <(> values among the cylinders is indicated
for this point. The maldistribution was relatively small in this instance,
thus allowing stable engine operation at a composite of 0.65.
Characteristic of all spark ignition engine operation is the more abrupt
fall-off of power in the lean region. This is seen a little more clearly in
an expanded ordinate plot of BHP versus (See Appendix). Given maldistribu-
tion, the individual cylinder power can be estimated from Figure 5, which is
really a plot of "composite" engine power. For the worst instance of mal-
distribution at wide open throttle on methanol, <(> varied by 0.72 between the
-------
5O
DO
40
tr.
UJ
O ^f\
z 30
UJ
to
ac
o
X
5 2°
a:
m
ir»
IU
0
PL
a*
N
1
k 4
^ i
o3
o
i
^
» X^
1 1*
„..-
^ J
•'
• ^
*^
, ..
0* .
%
o
GO
A 4
a
2000 RPM
MBT SPARK TIMING
D H-I4* HG MAN VAC
A A- 7" HG MAN VAC
O •-0" HG MAN VAC
O D A - METHANOL
• • A - INDOLENE
EFFECTIVE LEAN MISFIRE LIMIT
^ METHANOL
^ INDOLENE
X- INDIVIDUAL CYLINDER 0 VALUES
(TRUE LEAN MISFIRE LIMIT INDICATED BY §
0.5 0.6
0.7
0.8
0.9
1.0
I. I
I. 2
1.3
0
FIG 5: BRAKE HORSEPOWER VS EQUIVALENCE RATIO - METHANOL VS INDOLENE
-------
19
worst two cylinders (cylinder 4, = 0.93 and cylinder 3, = 1.65). The
corresponding BHP for these two cylinders was approximately 43.7 and 46.6,
or a variation of about 6%. In the lean region, though, variation of 4> in one
instance was only 0.27 (f>, but this resulted in cylinder BHP values of 34.0 and
45.7, or a 34% variation in power. This indicates that lean burning engines
must have good control over individual cylinder 4> in order to maintain maximum
engine power.
IV.I.3.3 Comparative Thermal Efficiency at 2000 RPM: From an energy stand-
point the engine brake thermal efficiency indicates the true merit of one fuel
over another, since it is a measure of the useful work supplied for the fuel
energy input. In Figure 6, the brake thermal efficiency is plotted for the
three engine load conditions; again as a function of the equivalence ratio.
For each load condition, methanol shows an improvement over indolene. The
increase in thermal efficiency at = 1.0 for each case translates in a fuel
savings on an energy basis of 9% for full load, 4% for the two-third load
condition, and 10% for the one-third load condition. At the 7 in. Hg manifold
vacuum setting, note that indolene's effective lean limit is located at =
0.82.. In comparison, at this $, methanol is showing an 18% savings in fuel.
Thermal efficiency decreases eventually for both fuels as the lean
region is penetrated because of the approach to the flammability limit.
However, if maldistribution could be eliminated,the thermal efficiency curves
for both fuels would maintain higher values in the lean region than is
indicated by these engine results. In addition, the results for methanol
indicate superior performance when compared to indolene, even with the
increased maldistribution.
In summary, the steady state experimental performance results at 2000 RPM
indicate that even with a non-optimized fuel-air induction system (which
results in a characteristic maldistribution of the fuel-air mixture among the
cylinders),methanol shows comparable power and improved thermal efficiency
over indolene. Methanol also shows a leaner operating capability than
indolene at the two lower load settings. Maldistribution phenomena can be so
severe (as is indicated by the wide open throttle results) that the true lean
limit for both fuels is effectively masked, and an effective lean limit at
higher determines the lean operating limit of a multicylinder engine.
-------
20
5
UJ
u
C
b.
bl
K
U
UJ
X
<
oc
MBT SPARK TIMING
METHANOL
INDOLENE
20
15
0.6 0.7 0.8 0.9 1.0 I.I 1.2 1.3
FIG. 6 COMPARATIVE BRAKE THERMAL EFFICIENCY - METHANOL vs INOOLENE
-------
21
IV.1.4 Engine Exhaust Emissions: Constituent exhaust species measured
during the steady state testing of methanol and indolene have included C02,
CO, 02, unburned hydrocarbons (UBF), NO, NOX, and total aldehydes. A
schematic of the exhaust sampling system appears in Figure 7. The system
includes conventional continuous on-line analyzers for detection of individual
chemical species. Non-dispersive infrared (NDIR) was used for CO- and CO
analysis, a polarographic sensor was used for 02 analysis, a heated flame
ionization detection (HFID) was used for UBF analysis, and a heated chemi-
luminescent analyzer was used for NO and N0» analysis. Aldehydes were the
only chemical species not continuously monitored. They were determined through
the use of a wet chemical technique, known as 3-methyl-2-benzothiazolone
hydrazone (MBTH), which measures the total aldehydes. Two instruments needed
special calibration for methanol exhaust emissions analysis. The HFID was
found to have a relative linear response factor of 0.85 to methanol vapor in
comparison to propane. In addition, the NO/NOX analyzer showed a slight
response to methanol vapor. This was a nonlinear effect when plotted as a
function of ppm methanol. The response asymptotically approached 80 ppm NO..
(indicated) for increasing ppm methanol.
All emissions results are presented on a dry basis. Traditionally, NO..
has been reported as NO-, even though NO- may not have been the primary
constituent. For the sake of comparison with literature and previous progress
reports, it will be reported in this manner. It should be noted, however, that
no NO- was detected in the indolene exhaust, and little has been detected
during methanol testing to date.
For steady state testing, emissions results are presented primarily on a
grams per indicated horsepower hour basis. This normalizes emissions with
respect to the useful energy developed by the engine. On this basis, emissions
are strongly influenced by equivalence ratio (<|>) and usually to a lesser degree
by engine speed, load, spark timing, and inlet air temperature (with the
exception of NO/NO.,). For this reason $ is chosen as the independent variable
for presentation of the exhaust emissions results. Emissions, with the
possible exception of unburned hydrocarbons, did not appear to be highly
influenced by the range of inlet temperatures encountered during testing. For
this reason temperature reference does not appear on the graphs. It should be
noted that intake air temperatures ranged between 70°F and 95°F for the
indolene runs and between 62°F and 84°F for the methanol runs.
-------
FLOW
METERS
EXHAUST
DUMP
LOW. CO
ANALYZER
BACKFLUSH
AIR SUPPLY
CAL
GAS
CAL
GAS
e-®-
CAL
GAS
MID RANGE
CO
ANALYZER
I—®-
r
PUMPS
t
EXHAUST
DUMP
1
?r
\j
?
>
]
>
>
L
INE
OL
S°F
>
>
3
ENSE
®_
•®-
ffL .,
INDICATES HEATED
PORTION OF SYSTEM
30'HEATED LINE 400°F
CONDENSATOR
DUMP
IMPINGER
CONTAINING
MBTH
SOLUTION
FOR
ALDEHYDE
DETECTION
FIGURE 7: EXHAUST EMISSIONS SAMPLING SYSTEM
K)
to
-------
23
IV.1.4.1 CyUnder-to-Cylinder Equivalence Ratio Variations: In the engine
performance results, certain behavior was attributed to the cylinder-to-cylinder
equivalence ratio variations. Figure 8 shows this maldistribution for the
individual cylinders at 2000 and 3000 RPM while operating on indolene. Note
that cylinder equivalence ratio ( .) has been normalized with respect to the
composite equivalence ratio (<)> ); also, this composite equivalence ratio is
the average of the four j. The variations appear to be systematic at high
manifold vacuum with little dependence on .At wide open throttle the
comp r
behavior becomes erratic. At times, cyclinder 2 and 3 are lean relative to
cylinders 1 and 4, and at times just the opposite effect occurs: 1 and 4 are
lean relative to 2 and 3. This behavior is illustrated in Figure A3 (Appendix)
where the maldistribution and BHP appear as functions of at 2000 RPM.
rr comp
This anomalous behavior is believed in part due to the operation of the secondary
power jet in the carburetor and its interaction with the test procedure
utilized for controlling <)> . The worst distribution for indolene (3000 RPM,
«= 0.9, 0 in. Hg manifold vacuum) shows a 26% variation in .. This
represents a variation of 3.8 in air to fuel ratio for indolene. The complete
range of maldistribution for indolene at all speeds, excluding 4000 RPM, can be
seen in Figure A5 (Appendix).
Figure 9 displays the maldistribution for methanol at 2000 RPM for the same
carburetor rejetted for methanol. Figure 10 compares the two fuels at similar
equivalence ratios. Note that in comparison to the 2000 RPM indolene data, the
maldistribution is greatly exaggerated with methanol. The worst distribution
(<(> = 1.3, 0 in. Hg manifold vacuum) shows a 0.72 variation in 4> ., which
represents a 3.2 variation in air to fuel ratio (stoichiometric air-fuel = 6.4).
Maldistribution results for the two fuels at wide open throttle (see Figures A3
and A4, Appendix) show that methanol yields more consistent trends but greater
cylinder to cylinder variations than'indolene. Cylinder 4 is always the leanest
and cylinders 2 and 3 are almost always the richest. This comparative evidence
indicates that the carburetor modifications for methanol use have not improved
the serious maldistribution of the OEM system. An iteration on carburetor
modification before the bulk of the methanol baseline data is acquired is
expected to reduce the severity of this maldistribution.
In summary, the methanol data represents results for a first attempt at
modifying the carburetor through simple jet replacement to allow for the
increased fuel flow. Yet, even though this modification resulted in increased
-------
_ I.I
I. 0
0.9
o _
o.e
1.1
I.O
0.9
0.8
,r
0 CYL
0 COMP
g COMP
V - 0.8
a - 0.9
o - i.o
A - I. I
X - 1-2
1234
CYLINDER NUMBER
2000 RPM
I.I
I.O
0.9
0.8
I.O
0.9
0.8
1234
CYLINDER NUMBER
3000 RPM
FIG. 8: CYLINDER TO CYLINDER EQUIVALENCE RATIO VARIATIONS - INDOLENE
IS)
-------
1.2
25
0" MAN.VAC
g CYL
0 COMP
1.0
0.9
0.8
I.I
1.0
METHANOL - 2000 RPM
0 COMPOSITE
7" MAN VAC ^ ~ 0-8
Q — 0.9
O — 1.0
A ~ I.I
X ~ 1-2
O- 1.3
14" MAN VAC
1234
CYLINDER NUMBER
FIG. 9: CYLINDER TO CYLNDER EQUIVALENCE RATIO VARIATIONS
METHANOL
-------
26
0.7
I.I
0" HG MAN VAC
2000 RPM
MBT SPARK TIMING
- INDOLENE
----- METHANOL
0 COMPOSITE
V - 0.8
7" HG MAN VAC Q _ , Q
X - 1.2
14 HG MAN VAC
1234
CYLINDER NUMBER
FIG.10 METHANOL VS INDOLENE
CYLINDER TO CYLINDER EQUIVALENCE RATIO VARIATION
-------
27
maldistribution for methanol, the engine performance data indicated improved
engine thermal efficiency, comparable power, and an extended lean operating
range.
IV. 1.4.2 CO.,, CO and 0,,: Exhaust emissions are much more sensitive to
variations in resulting from the maldistribution phenomena than are engine
performance characteristics. Figures 11, 12 and 13 show individual cylinder
measurements of C02, CO, and 02 for all the indolene test conditions. These
plots, presented on a per cent (dry exhaust) versus ^ basis, serve to
illustrate this point. On each graph three runs are indicated (four similar
symbols representing the four cylinder $ values for a test condition). These
indicate the typical spread caused by maldistribution. As is well documented
in the literature, values of these three constituents are virtually independent
of engine speed, load, spark timing, and inlet air temperature. Our results
confirm this.
With these three constituents and knowledge of the fuel composition, 4>
can be readily calculated after balancing the chemical equation which represents
the combustion process. The equation form and solution technique, which
provides an internal check on the experimental determination of the three
species, is detailed in the Appendix. The water-gas reaction with a constant
of 3.5 was used to determine hydrogen in the exhaust. Since these constituent
gases are dependent on the stoichiometry only, they serve as an excellent check
on data being generated in new test conditions.
The 4> _ values for the engine show data deviating from the individual
cylinder plots. These composite results for indolene can be found in the
Appendix.
IV. 1.4.3 CO,,, CO, 0,, Comparative Results: Figures 14, 15 and 16 show
comparative individual cylinder results for indolene and methanol. The curves
represent best fit of all the indolene. data and the symbols represent results
for methanol at 2000 RPM. Methanol data points nearly match the indolene
curves except in the rich region where CO- and CO start to diverge. The most
obvious difference, however, is the extension of the equivalence ratio range
when operating on methanol. The lean limit for indolene is close to $ = 0.7,
while for methanol it is closer to
-------
28
o
o
•
(O
o
o
in
o
o
o
o
to
>
o:
O
O
o
o
•
0>
o
o
•
CO
o
o
4
f.
4 +
4
4
V"
J#" *"
?
4-
LEAN
^ ++:
f*.-*.
•ft
_j.
4-
RICH
%
'*&
'7j^
'V
%4t
4 - INDOLENE - ALL RUNS
SPEED LOAD 0QOMP RUN
0 1500 >/3 1.00 U8)
H 2000 2/3 1.13 (5)
A 3000 FULL 0.92 (88)
j.
\
%
I
*\^
^
4
•4-
1-
"0.60
0.70
0.80
0.90
1.00
l.lO
1.20
1.30
1.40
FIG. 1 1: C02 VS EQUIVALENCE RATIO- INDOLENE DATA
-------
29
8
O
o5
8
CO
o
o
N*
o
o
«
o
0
o
o
CO
0
o
CM
o
o
•
0
o
°0
4-
LEAN
4-INDOLENE-ALL RUNS
SPEED LOAD 0C.OMP RUN
• 1500 '/3 .00 (18)
• 2000 2/3 .13 (5)
A 3000 FULL 0.92 (88)
, v ^ti
4- 4
1a&tfjt&4&*l
RICH
4.+V
/
f'
\
f
&
• sf
ijgj^f
$
if-
J~i-
HtU
/)-
?
-1
t
a.
4-
-I.4- +
$
J- *
4>
4-
-
60 0.70 0.80 0.90 1.00 1.10 .20 1.30 1.
0
.40
FIG. 12 CO VS EQUIVALENCE RATIO- INDOLENE DATA
-------
30
0
o
at
o
0
CO
o
o
t^
o
o
10'
o
0
•
a:
g
CM O
o o
8? *
O
0
ro
O
q
CVJ
o
0
o
q
r?
*
•
t
t-
t
\
4-
-1
*
\
\
*+
\
^
J
LEAN
^
*• f
I- i-
t^
'^
?1|
RICH
wfAA- ^ i ^
+ - INDOLENE - ALL RUNS
SPEED LOAD 0QOMP RUNS
0 1500 ^3 1.00 (18)
| 2000 2/9 1.13 (5)
A 3000 FULL 0-92 (88)
-»*•
] "'^^^^j^^f' 4
0-60 0.70 0.80 0.90 1.00 1.10
0
FIG.13 02 VS EQUIVALENCE RATIO- INDOLENE DATA
1.20
1.30
1.40
-------
31
16.00
15.00
14.00
13.00
rlZ.OO
11.00
10.00
9.00
aoo
7.00
RICH
BEST FIT OF
INDOLENE DATA
METHANOL 2000 RPM
6.0O.
~r
0.50 0.60 0.70 0.80 0.90 g 1.00 1.10 1.20 1.30
FIG. 14 C02 VS EQUIVALENCE RATIO- INDIVIDUAL CYLINDER DATA
METHANOL VS INDOLENE
1 .40
1.50
-------
§
o
o
o
o
o
o
<0
o
o
in
o
o
o
o
o
o
o
q
o
LEAN
RICH
BEST FIT OF
INOOLENE DATA
METHANOL 20OORPM
SPARK 23«-50-
LOAD 1/3-FULL
0.50
0.60
1.10
I.ZO
1.30
1.40
1.50
01
N>
FIG. 15 CQ VS EQUIVALENCE RATIO - INDIVIDUAL CYLINDER DATA- METHANOL VS INDOLENE
-------
33
BEST FIT OF
INDOLENE DATA
'0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
0
FIG. 16 02 VS EQUIVALENCE RATIO - INDIVIDUAL CYLINDER DATA - METHANOL VS INDOLENE
-------
34
other engine variables. Figures 17, 18 and 19 display composite exhaust
results for these respective pollutants. The data represents all the indolene
runs. Aldehydes are.calculated as formaldehyde, since the total aldehyde
technique by MBTH method of analysis was used. This method cannot distinguish
among the various aldehydes present in the exhaust sample but it is generally
accepted that formaldehyde makes up 70% to 90% of the total.
General trends show unburned hydrocarbons increasing in the rich region,
aldehydes increasing in the lean region, and oxides of nitrogen peaking on the
lean side of stoichiometric air-fuel ratio. The oxides of nitrogen show the
most sensitive dependency on (and therefore on the degree of maldistribution)
of any of the measured emissions species. The scatter of the data in Figure 19
is due to the combined effects of maldistribution and the influences of other
engine variables (notably the spark timing). By separating all the influencing
parameters, the NO., data becomes amenable to analysis.
Beginning with Figure 19, one important feature of the composite N0»
results is seen clearly. High levels occur nearly to the effective lean misfire
limit ( = 0.8). This indicates that control over N0» by lean engine operation
was not possible. If the maldistribution phenomena is eliminated, control may be
possible, as is indicated in Figure 20, where the individual cylinder results
appear. Low levels of NO., occur at = 0.8 and leaner, and the peak is defined
more clearly. This latter effect is just the result of eliminating the non-
linear averaging of the four cylinder values of NO...
Figures 21 and 22 are speed and load dependency plots of NO., at MBT spark
timing. The coherency of the data for rich <|> indicates that neither parameter
significantly alters the production of NCL. However, in the lean region
significant scatter occurs near the peak NO., equivalence ratios. The scatter
principally results from the effect that maldistribution has on the setting of
MBT spark timing. As previously discussed in Section IV.1.2.1, two cylinders
usually operate at retarded spark, while the other two operate at over-advanced
spark setting. In other words, MBT conditions for the engine are not MBT
conditions for the individual cylinders. The effect that retard or over-advance
has on engine power and NO., emissions is indicated in Figure 23. (The NO., shown
is that being produced by one cylinder operating near peak NO., equivalence
ratio.) Here, MBT spark timing is 31.5 degrees before top dead center (BTDC).
By over-advancing the spark by another 12 degrees to 43.5°BTDC, the NOX has
doubled. At the same time, indicated power has decreased by 4%. Retarding from
31.5° further reduces NOX emissions, but power and fuel economy (not shown)
-------
35
§
•
o
o
00
o
_ o
> N
Q
tr
i °
n 0
1*
^
?8
I"
(E
O
O
OT O
111
U.
s§
z n
K
S
CM
^J
•
O
O
0
0
-ih
i
•
4
+
* +* +
*+
T
4
4
LEAN
J
i-
4
+* + +4
-f +^4+
" 4
.60 0.70 0.80 0.90 1.
RIC^
+ 4+
44
+
+
4
00 1.
4
4
4 ^4++
^ 4
k
INDOLENE-ALL RUNS
4
t
4
+• 4
_4
t
•+
0 1.20 1.30 1.
40
0
FIG. 17; UNBURNED FUEL VS EQUIVALENCE RATIO-COMPOSITE CYLINDER
DATA
-------
36
6
S
0
2 o
is
CO O
V
O
1
5*0
s?
ALDEHYDES (/
0.40
O
O
O
0
g
6c
4
•f
-1
t
•i-
^
•f
"*" -H
+ f *
-1- ^
"'
4
*
LEAN
^.t
•4- * *
+ ^'++^
+
+ +* ,
+ "*1
*'*
RICH
-
*
tt *
•**
-------
37
§
*
8
0$
8
+
C""4"
0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.'
0
FIG. 19: OXIDES OF NITROGEN VS EQUIVALENCE RATIO-COMPOSITE CYLINDER DATA
-------
38
8
o
C\J
8
00
8
.
o o
• o
1 "
\
5 0
0 O
1 O
CM ""
O
z
OT
~ 8
ox«
2
o
o
(O
8
•
*
0
o
CM
8
r\
4
-
4
4-
f
4 +
4>-
+
+
* 4
,
4- *
4 .
r T 4-
t t
f+ 4"
4-
+ +.
4- + 4
4-
f
J-
4 +•
^
+ '
.
f
f
+ 4
4 -1
+ ^
"*"
4
4
4
f """
4
f f i
+ 4-
44.
4- * 4.
+ +
^•^-^
4 +
+ -4-
^
r
"""f1"
"•• +
LEAN
4-
4-
++
4- 4-)-
4*"
4
+ ,
t ++^
+ _^
4 ;
.1- -1
t\ ^
* 4 .
/4
4-
j. f
t$
4*"4- "*" "*"
4"^-
^ +t 4 *
4 +
t-
4 +
4
4
1-
RICH
-4
4-
*"4
^ 4
^
h 4*-
f * j.
4- 4-
f4 4 4
* tj
* J-
* ;*^
**"!
+ + -i-
4+t 4
i4-Hl^ "*" a
4 f ^
J- +
4-
,.
4-
t
i t
v
^t^V*
,4J-f4.+
^ J-
ift+r
4ft- 4^
jC*j.
-£
4 4
H--T *•
INDOLENE - ALL RUNS
*~j. j-+
•1-
•*^ j.+
X ^ rt*
+V +4-^^
f +
-1-
0.60 0.70 0.80 0.90 1.00
I.IO
1.20 1.30 1.40
FJG. 20: OXIDES OF NITROGEN VS EQUIVALENCE RATIO-INDIVIDUAL CYLINDER DATA
-------
39
18.0
16.0
IA f\
I4.O
(C
°. 12.0
o:
i
2m n
IU.U
O
1
o"
z
< 8-°
ox
z
6.0
An
•t.v
2.0
d
+
DA
+D
DO
A
O
D
D
Q
Q D
Q
O
D
n
u
Q
G
4
O £
J • *
i
i
a v
^
0 A
+
a
7
0
LEAN
&
C
+
O a
^A ®
[ °
/\A
'
\
&
RIC^
a
Q
D
+ a
&
a
A *
A t
V
+^
-------
40
18.0
16.0
14.0
12.0
s
o. 10.0
x
o
I
CO
<
8.0
6.0
4.0
2.0
0.0
f.
0
ffi
0 •
A
&
A
0
A A
3
O
A O
A
0
O + "*
•f
+
+
+
* *.
•
*
* A
30*
» ^
° «
A o
* + *d
/«
+*
o
•
, ^ A
• *
• ^A U
CT
*v%>
^ +
+
INDQLENE
2000 RPM
+ ' 14" HG MAN VAC
& 7" HG MAN VAC
0 0" HG MAN VAC
• SPARK RETARD
(WIDE OPEN THROTTLE)
*l
a a +f
o +
0
4
1
Cg 4
V ^
o
07
0.8
0.9
1.0
I.I
1.2
1.3
1.4
FIG. 22 OXIDES OF NITROGEN VS. EQUIVALENCE RATIO-LOAD DEPENDENCY
INDIVIDUAL CYLINDER DATA
-------
41
38
o. 36
34
MBT
METHANOL
$=0.959
RPM = 2000
MAN. WC. 7 " HG
INT. 79 °F
50
FIG. 23: EFFECTS OF SPARK TIMING ON NOX a IMP
-------
42
begin to decrease. (This is a standard technique used in the industry for
control over NO...)
Figure 24 shows the spark advance effect on NOX in more detail. Here, all
engine variables are fixed and spark advance appears as a parameter. Note now
the coherency of the data in comparison to the previous NO., plots. Each set
of our solid symbols (*,•,•,*) represents the individual cylinder
contributions of NO., for a particular engine steady state test point. In
particular, note the four cylinder contributions indicated by A at the far
left of the graph. These represent one engine condition where . varied from
0.73 to 0.86. The spark timing was 33°BTDC. The richest cylinder yields just
over 14 gms N0../ihp-hr while the leanest yields about 1.5 gins N0x/ihp-hr.
In summary, the indolene NO., data indicates that spark advance and
equivalence ratio are the two most important factors affecting the emissions of
NO., in multicylinder engines. The non-optimized spark setting for individual
cylinders, together with variations among the cylinders, leads to large
variations in N0>. on a cylinder-to-cylinder basis. The effects of speed and
load on NO., are obscured by the effects of spark advance and .
IV.1.4.5 Comparative NO.., UBF, and Aldehydes; Due to the increased maldistribu-
tion when operating on methanol (see Figures A3 and A4, Appendix) the setting of MBT
spark timing has an even more pronounced effect on the NO... This is shown in
Figure 25, where the 2000 RPM results appear for the three engine load
conditions. Again, the solid symbols represent the individual cylinder contri-
butions for an engine test point at full load. This graph can be compared with
its indolene counterpart, Figure 22. NO., for MBT is lower with methanol by
roughly a factor of two (note scale change in Figure 25). The possibilities
of operating at very lean for reducing NO., become evident; however, this
depends on the degree to which the cylinder-to-cylinder variations in can be
reduced.
Unbumed fuel, as well as aldehydes, is a wall quench phenomenon and the
description of their behavior in terms of the chosen engine variables is
simplistic. At this time we do not present any explanations for either beyond
the general trends that they exhibited as a function of . Figure 26 is a
comparative plot of UBF for the two fuels at 2000 RPM. The data indicates no
particular benefit for one fuel over the other when presented on this mass
basis. However, on a ppm volume basis,methanol yields about half the UBF
emissions that indolene does.
-------
£
0
(T
I
Q.
I
tn
o
(M
(O
0.70
1.20
1.30
FIG. 24: EFFECTS OF SPARK ADVANCE ON THE OXIDES OF NITROGEN - INDIVIDUAL CYLINDER DATA
-------
44
9.00
8.00
T.OO
IE
o
.,6.00
-------
i
a.
u.
o
cc
a
3
A
A
0
i
M 1
1
Q
i
LEAN
% • |
RICH^
1-
' *.V
«"1
> •" r
*
O
0
a
•
2000 RPM
MBT SPARK TIMING
Q.B H" HO MAN VAC
A. A 7"H6 MAN VAC
O, • CT HO MAN VAC
Q^B.A INOOLENE
O.B.A METHANOL
0.50 0.60 0.70 0.80 0.90 1.00 I-10 1.20 1.30 1.40 I.5O
0.
FIG.26:UNBURNED FUEL VS EQUIVALENCE RATIO-METHANOL VS INDOLENE-COMPOSITE CYLINDER DATA
tn
-------
46
The one aspect of emissions which appears inferior for methanol when
compared to indolene is the aldehyde emissions. Figure 27 shows the comparative
evidence. Methanol gives two to eight times the amount of aldehydes on a
gms/ihp-hr basis. In the extended lean operating range methanol shows even
higher values. However, it appears that an oxidation catalyst would most
likely eliminate this particular species.
IV.1.5 Conclusions and Projected Activities: The initial results for
methanol support earlier work of the authors and other investigators; namely,
that methanol reduces exhaust emissions with the exception of aldehydes and
increases fuel economy relative to gasoline fueled spark ignition engines.
This is made possible in part through the extended lean misfire limit of
methanol and its more efficient combustion characteristics.
However, the combination of typical production carburetion and intake
manifold appears to preclude best engine performance and control over exhaust
emissions because of resulting variations in equivalence ratios among the
cylinders. On the basis of the experimental evidence, this appears to hold
true for both indolene and methanol; however, with methanol the maldistribution
is worse.
In light of the steady state evidence, alternative air-fuel induction
systems will be investigated in an attempt to alleviate cylinder-to-cylinder
equivalence ratio variations. Three systems will be investigated as part of
the second year activity schedule. Two of these systems are described in the
following sections.
Additional tasks included in the second year activity schedule are listed
below:
Task A - Completion of the Methanol Base Line:
A few iterations on the stock carburetor modification will also be
included in the testing sequence.
Task B - Transient Engine Performance:
This will include cold starting, warm-up, and acceleration/deceleration
modes for both indolene and methanol in the four-cylinder engine. This
work will be carried out using the second engine and a fast time response
emissions measurement system. The testing will utilize the stock and the
modified stock carburetor (indolene and methanol, respectively) primarily.
Initial assessment of the alternative fuel-air induction system in these
transient modes will occur during the latter portion of the grant period.
-------
47
o
cO
cr
i
£L
I
O
CJ
V)
LJ
O
>
I
liJ
O
.
I
G
3
*
0 r.
LEAN
*
^
A
>
© "
G
RICH^
I
•
3
©GHHEfe
&
2000 RPM
MBT SPARK TIMING
(D, • 14" HG MAN VAC
A,A 7" HG MAN VAC
(•),• O"HG MAN VAC
(•,*,• -METHANOL)
(0, A.0 - INDOLENE)
•»
6 e°
•
g
m
6
q
c>
0.50 0.60
0.70 0.80
0.90
1.00
1.10
1.20
1.30
FIG. 27 ALDEHYDES AS A FUNCTION OF EQUIVALENCE RATIO-METHANOL vs INDOLENE
-------
48
IV.2 ALTERNATIVE FUEL INDUCTION SYSTEMS
There are two major problems of fuel preparation frequently identified
with venturi carburetor, intake manifold systems. They are poor nebulization
of the fuel and poor mixing with the air. The consequences are loss of perform-
ance and excessive pollutant emissions because the spark advance and air-fuel
ratio cannot be simultaneously optimized for all cylinders for given speed and
load conditions. Evidence of such problems has been presented in Section IV.1.
Two alternative fuel preparation systems have been identified for experimental
study.
IV.2.1 The WHB Shock Wave Induction System: This system is the result of
returning to the first principles in examining the instantaneous air-fuel
requirements of an operating engine and satisfying these requirements with a
minimum of hardware. The basic hardware consists of an intake duct coupled
directly to each intake port through a mounting plate which carries a sliding
throttle mechanism, a metering orifice in the wall of the intake duct, and a
floatless fuel supply system which utilizes a fuel return line for maintenance
of liquid level.
When the intake valve opens and the piston begins its downstroke,
expansion waves propagate outward through the intake duct, initiating air
flow into the duct and fuel flow through the metering orifice. The expansion
waves are reflected from the open end of the duct as compression waves delayed
by the round trip travel time in the duct, so that the resultant pressure
difference at the metering orifice closely approximates the time derivative of
the air mass flow rate in the duct. Since suitable fuels are 600-700 times as
dense as air, the resultant fuel mass flow rate can be made proportional to
the air mass flow rate on an instantaneous basis by proper selection of metering
orifice dimensions. Subsequent wave reflections are prevented by a viscous
acoustical resistance in the form of a narrow slot through the wall of the duct,
providing critical damping for an otherwise resonant acoustical system.
Wave motion in the duct generates a substantial non-turbulent pressure
gradient near the wall; the metering orifice outlet is positioned to maximize
the effectiveness of this pressure gradient in atomizing the fuel issuing from
the orifice. Mean fuel droplet diameters of less than 10 microns have been
produced at 900 RPM engine speed (CFR engine) at wide open throttle in a 33 mm
diameter duct.
Because an intake duct is coupled directly to each port, the mixture
transit time from metering orifice to intake valve can be reduced by an order
-------
49
of magnitude compared to a conventional system with a distributing manifold, so
the fuel droplets have much less time to fall out of the air stream. They will
not recombine into larger droplets or be deposited on the walls of the intake
tract as in a conventional system because no significant changes in mixture
direction occur along their path. Significant reductions in fuel consumption
and/or emission of unburned hydrocarbons and carbon monoxide result from the
improved mixture preparation. Using this system with gasoline as a fuel on a
dynamometer mounted, General Motors Vega engine resulted in a 20% torque
increase at full throttle conditions over the normal engine operating range.
The brake specific fuel consumption curve was lower than the OEM curve for
this comparative testing. Emissions measurements were not measured for a
sufficient number of test points to allow contrasting statements to be made.
Earlier laboratory work using a CFR engine with the same type of fuel
induction system showed it to perform successfully on either gasoline or
methanol by changing the fuel orifice. This combined evidence served as
the basis for selecting this product as an alternate fuel preparation system
to the OEM venturi carburetor, manifold system. The comparative performance
testing using methanol and gasoline fuels is expected to take place in the early
months of the second year of this grant.
IV.2.2 Electronically Controlled Fuel Injection System; During the first three
months of this year a performance and emissions test program was carried out to
evaluate a slant-six Plymouth Valiant equipped with the prototype of the fuel
injection system to be used subsequently in the Pinto engine test program. This
testing was done for Lawrence Livermore Laboratory of ERDA in exchange for a
Pinto engine fuel injection system. The familiarity and operational expertise
of the controller gained by this test program will prove useful for the
subsequent tests which are part of this grant plan.
To date, the electronically controlled fuel injection system mounted in the
Plymouth Valiant has shown good results during steady state tests in controlling
air-fuel ratio and cylinder distribution. The transient response of the system
showed reduction in air-fuel variations by a factor of two in comparison to a
conventional venturi carbureted engine, however, more development and testing
is needed in this area. A prototype of this system is currently being tested
on a laboratory CFR engine for further improvements before mounting the system
on the Pinto test engine. A brief description of the electronically controlled
fuel injection system follows.
-------
50
Figure 28 reveals the major components of the system. The injectors and
other fuel handling components are Bosch products, currently being used by an
American automobile manufacturer. Note that the fuel pressure regulator is
referenced to the intake manifold pressure and therefore a constant pressure
difference will be maintained across the injectors. Consequently, the
injectors are operated either fully on or off with the control of opening and
closing times provided by the electronic controller.
The electronic control unit integrates signals from a manifold pressure
transducer, air temperature sensor, tachometer circuit and water temperature
sensor. The linear output of the manifold vacuum transducer and tachometer
circuit can be modified by their respective function generators to obtain a
linear curve or a curve consisting of three straight line segments with two
break points. The slope and break points are adjustable over a wide range.
The modified air pressure and RPM signals are essentially multiplied together.
The resulting product is then further modified by the air temperature signals
and engine temperature.
The final signal level is proportional to the injection pulse width..
Injection timing is accomplished by a magnetic switch mounted on the
distributor shaft and can be varied over 720 degrees of crankshaft revolution
range.
. The flexibility of control of this system in air-fuel ratio management
and fuel injection timing should be apparent from the brief preceding
description. It has been successfully operated with both methanol and
gasoline in the Valiant vehicle. Hence, it is anticipated that the system
being acquired for the Pinto engine should provide an excellent basis for
comparison of the two fuels with a fuel injection system.
-------
51
FUEL TANK
FUEL
PRESSURE
REGULATOR
INJECTORS
I EACH CYLINDER
INTAKE
MANIFOLD
PRESSURE
ENGINE SPEED
FROM TRIGGER
IGNITION SWITCH
ELECTRONIC CONTROL UNIT
FIG. 28 ELECTRONIC FUEL INJECTION SYSTEM
-------
52
IV.3 COLD STARTING
Cold starting of methanol fueled automotive engines has been accomplished
most frequently by use of an auxiliary gaseous fuel. Both propane and butane
are acceptable. The City vehicle that is now entering its fifth year of
operation has always used propane for cold weather starts. These auxiliary
fuels present no special problems to an experienced operator. However, they
are seen as a nuisance and a source of consumer complaint if they were to be
incorporated in vehicles operated by the public.
Blending of volatile constituents, such as butane, with pure methanol has
o
also been shown to be successful. In Section IV.4, the vapor pressure-
temperature relationship for one such blend is displayed. Even though such a
blend offers a cold start solution, it is still seen by the investigators as
violating a significant methanol fuel safety factor. Methanol does not
explosively ignite at atmospheric pressure and room temperature. When blended
with butane, this quality would be sacrificed. Thus, a primary goal of the
investigators is the catalytic generation of a combustible gaseous mixture for
cold start in conjunction with high quality liquid fuel nebulization to sustain
engine operation after the first few gas fueled cylinder firings.
Progress to date on the cold start aspect of the program has been very
limited except for delineation of the possibility specified. None the less,
this problem is seen as one of the few germane arguments against pure methanol
as a fuel for automobiles. Hence, the investigators believe that it must
receive a significant portion of the research effort in the coming year of
activities.
-------
53
IV.4 INFLUENCES RESULTING FROM VARIATIONS IN COMPOSITION
OF METHANOL BASED FUELS
There are many possible additives which may be combined with methanol
either inadvertently through low cost production or intentionally to seek
improved performance. Methanol's excellence in pure form as an automotive
fuel in terms of performance, thermal efficiency, anti-knock quality and
emissions is such that no additives are needed except for cold starting.
However, it is also frequently considered as a gasoline extender by using
it in blend form with gasoline. Hence, evaluation of additives to methanol
resulting in methanol based fuels must address the following questions:
(1) Is cold starting enhanced?
(2) Are emissions or performance degraded?
IV.4.1 Volatility of Fuels; In order to have comparative data, the measurement
of the volatility of a fuel must be done under controlled conditions. The two
procedures which are being used in this study were ASTM D 86, Distillation of
Petroleum Products, and ASTM D 323, Vapor Pressure of Petroleum Products (Reid
Method). Since petroleum products consist of mixtures of hydrocarbons with
different boiling points, the temperatures at which different percentages of
the fuel have evaporated give a measure of its volatility. This forms the
basis for ASTM D 86. The other procedure depends on the vapor pressure of a
fuel, such as gasoline, which is measured at 100°F in a pressure bomb having
a 4 to 1 volume ratio of air to liquid. This procedure is designated as
ASTM D 323, Reid Vapor Pressure (RVP).
The relationship between volatility as measured by the ASTM procedures
g
and performance criteria have been defined and are summarized in Table 2.
It should be noted that the characteristics in Table 2 are applicable to
gasolines. Some of them can be seriously questioned for methanol based fuels.
However, as a starting point for comparative evaluaticns, these measurement
procedures will be used and then subsequently challenged where it seems
appropriate.
IV.4.2 Types of Fuels Tested: The fuels that have been examined to date for
volatility comparisons are given in Table 3. These include the gasoline used in
the test program and some methyl alcohol blends which may be considered for
cold start enhancement over straight methanol or for use as gasoline extenders.
The indolene-methanol blends were tested to determine the effect of small
versus large ratios of gasoline to methanol in gasoline-methanol blends. The
5% butane-methanol blend was prepared as a result of the work of Tillman on
use of butane in methanol to improve cold starting.
-------
54
TABLE 2: VOLATILITY CHARACTERISTICS AND PERFORMANCE
Low Temperature
for 10% Evaporated
and High RVP
Low Temperature
for 50% Evaporated
High Temperature
for 90% Evaporated
Improved cold starting
Vapor lock under hot
operating conditions
Increased vapor forma-
tion in fuel tanks and
carburetors
Better warm up
Better acceleration
Increased tendency for
carburetor icing and
resultant stalling
Improved fuel economy
Resistance to knock
Poorer mixture distri-
bution in intake
manifold
Increased combustion
chamber deposits
Excessive varnish and
sludge deposits in
engine
TABLE 3: FUELS USED IN VOLTALITY COMPARISON STUDIES
Chevron unleaded
Chevron unleaded
5% indolene
95% indolene
4% isobutanol
5% butane
winter grade
summer grade
95% methanol
5% methanol
16% methanol
95% methanol
- 80% Chevron unleaded
-------
55
The isobutanol-methanol-gasoline blend is one that is being studied in
Sweden, and purports to have a suitable distillation curve and no problems
with phase separation or vapor lock.
The results of the ASTM distillation are plotted in Fig. 29. The
addition of 5% methanol to indolene causes a depression of the distillation
curve as compared to indolene alone to about 20% evaporated. The curve then
rises to meet the indolene curve. A similar initial depression is noted with
the Swedish blend. For a better comparison, the ASTM distillation curve data
points for 10%, 50% and 90% evaporated and the final boiling point are given
in Table 4. The Reid vapor pressure data is given in Table 5. Note that the
addition of 5% methanol to indolene caused an increase of a little over 1 psi
in the Reid vapor pressure. The addition of 16% methanol (and 4% isobutanol)
to Chevron unleaded gasoline increased the RVP by about 3.5 psi. Of the four
methanol blends that were prepared, the 95% indolene - 5% methanol and the
4% isobutanol - 16% methanol - 80% Chevron unleaded seem to most closely
approximate the desired performance characteristics as defined in Table 2.
The actual performance with 100% methanol has proven satisfactory in test
engines, street vehicles and racing cars, even though it seriously violates
the criteria of Table 2 by having a poor RVP and a constant boiling temperature
12
rather than a distillation curve. One report describes a malfunction
involving methanol-gasoline blends which was labeled "acceleration vapor lock".
It was attributed to departure from their suggested vapor lock index (VLI).
VLI = RVP + 0.13% distilled @ 70°C
An alternate explanation could be the leaning effect associated with adding
methanol to the fuel without modification of the accelerator pump. Vapor lock
is a dynamic phenomena involving heat transfer rate, fuel flow rate,vapor
pressure-temperature relationship and heat of vaporization. For closely
related fuels,such as gasoline blends, the VLI may be satisfactory. However,
a more carefully developed index capable of handling a broad range of fuels
should contain at least some of the factors just mentioned.
Cold start problems with pure methanol are known to exist. Blending is
one way of avoiding this problem while at the same time extending gasoline
supplies. From the results of these initial volatility tests, the two fuel
blends that appear to be prime candidates for further evaluation are the 95%
indolene-5% methanol, and the 4% isobutanol-16% methanol-80% Chevron unleaded.
-------
400
WINTER CHEVRON UNLEADED
SUMMER CHEVRON UNLEADED
INDOLENE
5% INDOLENE, 95% METHANOL
5 % METHANOL , 95 % INDOLENE
4% ISOBUTANOL, 16% METHANOL, BO CHEVRON UNLEADED
5% BUTANE ,95% METHANOL
40 50 60
PERCENT EVAPORATED
90
100
FIG. 29 ASTM DISTILLATION
Ul
-------
TABLE 4: ASTM DISTILLATION, °F
%
Evaporated
10
50
90
FBP
Chevron Chevron
Unleaded Unleaded
Summer Winter Indclene
.124 108 "l39
202 179 221
334 308 331
400 380 392
TABLE 5:
Fuel
Chevron Unleaded-Winter
Indolene
95% Indolene, 5% MEOH
5% Indolene, 95% MEOH
4% Isobutanol, 16% MEOH,
80% Chevron Unleaded-Winter
5% Butane, 95% MEOH
100% MEOH
4% Isobutanol
5% Indolene 95% Indolene 16% MEOH
95% MEOH 5% MEOH 80% Chevron
148 120 114
150 223 184
152 331 334
170 384 400
REID VAPOR PRESSURE
Reid Vapor Pressure
(psi)
7.65
8.75
9.90
5.20
11.10
15.10
3.85
5% Butane
95% MEOH
148
150
152
170
in
-------
58
It is evident that use of distillation equipment and RVP procedures
provides some useful information. Hence, examinations of additional fuel
candidates using these procedures will continue during the coming year with
emphasis on combinations involving methanol based fuels resulting from low
cost,high volume production of methanol. It is also evident that further
study is required to present a vapor lock index that can correctly handle
methanol-gasoline fuel blends. This, too, will be part of our continuing
investigation. Engine experiments using the more desirable fuel candidates
resulting from these studies will subsequently follow.
-------
59
IV.5 ENGINE WEAR
The question of whether there is an increase in engine wear with the
use of methanol has not been clearly answered. A recent survey has
summarized information indicating both viewpoints. In one case, excessive
14
wear of piston rings and cylinders was reported. Other information indicated
that wear is no greater with alcohols than with gasoline. With this conflict-
ing background, it was considered important to incorporate in our comparative
methanol/gasoline fuel study the relative effects of these fuels on wear.
Wear data has been accumulated from the start of the current program including
the break-in period of both engines with gasoline fuels. By following the
wear history from the beginning of the tests a comparison could be made
between wear rates occurring during successive gasoline, methanol and methyl
fuel operations on the same engine.
The procedure selected for the determination of internal wear of the test
engine is based on the analysis of the crankcase oil for specific wear metals
by atomic absorption spectrophotometric techniques. The procedure is based
on the fact that the trace elements present in the crankcase oil are an
indication of wear of the component parts of the engine made up of those
particular metals. This is a well known procedure and has been used routinely
to permit forecasting of impending failure of specific parts of an engine.
The materials that would be expected to wear in the 2.3 litre Pinto test
engines used in this study are shown in Table 6. The elements considered most
likely to show up in the engine crankcase oil>as engine wear proceeded,were
copper, iron, lead, chromium, aluminum and tin. Hence, they were selected
for analysis.
IV.5.1: Pinto Engine Wear Results: Wear data has been accumulated for over
256 hours of gasoline fueled operation on the Pinto engine number 1 and after
69 hours (including 48 hours with gasoline and 21 hours with pure methanol)
on the Pinto engine number 2. In most cases crankcase oil samples were taken
every 10 hours and analyzed for the wear metals by atomic absorption
spectrophotometry.
The procedure for the calculation of accumulated wear metals included
adjustment of the chemical analysis data to compensate for the addition of
fresh oil. The adjusted values were to be as if no oil were removed or had
leaked out. Therefore, the values should show a continual accumulation of
wear metals.
-------
60
TABLE 6: MATERIALS IN PINTO ENGINE
18
Journal and Main Bearings
Cam
Block
Piston
Piston Rings.
Crankshaft
Cam Bearings
Valves
Rocker Arms
Oil Pump
Distributor and Pump Bearings
Distributor and Pump Shaft
- Cu-Pb Alloy Plate
- Cast Iron, Phosphate Coat
- Cast Iron, Phosphate Coat
- Ni and Al
- #1 - Cast Iron - Molycoat
#2 - Cast Iron - Phosphate
#3 - Cast Iron
Oil - Steel with Cr and Black Oxide
- Nodular Iron
- Al, Sn, Si, or Al, Pb, Cu
- Exhaust - Austenitic, Cr Plate Stem
Al Diffused Face
Inlet - Austenitic, Mn, Phos. Stem.
Al Diffused Face
- Cast Iron with Mn
- Al, Cu, Zn, Mg
- Steel Backed Babbit
- Cast Iron Phosphate Coat
-------
61
The data for the Pinto engine number 1 is summarized in Table 7 and
the data for the Pinto engine number 2 is shown in Table 8. The significant
wear metals in all cases were copper, iron and lead. The accumulated wear
of these three metals during the break-in period of the Pinto engine number
1 is shown in Fig. 30. A plot of the entire 256 hours of operation is shown
in Fig. 31, which shows correlations of the adjusted amount of wear metals in
the oil samples with events occurring during operation of the engine.
Specifically, the increase in lead content shows interesting correlations
with heavy load operations at 60 to 80 hours and a sharp increase at 180
hours when a loss of oil pressure occurred. There was a steady increase in
iron and copper wear prior to the detection of scuffing of the cylinder
walls at about 50 hours of operation.
Except for lead, there was a decided drop in the major wear metals after
the change of the oil filter at 182 hours. It is possible that the new oil
filter picked up some of the wear metals left in the crankcase oil which had
not been picked up by the old oil filter. Why the lead continued to increase
while the copper and iron decreased might be attributed to either continued
wear of a component containing lead or a preferential pick up of the copper
and iron by the filter. The wear study on the Pinto engine number 1 included
only gasoline fueled operations up to the time of this report.
The wear data for copper, iron and lead for the Pinto engine number 2
are plotted in Fig. 32. For this engine, the gasoline-fueled operations ceased
after 47 hours. There was no change in the trend after the change over to
methanol. However, there is insufficient operating time with methanol to
expect any significant change. Future operation of both engines with
methanol should reveal whether methanol causes an increase in wear above that
experienced with gasoline.
-------
62
TABLE 7: ACCUMULATED WEAR METALS IN ENGINE OIL OF PINTO ENGINE NO. 1
(MG/1000 CMS)
Events
Oil Leaks
Cyl . Scuffing
Oil Press. Drop
Oil Fltr. Changed
Oil Change
Oil Leak
Extensive Leak
Oil Leakage
Sample
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
17A
18
19
20
21
22
23
24
25
Operation
Hours
10
20
30
40
50
60
82
90
101
110
120
130
140
150
160
172.75
180.25
180.65
182.25
185.25
198.05
215.9
225.3
235.85
245.6
256.5
Copper
Cu
32
34
33
—
38
29
24
21
25
23
21
23
23
21
21
23
21
18
17
10
15
11
11
12
12
11
Iron
Fe .
16
20
22
—
31
27
23
20
24
23
22
25
27
25
25
28
25
21
27
17
27
22
25
25
27
27
Lead Chromium
Pb Cr
4.1
5.3
6.3
--
8.3
6.8
19
20
22
20
21
26
31
32
32
33
32
30
40
29
42
37
40
45
48
45
0.2
0.2
0.2
0.2
0.2
0.2
0.5
0.4
0.7
0.8
0.8
1.5
1.9
2.4
2.1
2.2
2.3
1.5
0.9
0.2
0.2
1.6
2.0
2.4
2.8
2.8
Aluminum Tin
Al Sn
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
-------
63
TABLE 8: ACCUMULATED WEAR METALS IN ENGINE OIL OF PINTO ENGINE NO. 2
(MG/1000 CMS)
Sample
Events No.
1
2
3
4
5
Oil Changed 6
7
Methanol Fuel 8
9
Operation
Hours
3.42
6.07
8.05
10.02
19.35
30.05
47.77
59.22
69.38
Copper
Cu
30
34
32
34
40
43
49
51
56
Iron
Fe
15
16
17
18
26
30
43
50
54
Lead
Pb
13
15
15
16
19
21
25
27
29
Chromium Aluminum
Cr Al
0.17
0.24
0.36
0.48
0.74
0.76
0.93
1.01
1.12
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
Tin
Sn
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
< 2.0
-------
CYLINDER
SCUFFING
10
20
30 40
OPERATING HOURS
50
60
70
80
FIG. 30 WEAR OF PINTO ENGINE NO. I DURING BREAK-IN
-------
40
60
80
100 120 140 160
OPERATING HOURS
180
200
220
240
260
EVENTS
I. EXTENSIVE OIL LEAKAGE
2. CYLINDER SCUFFING DETECTED
3. OIL PRESSURE DROP
4. OIL FILTER CHANGED
5. OIL CHANGED
6. OIL LEAKAGE
7. EXTENSIVE OIL LEAKAGE
8. OIL LEAKAGE
FIG. 31 WEAR OF PINTO ENGINE NO. I
en
-------
30 40
OPERATING HOURS
50
60
70
80
FIG. 32 WEAR OF PINTO ENGINE NO 2
a\
-------
67
IV.6 THERMOCHEMICAL ENGINE PROCESS MODELING
Effective research relative to a complex phenomenon, such as the thermo-
chemical events in an internal combustion engine, requires an intelligent mix
of experimentation and analytical modeling. Modeling of these processes can
provide an explanation for the trends in pollutant emissions and engine perform-
ance produced by the experimental Pinto test engine and found in the literature.
This analytical process has been accomplished by creating a thermodynamic and
chemical kinetic computer model of the actual processes occurring in an S.I.
engine.
The combustion process in an internal combustion engine is a complex
phenomenon which directly relates to such operating variables as power output,
thermal efficiency, exhaust composition, and cold starting. In particular,
exhaust emissions are affected by engine speed, load, air-fuel ratio, spark
timing, but these effects cannot easily be predicted. It is therefore necessary
to accurately model the thermochemical events occurring during an actual engine
cycle.
19 20 21
Recent investigations have developed fairly good S.I. engine simulations ' '
but they do not include vaporization of liquid fuel (methanol's low pollution
characteristics are somewhat linked to its high heat of vaporization) or complete
chemical kinetic combustion reactions which are essential in accurately studying
pollutant formation (especially in the quench zone). We have included these
important considerations in modeling of a jet turbine combustor.
Researchers at the University of Santa Clara have been developing an Otto
cycle computer model with complete chemical kinetics to predict the kinetic
combustion and pollutant formation processes in an S.I. engine. This model
consists of two programs which work hand-in-hand to model the S.I. engine's
thermochemical processes. The first program is a chemical equilibrium model of
the Otto cycle to predict the thermodynamic properties of the gases during the
22
cycle. This program was developed from a NASA combustion and rocket program.
It has been modified to contain the following simulations which, are diagramed
in Fig. 33. During the intake stroke, liquid fuel and air are mixed and heated
in the intake manifold, allowing an equilibrium amount of vaporization of the
fuel. This heated charge is then induced into the cylinder where it is mixed
completely with the exhaust residual. Further vaporization of the liquid fuel
occurs at this point. As the intake valve closes, the mixture is compressed
isentropically and adiabatically to the point of ignition. By this time,
the fuel is completely vaporized. The unbumed charge is considered to be
-------
FACTORS BEING CONSIDERED IN COMBUSTION MODEL ING
FUEL & AIR
(A) INTAKE STROKE
0=0
ASSUMPTIONS
1. LIQUID FUEL AND AIR MIXED AND
HEATED IN INTAKE MANIFOLD
2. LIQUID FUEL ALLOVED TO VAPORIZE
3. CIIAPC.F IMXED KITH EXHAUST RESIDUAL
4. FURTHER VAPORIZATION OF LIQUID
FUEL
1. INTAKE VALVE CLOSES
2. ISENTROPIC COMPRESSION OF CHARGF
TO POINT OF IGNITION
3. NO HEAT TRANSFER
4. NO REACTION OF FUEL AND AIR
5. LIQUID FUEL COMPLETELY VAPORIZED
PISTON TOP VEW
BURNED T*M T- UNBURNED
SEGMENTS \JS SEGMENT
,QH
I. FLAME FRONT TRAVELS IN MODIFIED
CYLINDRICAL BURNING PATTERN
2. CONSTANT PRESSURE COMBUSTION OF
EACH ELEMENT, HEAT TRANSFERRED
TO CYLINDER FROM BURNED SEGMENTS.
ISENTROPIC COMPRESSION TO NEK
VOLUME
3. UNBURNED SEGMENT CONSIPFr.EU NON-
R FACT I NT
4. CHEMICAL KINETICS OF BUPXED
SEHMENTS CHEMICALLY PETEPVISE
SPECIES CONCENTRATIONS
1. I'FAT TPANSFPF TO CYLINDER
2. EXPANSION OF r.,\srs TO NEK VOLUJIF
3. CHEMICAL KINETICS DETERMINE
SPEC US CONCENTRATION
(B) COMPRESSION STROKE
(E) EXPANSION STROKE
^•Jk^^L
SPARK ENERGY-»=^ |fr ;
1. SPARK ENERGY ADDFD TO CHARGE
ENFRT.Y OF FIRST INCREMENT
:. FIPST INCREMENT OF FUEL AIR
MIXTURE COMBUSTED
(C) IGNITION
FIGURE 33
(F) BLOW. DOWN
1. EXHAUST VAI.M OPFS'S
AUiADATir i.'TNir.onc BI.OK
OF HASIT 10 ATMOSPI'Eni: PRFSSIIRT
CHEMIC.M UNT.TICS 1)1 FIRM INI
SPI:.CII:.S CONCI NTRA1 ION
oo
-------
69
frozen chemically during this portion of the cycle. The combustion of the
fuel and air is initiated by the spark firing. Ignition energy is added to
the chemical reaction which starts combustion. The combustion process modeling
in this program allows for finite-rate combustion, heat transfer to cylinder
walls, turbulence, and piston motion. The finite rate combustion is
23
considered to be of the stratified zone type where each burned segment of gas
is assumed to be separate with homogeneity only within the segments. The flame
front, which expands in a modified cylindrical pattern, has its motion defined
*\ A
by a modified Semenov flame speed equation coupled with a turbulent factor
25
based on inlet velocity. In the combustion processes each new segment of
unburned gas goes through a constant pressure combustion, followed by a constant
pressure heat transfer from all burned segments to the head, piston crown, and
cylinder walls. This in turn is followed by compression of all segments to the
new volume determined by the piston motion. By the time the flame front has
completely swept the cylinder, consuming all the unburned charge, the expansion
stroke has begun. During the remainder of this stroke heat is transferred to
the cylinder and then the gases are expanded to the new volume defined by the
piston motion. This is continued until the point where the exhaust valve opens,
where an isentropic, adiabatic blow down of exhaust products to atmospheric
pressure occurs..
The other complementary computer program is a complete chemical kinetics
program similar to the one used in Reference 2. It follows the equilibrium
program from the point of ignition through the blow down process with a
complete set of kinetic reactions and rates. Thus, the thermodynamic properties
of the gases are determined by the equilibrium program, while the exhaust
composition is determined by the kinetics program.
The reaction set for the combustion of methanol has previously been
developed in Reference 2. This reaction set has provided good results in our
gas turbine simulation and therefore it can be used directly in predicting
combustion reactions in the S.I. engine model. Since methanol is a relatively
simple molecule chemically, the combustion reaction set is complete. The
complexity of gasoline combustion, however, requires some simplification.
First, gasoline will be simulated by isooctane, CgH.g, since its thermodynamic
properties are similar to actual gasolines. Second, the reaction set for
isooctane combustion will be developed using a global combustion model similar
to that developed for octene in Reference 2. in a global model the fuel and
air react to form simple hydrocarbons, carbon monoxide, and simple free
-------
70
radicals, thereby ignoring intermediate pyrolysis reactions in fuel combustion.
Experimental ignition delay data obtained from Reference 26 win be used to
determine the initiation reaction rates for the octane global combustion model
reaction set.
As a first test of the S.I. engine model, methanol was combusted with air
in stoichiometric proportions at the conditions described in Table 9. As can
be seen from that table the agreement between the predicted and actual results
is good, especially for the NO,, exhaust emissions. Therefore, it is expected
that with further fine tuning of the computer model this can be used to predict
the effects of varying the parameters currently being used in the methanol test
matrix. This fine tuning might include such additions as valve timing, quench
zone modeling, a better turbulent flame speed model and more accurate tuning
to the Pinto test engine.
Once predictive confidence has been established, the S.I. engine simulation
becomes extremely useful in studying the effects of changing engine variables
on the exhaust composition, especially NO., and CO. It is possible to study
combinations of liquid and gaseous fuels to determine the degree of liquid fuel
vaporization and the heats required to maintain a given inlet mixture temperature.
This is a very important feature since methanol's high heat of vaporization
radically affects the engine's performance and emissions characteristics. The
effects of changing air-fuel ratio, ignition timing and compression ratio can
also be examined. The net result is a low cost means of studying a complex
interactive thermochemical cycle of events. It should greatly reduce the
expensive experimental hardware modifications required to find the best mix of
engine performance and emissions suppression for a methanol fueled engine.
-------
TABLE 9
71
INPUT PARAMETERS:
*
Fuel
Intake Manifold Temp.
Intake Manifold Pressure
Exhaust Residual Fraction
Engine Speed
Ignition Timing
ENGINE DATA:
ihp
isfc (Ib/ihp-hr)
nvol %
"th %
EXHAUST DATA (molar %, dry
CO
co2
°2
NO,
= 1.00
= Methane 1
= 283.11°K
= 1 atm
= 0.0404
= 2000 RPM
= 30°BTDC
Predicted Actual*
12.0 12.4
0.905 0.887
78 80.5
32.8 33.6
basis):
Predicted Actual**
0.12 0.49
14.90 14.31
0.12 0.28
0.03 0.41
1882 ppm 1899 ppm
*Taken from Pinto test engine runs 517 and 518.
**Taken from Pinto test engine run 518, cylinder 3, = 0.999.
-------
72
V. REFERENCES
1. H. G. Adelman and R. K. Pefley, "Methanol as an Automotive Fuel", presented
at the AIChE 80th National Meeting, Boston, Massachusetts, September 7-10,
1975.
2. H. G. Adelman, L. H. Browning and R. K. Pefley, "Predicted Exhaust Emissions
from a Methanol and Jet Fueled Gas Turbine Combustor", AIAA Paper No. 75-1266,
presented at AIAA/SAE llth Propulsion Conference, Anaheim, California,
September 29 to October 1, 1975.
3. G. Blair Martin, "Evaluation of NOX Emission Characteristics of Alcohol
Fuels in Stationary Combustion Systems", presented at the AIChE 80th
National Meeting, Boston, Massachusetts, September 7-10, 1975.
4. L. W. Huellmantel, S. G. Liddle and D. C. Hammond, "Combustion of Methanol
in an Automotive Gas Turbine", 1975 GMR Symposium: Future Automotive Fuels,
Warren, Michigan, October 6-7, 1975.
5. E. Faltermayer, "The Clean Synthetic Fuel That's Already Here", Fortune
Magazine, Vol. XCII, No. 3, September 1975, pp. 146-154.
6. G. D. Ebersole and F. S. Manning, "Engine Performance and Exhaust Emissions -
Methanol Versus Indolene", SAE Transactions, Vol. 81, Paper No. 720692 (1972).
7. B. A. D'Alleva and W. G. Lovell, "Relation of Exhaust Gas Composition to
Air-Fuel Ratio", SAE Journal, Vol. 38, No. 3, March 1936, pp. 90-96.
8. R. M. Tillman and R. G. Jackson, "Automotive Uses of Methanol Fuel",
presented at the Engineering Foundation Conference: Methanol as an Alternate
Fuel, Henniker, New Hampshire, July 1974.
9. SAE Information Report, "Automotive Gasoline - SAE J312b", 1975 SAE Handbook,
Society of Automotive Engineers, Inc., Warrendale, Pennsylvania, p. 439.
10. R. M. Tillman, 0. L. Spilman and J. M. Beach, "Potential for Methanol as
an Automotive Fuel", SAE Paper No. 750118, Society of Automotive Engineers,
Detroit, Michigan, February 24-28, 1975.
11. "The Swedish Methanol Project with Operating Plan", The Swedish Methanol
Development Company, Stockholm, October 1, 1975.
12. "Methanol-Gasoline Blends: How Promising Are They?", Automotive Engineering,
Vol. 82, No. 12, December 1974.
13. R. R. Adt, Jr., R. D. Doepker, and L. E. Poteat, "Methanol-Gasoline Fuels
for Automotive Transportation - A Review", U.S. Environmental Protection
Agency, Alternative Automotive Power Systems Division, Ann Arbor, Michigan,
November 1974.
14. R. M. Tillman, data presented at Bureau of Mines sponsored technical
meeting, Denver, Colorado, September 5, 1974.
15. J. D. Rogers, "Ethanol and Methanol as Automotive Fuel", Report No. P813-3,
E. I. DuPont De Nemours and Co., Inc., 1973.
-------
73
16. E. A. Means and D. Ratcliff, "Determination of Wear Metals in Lubricating
Oils by Atomic Absorption Spectroscopy", Atomic Absorption Newsletter,
Vol. 4, No. 1, January 1965, pp. 174-179.
17. R. L. Klug, "Scheduled Oil Sampling as a Maintenance Tool",Paper No. 720372,
presented at Earthmoving Industry Conference, Society of Automotive
Engineers, April 1972.
18. M. Stroben, Personal Communication, Ford Motor Company, Detroit, Michigan,
November 18, 1975.
19. F. J. Zeleznik, "Combustion Modeling in Internal Combustion Engines",
prepared for Combustion Science and Technology, December 20, 1974.
20. H. Hiroyasu and T. Kadota, "Computer Simulation for Combustion and Exhaust
Emissions in Spark Ignition Engine", 15th Symposium on Combustion,
Combustion Institute, 1974, pp. 1213-1223.
21. G. G. Lucas and E. H. James, "A Computer Simulation of a Spark Ignition
Engine", SAE Paper No. 730053, January 8-12, 1973.
22. S. Gordon and B. J. McBride, "Computer Program for Calculation of Complex
Chemical Equilibrium Compositions, Rocket Performance, Incident and
Reflected Shocks, and Chapman-Jouguet Detonations", NASA Report No. SP-273,
1971.
23. G. G. Lucas and K. Varde, "Analysis of Nitric Oxide Formation in Spark
Ignition Engine with Heat Transfer and Effect of Ignition Point", SAE
Paper No. 740189, February 25-March 1, 1974.
24. NASA Report 1300, "Basic Considerations in the Combustion of Hydrocarbon
Fuels with Air", prepared by Propulsion Chemistry Division, Lewis Flight
Center, 1957.
25. B. S. Samaga and B. S. Murthy, "On the Problem of Predicting Burning
Rates in a Spark Ignition Engine", SAE Paper No. 750688, June 3-5, 1975.
26. B. P. Mullins, "Studies on the Spontaneous Ignition of Fuels Injected into
a Hot Air Stream", National Gas Turbine Establishment, England, Report
Nos. 89, 90, 97 and 107, 1952, Fuel, 32, 343 (1953).
-------
74
APPENDIX
Equivalence Ratio Determination;
Calculation of the equivalence ratio () was based on measurements of
the molar concentration of the exhaust constitutents. Known fuel composition
combined with these molar exhaust concentrations provides the basis for the
solution of the combustion reaction equation. Redundant exhaust emissions
instrumentation permits cross-checking (two solutions) of the equation. The
difference in calculated A/F (or ) from these two chemical balances gives
an indication of the error introduced by the emissions analyzers.
The combustion of a fuel with air may be expressed in the following
general form (assuming the fuel to be composed only of carbon, hydrogen and
oxygen)
a C H Oz + b (02+3.76N2) -*• c C02 + d CO + e H20 + f 02
(1)
+ g H9 + h (3.76)N,+ i CH 0 + j NO, + k HCHO
£ £» A y z £•
where the coefficients a through k are molar concentrations or molar percents.
The last three terms on the right side of the equation are unburned fuel,
oxides of nitrogen (as NO-), and aldehydes (as formaldehyde), respectively.
As a general rule, HCHO is so small that it is within the measurement
uncertainty of the principal combustion species of C02, CO, and 02. Therefore,
it can be neglected in balancing the equation.
C02, CO, 02, C HO- and N02 (c, d, f, i, j known) were measured directly.
H_ was calculated by use of the water-gas reaction and an equilibrium constant
of 3.5:
d CO + e H20 j c C02 + g HZ
where
[CO] x [H20] .
v _ _*• _ - ° e
Keq * [C02] x [H2] ' c-g
The constituent species in the equation are expressed as moles or mole
fraction. Equations (1) and (2), together with the measured species, provide
a like number of equations and unknowns. The unknown coefficients (a, b, g, h)
may now be calculated by conservation of the constituent atoms:
-------
75
Fuel (carbon):
a = 2p + i (3)
Water (hydrogen):
e = (a»y - 2'g - yi) ^
Substituting (2) into (4) to eliminate g yields,
2(1+0.285)'-)
Air (oxygen and Nitrogen):
(2-c + d + e + 2'f
(5)
h = b -
The air to fuel ratio can now be calculated. In terms of the measured
species and known fuel composition, this carbon balance procedure yields,
[2-c + d + - + 2.f + i-z + 2-j]
II. .ir .... 2(1.0*0.2857-1)
Ibm fuel ' r(£l£!+11 MWfuel
*• x ^
(8)
In the experimental emissions measurements the combustion water was
condensed from the exhaust sample stream before measurement of C02, CO, and 0.
This was at 33°F, which yields 0.006 mole fraction of water in the sample
stream leaving the condenser. A new balance of the chemical equation can be
derived which will yield a check on Equation (8).
This A/F we term a nitrogen balance and it requires an iterative scheme
for determining the unknown coefficients of Equation (1). Conservation of
constituent atoms begins with a guess of the combustion water mole percent
(TH20):
Air (nitrogen and oxygen) (dry balance):
. (100% - [c + d + 0.6 + f + g + i + j])
n = n -r -if
-------
76
Substituting (2) into (9) to eliminate g yields,
f • r , ,
h = (C'5.5)(1-TH20/100)
and
b = h + (11)
Conserving hydrogen atoms yields a new estimate of the water,
dH2°
fa-v - F2-
(c'3.
5)(l-TH20/100)
e is then compared with TH20. If the two are within a chosen tolerance,
then the calculations are complete. Otherwise, the value calculated in (12)
is substituted into (10) and the calculations are repeated. After the
sufficient number of iterations, the air to fuel ratio can be calculated
using (11) and (3) and the respective molecular weights.
Ibm air .. _, 3.76 air
= 4.
fuel -•"- (c^d) MWfugl
(13)
Equation (13) can be compared with Equation (8). The difference in the
two A/F's is indicative of the combined measurement error in C02, CO, and 02.
-------
77
ENGINE SETTINGS'
2000 RPM
14" HG MAN VAC
SPARK - 35° BTDC
0 ~ 1.06
0compo«tt
0.98
RUNS:
I 2 3
CYLIKDER NUMBER
O-I32 «-45 A-125
B-124
• -89
A-48
8
111
i
u.zz
0.20
o.ie
8
FIG.
Start
—
—
^^
^-~ — '
— — ~~~~
48
lnd<
488 124
I«M Mcthanol-
li'25
m
^
,132
0 100 ISO 200 250 300 350
ENGINE HOURS
A-l MALDISTRIBUTION 8 PERFORMANCE TRENDS
FOR INDOLENE
REPETITIVE TEST POINTS
-------
78
o
£
o
il
u.
u*
I
h-
u
K
.300
.275
.250
.225
.200
.175
.ISO
.125
0.7
LEAN
RIC
\
14
\
\
INOOLENE FUEL
RPM
3SOO
\
" HG MAM VAC
\
0.8
0.9
1.0
0
I.I
1.2
1.3
1.4
FIG. A2: TRENDS IN BRAKE THERMAL EFFICIENCY - PART AND FULL
LOAD COMPARISONS AT VARIOUS SPEEDS
-------
48
46
| 44
i Az
I 40
ui
INDOLENE | M
2000 RPM 36
0" HG MAN VAC
0 CYL
0 OOMP L0
0.9
(
33
1 32
S '°
INDOLENE 5 M
2000 RPM 28
7" HG MAN VAC u
acYL .0
0COMP
0.9
0
IB
17
L
INDOLENE g 14
2000 RPM S
14" HG MAN VAC "
0CVL
0 COUP1'"
0.9
0
/
2J
4 1
/
/
'
4
1
i
3
3.7 0.8
i
•
LEAN
/
'
4
I
j
3 1
4
i J
2 4
0.9
/
/
4
Z
S
RICH
^
\
i
l
i
1.0
/
/
LEAN
3
2
1
4
2
1
4
^—
1. 1
x
RICH.
4
3
2
. .
2
1
_4j
^•^V
1
4
.4
3
1
2
1.2 1.
^*
-~^
2i
2
1
: :i
:7 0.8
2
3
!
1
1 ' 3
4 4
0.9
/
/
3
2
.0
LEAN
S
/
'
l?
.7 aa
0.9
2
1
3
.
2
)
1
a
•
RICH^
s*
*--
^
III
1 V
.0
L"
T"
»^
2
U
4
i=r
1.2 1.
I.I
3
t
4
9
t
4
1.2
79
FIG. A3: BRAKE HORSEPOWER AND MALDISTRIBUTION-INDOLENE
-------
80
METHANOL
2000 RPM
0" HG MAN VAC
METHANOL
2000 RPM
7" HG MAN VAC
47
45
j
41
39
1.30
1.15
1.00
0.85
0.70
35
33
29
27
25
I.I
PCYL
tCOMP
1.0
METHANOL
2000 RPM
14" HG MAN VAC
flCYL
0.9
18
16
14
12
I.I
I.O.
0.9
LEAN
RICH
13
2
4!
2
1.3
-
4
2
1.3
4
2
3
1
4
2
3
1
4
3
2
1
4
3
2
1.4
3
2
1
4
3
2
1
4
LEAN
RICH
LZ
3
4J
2
1
3
4
1
2.4.3^
f? i]
3 S
4 4-1
1
2
2
3
3 4J
4
1
2
3
4
2
3
4
2
3
1
4
^
^^
^~
LEAN
RICH
*
3
\\
4
1
P 1
h *
li '
3
2
4
1 I
3
3
4
1 2
2 .
4 '
3
4
l£
fj
2.41
'
2
.
4
.2
3
4
Ll
3
"
0.6
0.7
0.8
0.9
e>
1.0
\2
1.3
FIG A4= BRAKE HORSEPOWER AND MALDISTRIBUTION-METHANOL
-------
a CYL
I COMP > 1.0
I 2 3
CYLINDER NUMBER
1500 RPM
I 2 3
CYLINDER NUMBER
2000 RPM
I 2 3
CYLINDER NUMBER
2500 RPM
I 2 3
CYLINDER NUMBER
3000 RPM
1234
CYLINDER NUMBER
3500 RPM
0 COMP
^ - 0.8
Q - 0.9
© — 1.0
A - I.I
X - '-2
FIG. A5: CYLINDER TO CYLINDER EQUIVALENCE RATIO VARIATIONS- INDOLENE
oo
-------
82
o
o
(O
8
£
0
o
3;
8
£
8 8
CM
CM
O
0
# 8
o
o
g
8
0
0
o
00
8
No
J
-J-
t
^
T +
r
/
V
^
LEAN
hi* ^4.'
rtitt- »• ,
* t
4-
-i. +
"^
j
RIC^
- ^_ 4-
4
\
4. 4f
f I
4-
•(-
4-
1-
Jt
^
t
•i-
V/if.
INOOLENE-ALL RUNS
" X
*
4-
t
i-
60 0.70 0.80 0.90 1-00 1.10 1.20 .30 1.4
e>
FIG.A6: C02 VS EQUIVALENCE RATIO- COMPOSITE CYLINDER DATA
-------
83
o
o
8
CO
8
8
0
-~ o
o
88
0 C
O
0
in
O
O
CM
8
—
8
INDOLENE-ALL RUNS
H-
4-
4. _,
fcft**»» '*•
LEAN
4-
*
^ fc
t* ^ ^
*. *T>FV ** '
j" **f *
RICH
*
4. .*-f
+ 4-
•f
t
j rf
t.
4-
•i1'
•
t-
•f
*
4-
i.
V
"•-
t
a
/
.
"0.60 0.70 0-80 0.90 1.00 1.10 1.20 1.30 1.40
0
FIG. AT. CO VS EQUVALENCE RATIO - COMPOSITE CYLINDER DATA
-------
84
0.
8
IB
8
8
10
i ""
°8
s* *
o
0
10
8
CM
8
8
d
0.
1
*1
\
v
^
il
LEAN
t
V*
t.
yi-
*
JX
*V 4-
•**\
RICH
4-
4-
44-
* W*.*
H 41-
INDOLENE-ALL RUNS
*- *+
t*ML -f 4 *+fft
feti
i.
60 0.70 0-80 0.90 1.00 UO 1.20 1.30 1.'
0
F|G. A8:02 VS EQUIVALENCE RATIO - COMPOSITE CYLINDER DATA
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverie before completing)
1. REPORT NO.
F.PA
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Characterization and Research Investigation of
Methanol and Methyl Fuel in Automobile Engines:
1st Year Report
5. REPORT DATE
August 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
R.K. Pefley, A.E. Bayce, L.H. Browning, M.C. McCorma.ck
M.A. Sweeney
8. PERFORMING ORGANIZATION REPORT NO.
. PERFORMING ORGANIZATION NAME AND ADDRESS
University of Santa Clara
Department of Mechanical Engineering
Santa Clara, California 95053
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
R803548-01
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
Motor Vehicle Emission' Laboratory
2565 Plymouth Rd.
Ann Arbor, Michigan 48105
13. TYPE OF REPORT AND PERIOD COVERED
1 Year Feb. 75 - Feb. 76
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Methanol is compared with gasoline in steady state dynamometer test evaluations of
power, thermal efficiency, and emissions. The reported comparisons are from OEM
equipment with gasoline and low cost modifications of the fuel preparation system
to accommodate methanol. Maldistribution of fuel-air mixture among the cylinders
is severe for gasoline and somewhat worse for methanol. Even so, methanol is
found superior in thermal efficiency. NOx emissions, and equivalence ratio range,
and is comparable with gasoline in power, hydrocarbons and CO. Methanol produces
somewhat more aldehydes, particularly from lean equivalence ratios, but the
absolute levels for both fuels are low.
Also presented is some engine friction-wear data based on analysis of engine oil
.'.'or various metals. Preliminary computer modeling of engine thermochemical
•processes predicts both emissions and performance in good agreement with
experimental data.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Methanols
Methyl Alchol
Exhaust Emissions
18. DISTRIBUTION STATEMENT
Not Restricted
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
85
------- |